olena-2.0-75-g6931e27 Add a new figure on the main page.
by Guillaume Lazzara
* doc/img/genericity.png,
* doc/img/src/genericity.svg: New.
* doc/mainpage.dox: Include this figure.
---
milena/ChangeLog | 9 +
milena/doc/img/genericity.png | Bin 0 -> 135325 bytes
milena/doc/img/src/genericity.svg | 354 +++++++++++++++++++++++++++++++++++++
milena/doc/mainpage.dox | 93 +++++-----
4 files changed, 413 insertions(+), 43 deletions(-)
create mode 100644 milena/doc/img/genericity.png
create mode 100644 milena/doc/img/src/genericity.svg
diff --git a/milena/ChangeLog b/milena/ChangeLog
index 6aec8b7..31ffb7c 100644
--- a/milena/ChangeLog
+++ b/milena/ChangeLog
@@ -1,5 +1,14 @@
2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+ Add a new figure on the main page.
+
+ * doc/img/genericity.png,
+ * doc/img/src/genericity.svg: New.
+
+ * doc/mainpage.dox: Include this figure.
+
+2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+
Add new documentation groups.
* doc/Doxyfile.in: Use .dox file.
diff --git a/milena/doc/img/genericity.png b/milena/doc/img/genericity.png
new file mode 100644
index 0000000000000000000000000000000000000000..e9d7378ac31e7199708d7556def6205fd21eed8d
GIT binary patch
literal 135325
zcmX`T1yt1E_x)|40)li%3P_i9gER~!A>Cclol18OUD6`b-O>%x3@s_bfaK8pFQ4!4
zd6vt?f&~NI_r2%r^V<6o{y|Y19hC_6(W6J`vN94Xj~+eIeDvsXAo6qY5rn&48F=&9
zSw&jxQTZ6jE_mm;v7EHTqx*+{U)zck!ADRWWVD<gJ$f1b@c%K-;PAuOyl|0KkbJR-
zN{-1uG1zTR`RLJ`N3s&4Y90&wZJtoQInU)w0e6p+56*a>#U4u-y?L)f(yLYQR=Y@Y
zDB`VLGCYQsT9IUN!by8MJh4&3;;&mv{+M>Wz-;xh96T9Ywc2cuh$%deDL&8i73KQ~
zs`mk@-#;@<>)MkYv=!ge`GoH1;l4+E*5|*}re5ZCw&IJ(_9nX++rB%8<y`Gsh?LgN
z(iQ%3Yk6c@Hld9pi3OqeS?x#oPVRCV5n@YXB|(CBcdg9L&B3RbvcD$j?{wAM^mzoK
z?|Rqcs^`>1qKun%^0=}5<>wtw3FT>L-#J%hNG?>&CVGR{T=%&7Cb?>%V%TIAG1^4h
z&t?*thdh1)Udpm;v~DY%{oozi=-y|(_8IF5OHMlcNVHe462^B1&1<%WvJ_u@R%0;h
zqdD2(jY%W-me;oNS+zJ^X~RW*COA%o{3P4&%F6Gen><0WcyP<9u{?5vS7;}JZAf?M
z#E5YHTG)v2$uH|V2a-@`o%(m&+<jl3zF4`x?zq1;pb`y0Lcz90z8m}_F7otm@#lly
z*75GG_%+Axu(Ge>@7o3iI0Je7Gs4|T+diAHK1F@L$9&+eLF8LVtJJw9A3c<EPc1Gk
z$^AFrWS9FlT;OQzUo0K6M=pEt^<B&Fkt>g;UqsM7{^*ijlk7N53X)@;fv6jN%FSjL
zQB1-F3K3DvIK^W0rYe-=OpXXjakL<``GNr&(c=b{Ax!3s&ySHrD_I^VI1lmed*dFz
z-65DXB4?eetDvqN%xJ}z#1h3!okMaW=hjNfxmcPXcjqS!c=K+@NiU_nU06JGRv?GO
z2B+b|cfhfH4=lr(S5|ShR|J8958EZTeB2w+0Q4ZKHgT)F-@Q)Rn4>J*M`~)+Tbrw<
zY;$y5>v%gAOTGhM2XucKA1gz}`2G%!=@r0Qg?jhoBBRh!!o#>XTjT4>pYWE4&h6)s
zz90=UZ)?qLa?(wi(J;urOHmTut6lsg{>oO88cqCU@uamT>;qxQcGDYVwBJ>Op$h2|
zk`auO=ias|#)<DT-lOc|tW$4F>A&p$MM6$1c)rJ8{s!3#Iaa}N)8qP9%)|61duPIy
z+lrG2yefZI3`Iir&Tr3b<#SUz-_?Fc67djQGsT0Q=@Qk4@0vhELz9}BDM2sUvMg+q
z{Rx}6!#<<cGh3P^C4brOWZoq9Xn-pAorp+|4>65Q+?&q6h<CpWks<M~V1MC;wRJwT
z_iH(>=vxz(!s4Eb45BvCvt_hsx(r&p&RG0q$cUMbE7ykS5s8diHC=8IaFJ_C_3ugO
zEwmO@Qfl_vN^}t%rM2ce$`-R5!<2dD;#&jC2Y1Jh$eC`${tQr(YlNDLp8JxIpNxo+
zg-|@~1WKeg<=>F4!V<}@iO%gaDq#^(3La?P4QjM4xaN$!$C@&ELM(Hx=(5#=*_aI%
zu}c}5<20V^Jfr4`ip3_|z3s_yq7>@fxzLm0n*mLH0s<{<ZT8Ko4|ERjy18`rPsLMw
zwBfbmtUEYe`<`So)=5uayn4sOV*#l{!G0?fw|4kSC<7tnfg>ro_gVKG|JB{@_{pTj
zO7BEygNo4d5+-&5)cDLLrPD6sKqZBEHJc{Ml{@H<F6KlOXZPabbpL(A)+{H4FU5B(
z{!aP(ss}z#ijEF#ei*6ylI+p%r=0lstd}dZ14(j-v)6V$jX$>)ymv!Bk$AJ!)2gcx
zt|Gjl3_3~GvT<#VFFmV1(q$me$8MO_oMjVWzuBq9E4a80LMl|D&kq?qI&a$I&6tL5
zDP&Be$VSA<^=&t^l&MFW*T`ZbnP;n%K<UF5zGr;X<c`PX8l$P^;NYm$W)`Pxuv;E#
z_r3NiHxftA)Kka98%6lCr=h)k`BLzx2WMn-6iX6o*)GL}Kg%bF`e8$?KC><)Vd!M~
zlIW%SSe`zZrbqc!B6|6U;5~x(IH|?)cI_CC$?y%{(Lsxk@D*7sT)?;My@A_nxG{3}
z4rBTKu9f~LUz8Vz)#EU9L(|hedH;I~XidG8n*b?I!vUpG-Fyv=!L%l6Xx|aN-EAaE
zVz95JYe?+8L@y0%^@qqOrYR$vLfl{CWC@g@3hK6EIJAy?@d6o~upY(fQV;qO8kx^W
zVJK``v6u@uuW8fjLr2tu*89UB>m8KNSikO-?K3C-wUx<UymKW@ji|1!?&|6aXi8CI
z=mo1lNJ#kZEtBGqtubAp-uTIYdCkc9xbxc2$JTY}Y%7rsEdN!-n=B(OT>fE#eYkpI
zKZ?NPvza-<#)C)79+##PRtoP@w<A~oeEoR1aHCEc@g{49aLcyPF-pgN&*65>FTkz4
z`sxBLN3(FsR(D#+x~$XHQt<o<!p{W9L<e#1)jASLh!w}4-fEt5Wtvy0W=Y)oiB-c#
zY#*@@3s*@|$Tk(^j<3e!pumXPNM#Fq0qMgh&k>y4b@4*`?U;wBv9%tgye%0Ptrru;
zD3`sHw%2#KNTcTXOvz?+dK&l4Ynepg-MgR64NH@%SGDds3(n>#&vtfpJP(?7G_|xS
zke_U99Q&@ub%EmlR;H4zQH}qrtHcDUl=s%l_Frq+_$_`%v5nUvJ{d=g`E{eA#P5&Q
zki6cFuKH9c;O;j35VWNFLuYsWJ`vWNHr(#l)bRvqf3bRC-~FC-Nv*H7-qj`G+(77g
zySwPzg!V>(k9x-mC%Ffl6b+XfD`c50U_0!})cc#Ma&kf<yrtS#QjBK3p!1v^hD4(n
zv%l<j9@BE>MuD1JgCem+OuZ3b)rTUj8y`VD4?q4Yw0&VyFGfw>$ct6i&nV@k6CBB?
zZX6N2>fWZ$qfDr6qf&yaq34QcVq#+6kPDZs8K=*g*_|r(To0$^^}TY5U6psE-afh9
zTzvQNl(|)cnoaec&E1<~SS~Jtcgp)5Yb2E2yFbzt;xeXjUgzwtUH@gK)^aM-^^e@T
zt2k5ZWC{QA%%V|j^WVzet16Fu%&QSvJShquwEeGMYrQ;K;%v1I?S7x#g&GKGGx#ET
z9`WvBy`FPbxMVGtg8hwxyI03nu^Ok8t;av!E79AdmpX2Rk_WHBw!~p@%S29rRfcSF
zJX$Bm0ONa7T8;W}gMnyF#n*{+R(s3tH@Zjc-}M%rre<fGJ37X@wI~pVK!?p%F2~#x
z22DrW7B}K9k9>1xW<H)Du8#lx`<-GW!HUM2;nL!NWv-(kpwQ6)Wmk9Y=zV1pmH(Yw
z>+vs^ybmVMUx%qBv~Ad_QwYrUv}gV?GvhE6ew*XJ`BO9ZuEhVS`mh(H;<tC~3y<tP
z|8FCUT_SmUdR`mCk&@Lfl)msITCK_)tF6@UZ`f9x*SSNa-7qEh#v@G*vvGRG@y{Dq
zWo44!{t-C%6x`Koi)7>(mGgcjjgR$^1AUpV)6d5w7YXoLf5W!jl)cv_7!=pnhZejN
zZADnAhj<6EshQi^4BMJ?mCS1TXW?#I`uYd3U0zPko{t)Ol5P5jgQE#Y#VYj4$;lqa
zgS2TmIXt=FVYaiF%CGI)^p)DtasDgtp98k&VmIJVC05p+Wr@>&pv(J}&*e*A;d93j
z;Fi)vPM{?SmE=p&SWt!JQYly^j=TF2I^4*l3_&iEM@?#8=V>&oHI)pjbmG3swNqgU
zXO`vAwZob*AF)#?#MRj0sXG3n%1eL$!*fbIP=)h;9mTKwn8WmBt45+Qdz}4XHU8NW
z%S%+P>XS-aCu$m73UhAHa7O(jr>_^q>uHK&|7??J%iUmem9j~gSZpoBp0;(gb}<bo
z(t<XO`s=kPO}3LJu?E}Y>bL6%%S6`2rxq_#P-DnN>~2&a?Yg9A*foa(&C~U+U=dCf
zESE?4@cco=JDP%-sinPKj7$b<2MOg)fKgKd%?qbdc`t7)_d{d0wD(o92fU)5?RFqg
z-gnP;+WdA{J!@75pDBbzg!k|99}WspjSb#4Ufrtd+}ZRA1+5$3)!Z3tJxo4Vs7=`+
zOAJ`8*zOu773cP@QJzL53pep?Rnq79?eBR!#HR<)(b><B1<NjbYZ!DD`6c01JB{w=
z_<6N~F=GaOuInhPC2A3N4RXa(Y@KQAT-ENF<teZ0^aavA!isMjpgsG%FDQ8G$p=}e
z%lq*^P0~tykdg8C7wI*gsI9`nrbd`sXbjG8W$yJV4EXk-C=2R-^lM(WQ$k_Uf0AS~
zf5+x6@AhVIdK!vC{tVQ7!|`k(p)}2yjSr<Wybf!RA|oTsS*CV5vv<z+ub`xHMuar0
zArA-TpZojH1os9t7ug+$*0A&G7mjrKErBFA6^My)-ZcqbIt(u=Q{3A82}u4#!Ikf%
zS<2Vv(qw}U|4LuK@}VBR8N^b@7t8w=xgkYY$M&TdLLX4Ry&zyDE9Aan(juB7ev-3g
zdZtqwhzeKyX$!|np1xNQx~4zy8(;lber@hRVk1uY3Zd_uzT1C?yH%!@)%C^@G1@S3
zvod8V&^V*4mg5(mb2Z_5y3;>as*56+aUs*ePoW{kf*LG_L64nhw!?xV8<w#R&!nZ`
zaceXaBwePl+k3t8iINsc;_3~g9=g}OdEz<MpmZa@{`>okH4`f<hk$@w4UALDRUm7b
zM)hbiH_L^bR_VN8SP-VA>7?7rM9-k_(_bL;0w@M17MAp=3VE7tbe^T36Zz65Y7Af=
zB8q-~jX31l98xFyWLCKdKmX?UU(qo%!TTXcA)_t5x$IZ19OTBjMS{A8WI-m`=#`Kn
zl4$WS*`DJlhM%9`KRM|K54=*&nfoo`=22Nw*!rqHmM7Ne29~%;rs4j85Z#)k+J>Bf
zh}f!O=)k3p)2)(IcZ>f>k79Jr5T?s2C+nGlY;wk=hmL{7=<1i4x=}5iL?J;|<GF!A
zm*M*~r{C1gaLQ{M^+EGFA8T!kKQ1ogn$Q5L{+&*n?5&bC9NZA8x!K!yhSHj*^piGN
zc-e9n$Mh?g`B}I;_I$-Vf0+WzLaq2?#7k%LmhGmj>waky&-AMo+2-M}Ho$sl^@VY`
zde3=aaJyO6+D%uA=PLS8Hn7g@X;I26P?b$5zKGsDdViqfy54ajI6;5@^M|g*{m|Ub
zE`~}v=0lmfw=wywNwFJ?#*&gb&WML*3E`mcsz@X+tjuej*>7Fq$JFjSE-uL?-9ASW
zFqH-d1}=H-WC>pV$(K;>dCA&pI23TCA3W)!J#xP@L@D<$;UL}n;{K@n_dy{e?DTn*
zSP60$mO1%7N=U^eD(vyE39ka_3UqDJ(y8Cy?+>5XXJr~P3CMh$x$FzCSgTcCLKI#T
zo;=^pM&Vd49%dTxi`>1b_$_d4Id(&{dK0KYAA1`YzSq$OM~tr*i{;7aaZXHX=aYP6
z=JzN*pgit!b^F_tsd>d>SDPe%KczckVCWMmwTG&Vq+H;&?Iv`JC#E9Z9!_-m;t$8B
z)o>9?jL@k9q~q8xeY|6AIo8(L(t$%;nn2gdRpj3nj2q%YC@g-~mb+wbPmKXjT-`L~
zYaCM<7u+pFD!+4mJr4$>g~<50>kGyUDNfH)nbZeR2ffdu3L=C;sKwEGYu+{~*0aBT
zJK7#XXXWebyFXtQXk0bW)%B!Iy_D8dDEqyA2E34ywqKTBQD)bx)7DMgg5qQS4ujoV
z?DZ%^5-KROU%oufZ%u+wt7~Xn9=LZ5kBp!_0u@Y+VJ0!ap0=TXx;|7fNBGwII4SjE
z?TG^$2!wq^9Qq5}?_JG**=0x1*A!29D3ZTq{glV+805mHYobyTW`|qtQ!RU)ns<Z>
z)h9E)DUr3)X^gk!mz0qq%E{H>sk$6cmwNqH%|hZ8ZhXM+cWVy(v5?rWuA%bv7y-no
z-nXK!M0mbH{A0@mxQrYe{R|3mOG}w!ZZ5B159>$PD876XJZ!tMxa4vD_#9rU!oU~C
z^;axBZiDW5)|`jcnL4WP7MFd}-n;#6zMJEhOQ$1*iy;*hQ7k?LX8BECLpZB=?^~9O
zmhEb)tHsLHO-xM-id)SxCO$bge&ghBis*kHwqSO+pGfNDsA86}mQQDFX*m#;KVg}=
zHY$}b%Adt;N}V#CX5-ePr%Hq~2^}<z8hYC!-q-GVV$S_(p$Eh7-&3%l=xmAJ366_W
zo&vU9D+3D$M|9&aL=CsKha%5~79}j(ki00uXcaRJU0kN2mNhV@Cf%4sTBY*Agjw^z
zMiHhY$l15R_md2?|Mj^TiuX0j+vveh@ba3eI0+(gxtXzQsWR`m2e!xFk@NeW5P2%w
z&|J${H8Z?-EYc1(=Oy{f(JG&d=BTRun=!Vmn0&|dGK8xa4gI?ue+H%P4bqGD8;(pL
zIMvd1u7<1`uC_sHdSRWp2WNl&--`L)E09_{mtdL{BJ<&*k6pihk4SYy_N_U@ue;wr
zgHMddALH^z44|V`UPiuvUDgw$<gexNyyV3meX~+(T{pC}L{g_+7W#N7k9>84*0cy_
zn@;NOhB8DWceJJ;fHuwYYyqOkol&S=_AtoY+z6JImI?`>2xPq~Q@3Ge-H<o^5+@)~
zrQ4S<yd@t={5mme&Q<TA4WYksUf*z*sWgTcD^@xB`$$Enl`9UD1hz|nVN>Uh)>l)t
zPZm<XH)+?v=XLrmE#IwW)V<?+IeW51A&Rtrm#yEuMejtbv0ScBTcLgQzd;#QEPvOH
zQ*>%a?3~cNYYdMtuzS~{eyxb=I5z6IeHY=dbHi-OO6ejdhKEh>+xa(kGQy;}Z!WZA
zt0x)pWbg31b1R;?!3~<h^{L3#+8s%n4<Y~!Vbk`6i)Sgzyo%i&|MH2YJLjc~$mr~r
z>)Ty7s+E&5VQW@jt52U!eHHI6%*q$fS899=2;?gz;FzmPm_IxQis1ASI8PQcSt9E{
zIS?2dH<ZOo`n8SFQwRr(ed+C$1Y_OqiN%X3h+vQ}I>ozaslYeLGR0FxJ@gTY&=XrR
z<HW3^Bgcc5L*bz<b}&5Y8X%kM;$e$XTc=_&(q}A~NqVOk3jT6w^A~io`?3ig9UH4|
zZicihZ|8=Rfw?04-HBc7<JqBjl{~z>Bmm(|*`><<158L7pEa%8X#QA+PULu|3f&)I
zpy;H%d_%Aj5-XYeXQU&m|9*64$>rZC-WW=ku3<ik;BRtVLQxs3o-0pDc?#NYu>4L7
zJ2p>pTFK_72qZXNkuuza3%E()S_DZdwF0DywSKV;tG8(FDb1}-$~NC;qPGqwetb;z
zy!}|N)Ryz*cvkM5UcmV0feZh%HJ4(xpe&GDRcMYhHi*?o#4-+^jNiUf3hD5hz$A7p
z!IumXlo7p@Ko>otQ5ThgW|7%9oM9bL+ck`O?h6OG)kF6Ka09IR*}#@)(%t^@g3Pc#
zh7`Q;_Zfy!^Qky0f$A<A3*%XWnpXJhM9VO6`WqJEgT)cv^|Nwb4Lk1=2OTc{842I+
zDxFyPpIew)<SwtsAQ$g{haxO$w63nMEIDV>^;TS_6b2g%AIPNc;~ypOe%*hm_@{GE
zKRopNu^?hpXN*_bX5&RZ*O$bdPRk%RsfoYGsw>iMkL?7`^t#5nGRC^oW#UY>=5zkd
z`5!;}M&tMW<<(96tL38p5^+gES_RDa`LCU<&m5~aa@M=aWqw~rM5Cl?$Va(o-u}*E
zuW}-8&uB8_8kZ`-JKHLDC=hijo-zW{7Nl?z#~st)&s-#BrE~4N?~~DI<AMY{S&<32
z8hy{7KbQAZQcG$4PKAXc7F7+i9p?jWzR34=O%z)>XYa5dz{hqC!@h{+{ud$_qUZ2M
zX)v%#jGybnPN>WTmD4w+A)a;G(O}I=XZ}(o)mQnowTIPHQk!2rKuc*d&z?AmBTblq
zSogcPNszQAgkY?X@c7UWDxg{3hixf2x3C<A_Gb^%jv-Z3)LA1w%ke=D0$GO);WNMC
z;3@*iTK%_N0kJtAm5Pz2f8N~R#9Wugl6$10_sSD3Y@W!)@T~WxT|(xVix&SC%H0J{
z&UWtoi4a|duRiuhp?IJD-4iA9>!;@F7|Kg`IUWe_9HViT(AnoFzuX!Wa!y#Jh89^B
z%8<SfvsTkzAUsrvbhWjO-P~rG#_IX0T29u`@1c6wHL{9=P!2LCqSa`+Eq-M^iD-V;
zzfk2Yi7M=bH&rjW2O-5%LU7OhtN<~$o)Komv~}HpDLc)WD01v<dS|C7U{>XOPC$7&
zY((=^H#QCc{Yjq$5!&o$d5l_B6P_ZW3bfGiakX*^K;tRHuO)wFcay=6<vl-c51t4D
zdVld7!+$=%9Y*7K6ti{Wsj7&Fq0Nw#2eWlrcAaw^W^RZBP|?2ofsD+{Y(d&Hl7`PZ
zPElLRS+J1O^}OX=0!|SA=&mU;pZ;xAm5K79NxAB}Z?vC~;^K)3_DdUt*$+EBi8a1c
zUwG+$JUU+6c%PEZOFI}LrogxJugaigYFEvXB=()I9MR&p*)W^EZ^rj85jD;`{m9*8
zy|inLDMRb3wpPATONPY7i*S{SIW0}iuQ1#5IEtokpJW<@B#$%U;t9`IXkIvWnY~Ep
zi^UpH6kbK3KgLXw9{QB`1lC#T78Kx7WZMB%I2Ih|Xs>yLOqT?Sii-Ncgzu=A@&?(G
zJEI7Co+xseh$Q+RY{EhXdV#*h+aD>;Tb^X#mj`W?9gG7&=1%RN(gmM~wq<+K_1)2C
zdvuM&Uz7%E4((9={R)-ub)T{2hE<uoXjt_VXg^dfQA_y0b^Gv4ZX;p+!x@V3ML3Ic
zfumlpNo=NZwFAe_F?o}R;FM<+%;iYoV|gS6hPG#c&WomA*oC20+1Ip#4$F#!S&AVm
z=bI*jKF4Xk=);(oC7F3s%Ae&3|Du25Z=><xZFQ-X%Dw6M>Wau4uc#|h&)Qm;w&&m@
z!j?iURxhmrn*#72N4<uA4qSU>6V|KxTpftv(TaCHk$EC{%d{n<nNv2T(AdY&&4x&M
zc#}Wj>@89^S!A3jjExRyz>d-R>eQ<1ovE>L0Psr)yUY+zp`;$5X$%g^3SV?$fJYJ?
zJH*|*2uaqvsu8k)&~NYTq@|@9nVI!MMS~mr0@XIGF3R_gyI!(7Yz|<FGi^C1+Hhwm
zBz^(w<)}ATiv}rFK|@D(oYQkxW=Q%rV;!PN3z#eLBUCDD^CsNmRz3b8U3R5E7C_@V
zZ1>Mpy-T$DVJdVM@tOEmV1Oq!Rh>CSj_gB(OJfgwf?>|Pn^;QG)W<5RPoqSRzhXAs
z_=raMq@_cqcv(BgBLsEX<K*Wrxw{{@@;TpC;Wu2pVn^8W@Q8Pl9w#|pq55z6Phk9@
zh@_>x&iCk{BHS1D+{0o~t~SkWc!BS)<Wu~$%s6~ftxVmjPFqOe-e;93Ncu+excMTY
zXOVhKVlcy+J42`km>kI|DZrkHijA!S%1(EVt7nKNP;6x1?@Pn78*#&uQ{<7!pq)&M
z1R-lQY7@YZsYd`wre$R0YnP?ZKhJRJ>(-{Ou(h_1Aiw=14ZJPgIqED$!1uk~R8gC-
zUFEmMMh@omy_pMDN{Tnr{{tG_ud-sykMB%cZ|fD1mU#pR;chZjaeg?HSQ7*czfx9d
z)#;xpCt^sIwsN$D*1zK3CreV7mv$vvcEM1zyDp54eQxEKB`tC<n)k=HyT$3vRte)V
zU)jfW+zPT=LMrhMf~DvEe)p5dXvHx;$$Qf6u`2d-K{2NK;R`s2(FBKV^WqHSn2asq
zm9>HK2-ubsg<O4uT-cY^am8SQk8(r4>kZgB$#E|hby>ns&&$OrWEdG4Gx=Rn$4}<l
zeFXr=*xyeyX<ja;e~qe)vsDVSUFUs;A^7i629JG_dZ}{h%t+mO&$RmLI`s^IeRbM}
z`V|=}J59*np;9_!_)5Z;Z^47pE(=i^ic%8O6v&D>N=Jd=f`O;J{mNQON=jZo?QPfp
z_EvPa-u)tv$G!cxY^UVebRX`~e+O55!$vG*N})=Qf;?!%m|R5bl{F{)8s3tW-1W6Z
zpG_2HrV;hSOco5af8TlkY2dv$b<>VMO&6vKkp5P}laGEMRXgiZzW%R@-S_GJTHe46
zaX^He%U_-*`>77vM3z&#zKoolL#vl2tjBv>daxf;=Qs0Z`nK%&8m0yN_=j~Vip3=}
z)*ji?3`y)kk}&{Xy6}Hed!xgQeI~b#^A;hyve(Q~Q&Yo9gzk2*G`KyQp`f39cpZkJ
z0Pz!>+}|<V2jxmE_h-Mzog<mgKT+WR^+>IB=tmM$1f2V13VJ47UV46?`TpA3hafjj
zeih-lTbRP@d0gPBFMr{d=OdiMLyyRd7#!;g4@bGWzE-gBQ7Lgzr~30B7k9Yy7!MtW
z4wcj__VGC!J9sEMyUF}qY}0YfJ2!Kc)z=9mQL{+L@GukOzVp_9yphEpy^YeiL>f-#
z)A3W);L88vFS$>Eb}xkT!&mMV{t%Ujrb|!rlso1v<KKC+VuiTN9~>R-X&bM<lPAhr
zgc&E%N<&|H=SJ^*g!P@CL41JTf{RxtF=WWY{1{;h8WnR2C?C`q!s9Btg*ejCc3l=Y
z1(KaoGj$+taJR*8{LaqKXc!pI!0{qFrrAB2bn9>L2X3QhyiZA^r#^k}Qw+hMS?(*c
zb=q7(%w#4MC&q~^%Bg_csH&<mF*B#hJo6s2b#(wzhc?G&GEVGB7sV)vEi#rVaqu**
zNyz)W(rSvPb{6<a$CrY;E>}M!uKsVGG!Jb)&VH6ZUwYXT*dKo$`;*y#-AwEHz#Vx(
z`hpI<+Rjn@Bl+j@SG@3M!S)S{L5qfbp0_GuiDMToHiug$Djpom|6dlM<Nm~>q&#id
zPQ!knarFXS6!Z9|Gsoz<hcrkT)O*+C0r8uzXoG%W&y`(2icx^ta?~yVH4{o~{zm0U
z<<>f$RVIP*r3z;`W0YpG{u775`nqWki_StJ6HJdTNO~Y3>p;Im?Vn(5mVo=@+?<6>
z?l&Q^>dj`c4{r|);eqJh(}K!-kF?A1$*+Eef<xMJWFkxN&<6F;Im2OZ2EFC<loK4{
zK34O7!2!qB`#Z!#S-jNd-c4Flqc=gNkID-JjlvDSU^bAl*IZ<uf2;LisEGzJz8H7E
z*y~8`%vv69a>HWyDp_eR*o<aY^m_IZk7)lal$9^sXW9G1E5_zt1C0I6cv5l-1Fyve
zyPxPkf_N2!$Zxn;uT`?u`1AP5tF8X2v~>}!#tv*zi7j^_xJq2YcDhPnej!rAiny*U
z-#Pzyf~bQ2W=!X5F){xRE)i+V2(83V8u}wYuRS3PoadJgpD1Y<l0I!FFMHnJ-hy%q
z%8+s%f0kk_LMAM>jbj|)tFEr@mG{zg3G}|;;AbHrA@g-M1cfT$AdGOM0*t9F=O*3q
z;%R#QVh;?-Gy<J?Y_7jM{&zbH<;guXKM3lrcMQU<2}3BRrlv@Qyr5tb)15Q3wH<+;
zK~+eEeNutYATp-D<Iwq76MX=-Gn6GXB&2aW-9|nm)0SJVN|SRVJmt%SqJ7huB#LWo
z*D4U3A%Yl#zjtrqyH+h|d7080`XPVdxwCH8g4nnGVHy#d4du`qUhVZ#lZO2M6}djf
zz*pW9$>>y0CET=rr^T5b%nJ9Z{SdCkg0wKo-kkeOWb*q_;>+9nvwWWo;=r5V<nrv!
z&D}?;9a6cKmy}7Oy-T~~&WQ-&5V@U0g57X_6C70K4BJUlrD}O|-QN+3$Sp{8{o$Uv
zQNP`n9ACV^AmbcT*WHTO^Eqyfdne$aR=x^M`#SB5_Hkf5J{0}W6d&6UGZMb)Sx4=7
zd1lW|RcePNkl0@j2##;0+mstQIVG6%$DEDC3b{LB)UUtoXJKXC;C+Qy^%v2pw`D{B
z0ggVcQZ|?g?^_3va1!unD=TN6!Dxkv^cWbyzkeIqwO^W}1_y-O0NE#$<0>f4^uR>0
zbtwYytoj;Gl^5KD)?kj){AJcq6-&;iqCYbU^b!*A{>HVOq3S*74$^HJ%3aag4F4jO
zxpIo2zYLU^cx~^rndchR(HV?ch->5?q=7IPRX_+zUq&IW#11SJJd{V6T`pJV?#Z%-
z;7j!afwNdID^?(bXBQ*nzF7vh#pL<wI<0GbDv$}gF~^S=duSNj;dv*q{qY2FVSl%Z
z&NCS(=Bp!inD50VN#4c0@3jO9cJOyirpeXS@ySUQCEjgf_S~r{C`{1@qdQK$G<e4o
z<Q%rQw>5z}1~&gF*IcF9Fn<2J%w@TD88E|usVjsuM%F=R^f1tiWGNC)QWHOZ+a>Du
zXvP2jHiU5Sc7fSr)1EJMUlT3h^8@#|+4H}xpidbFX7Gk5S$=*#&}dcehx$^|(%#$@
z^Ac)KZMG_m&JFgf!rb;Nwt^<XVx1Qqo6b=<o&T$YW^0WRDF$g~M@PiUW0!b2L}hX;
zIQ+la=e((cbXVM4Fw@fxo~$-rkyo<cQLyV@qC4T$!?13-uW$V2c$~O(d!2N1sIEn#
zQuZndU0CCEzwFTK^G*M@$+^y%h!&8&WD%XGh<f5<ke|ttD1Nv_9P5KU*Y;2E6vD?e
z(k3&QLlLne7&{y;%`AfgQNunWhDEvyjzb0lmX*d21}#|Ek6ULKBuMAqLI$$Rj4#{m
zwYR4CT>qZ&gC+|4u8abhbFa_g`^zmUV!4K1j6))E<~1-J5sHGzu_o&}ZO#k@U@(24
zBcvXW2@k{R1I7(NAR)wdYv+shJVA2PTp5`|x?C9w`QLp{^U1e1$nSYUPXhfym2BUw
z1-u0zsY9>1mXRCTM9YkCLow2x!pKi(XlY?>-VIITx-1ruP<H&wADd18ThFbpRD`Q4
zMQs0$GZNxwscFh>%3gmEz8wBo{uwtS@cU@DfN&HfmgqUo?}S~ylpLlzlKX$3UR(wX
zjXTucmkb|Amnm6tA|d|OiL{4)o${Qze*QtR$GyERt>gCZy(k-pmwJ6MA5Yw?VFIq~
znZKyFXa0R3U@c$Kvt#s{v9YPYCzX*5+EeJCe-73xODV8Jys7mC?$JA5UKvcJALZpl
zAD_>n?ax@x^l%_Q#yM7H!dD>g@@gCV`<EdxF;QMoaosa}uz1M4CPWhqn&MJDn5@6^
z&W~WW_`nYKYEh7q(5o?_1C)Rb9Ygry>tJEE_X|`<Tz1WhIeJ{wn}fqU7V^shX<nyI
zWKf}_Vq%DlKUT~E!Q|4S2eh3R<V(FfB;9Cn4)7*j_hnb$FsO1vCv(f@2d{Awy&uke
zE7!n|%AX|$d-F=|mqcwIk}~ud9cQS|b|$aUtvO@V1^4LmSI%PT@z3+d;;`YkF#H>u
zjyi;hGeSb*tM7ZZZILmlO}AcIRfa;#@%xw6&E$or&8n2XNCKjP_s1orY5wq>4|Kc|
zJ1EvOmbzK(VV(*bnRmPmhX>eF1XB8>M>Hs}<1U_3@QC3DY0A(;^l_4g&UJfb!^*S=
z*#Nl*)oKGhq4xY88ZeQ7S0Zk`0I!#I7&osmN1Kdw0EQj-wbTD8hO;9Fe0FNGMLUDp
zCaW<3+n1-k2H4#ES+5?fUI<&w@O6(nwNd_zbsg;q1nuC#8*!b}lbWd6pl)7f0UZo{
z_Ce#uE8)vQrD|`4T~SfdeElclOP_4TV)2$`?hTw7MVX|12V9euZ&qSHl3x~E+*vm^
zHu@Nohqt`KYT`^<-e&$cbCe4<lV0?$o`o-+U&gjbJYy8A>1`4;_=IVl{M^9sda*-_
z`$~s>`4C3}wfLIEWYoXMe$@$@Ss|~eArQu^=Kq<_`cq|qtCX)(i07Z-j{Yy?v1`pn
zQJPli@XtHhq?<~2QzFKySL9h|cPD!;_KL?)r&OU!p8eN8*^yc-LW{I-Tae-Pj7L8h
z>BW1U%WUz6@35TSNLeqEhzJUfoQ#Vofyw>B1}^<Tmx{*d=}(i$76fLj^=1|zT=rMm
zGE_<=3ROS|1o(xq5b81);3a`<oGrJ_fpS=8T#Hwmof}@nrOtp^nR>o{1&5%Zt=rLH
zOeaWM4eUI4eggx@tQ|K_u?F%(0U`)zk^ZqPffDsnwbG9ucJ=|1=ACpjq>Jo{11j`J
zenRyRx({GV0fZGmFALlO@O}Es8&0b7c^-)Jn@*jlbAc8e+1)7BYuVmkDoWP!0;T5C
z`5Eh*gXI>!e47aRB5ro``k>V}=Et7n@-P1a(o3J~o(9j=&Den}vLwX{d-^TDn=3AZ
z^@jFhSnVY_j$|3SVpQyax&7t@?LE~-hX0q}6fZs$sJv1Vnp(f4U6oE2Afzuw$=iOW
zr}O)l!!~Q!=EuLxA*CH%KOG_>QPj?C?-S$5FPUF9EZb@!f6kIKJ4~<f)dnJ)gvwOG
z7}Hc^31lmje+BK=Y1@!1vZQdB%0ls#V<egwQieg9tJ$-&vI+;ZD*uP9ZD4IQ7JDtg
z#U*N#iG!(ogYy#cI8IHqwSZEwgOo~9EzM9eChlS}b#xLuZu-+=z2p5YVhRKuw6!U7
z##LETyw}NlddM`8ii?WSc0uU1&)e*99u!n?-pgAK_#uS}R`*m#=ttWNN8^J3YxSO%
zWdkv4D6XH~V-mD!t=Y-U(<XmQN~SVy{o0l1@Dg1Zff*}mNKd=pqa#-40|uz4<CKc&
z=z3Y<-Z6@)`Y9)p@kOKzevWeu7Ypmxg9R_RJ!Wwsk`fb-P9mwI&{(T#Gq;auRhZh2
zJ5&e63!Z);SV8?pTuX<0H7=lRvu6?!ypglLtihPR9J$e|@6(AFKrQ>*kvyq3Ih@A_
z-#+<<`EBR9I?6->nTMF0N{JdqF6NmG+MmfNnB6NOK0d1(Ri`^pN!QlCQZTyx`-aJ_
z+zjaOh|MIS)<8a;+3@i2_Tk-~UyY5jOmrEOET6LWE^egc1K&0<AP2Ia3aqpZ1_lN(
zyF~**LQ_3(yLemguJ(l5UQ4!B6B_~A{9rw8ZWgc|%kj;qm+W)i@!W05sQq~8JRXsB
z`20>Qn_+Qtx7_rZk%Am(f1hXEc1A9Ue49WDlJM>r_nmX!eV&bFEuRAYv;SV9jLS;t
zM599Kcw-^dl0%eX#{Jn~N&kk@Eyf6|@W|A-Z<t}z(1DQxp`Pup5@Wl|kHS9J_Y$RA
zB0JtmDokTSoug0vB?RR4Ty`kO73CX<@fQW2!FyL(UMxgIKRKDE6a@K%FHrUvJ2_24
z?}A2TnS$sn#3n&cRw7P)fK+GSCM*d>l2f;VI0TCW?D(mSaYqtR<Uo`d3MnMZNv_?g
zT08*#Uc6i)7i`mzAF`)`3ReXqS|w_PuU~gMlGJ}%?3<j#29ezHo!z~?Jrw6UaZSnB
zAe<otnF`0D^+MY{-=RdR(q*Xu3aA>!1kM9e($J1o@xR5D(u=W({}VxYKe`-V!jUqE
zakJe}Pu*LlZzGwop1fY})gXhvoT=org#XIC5q39N%$xd1ApaJsGV(*+!{xUHJ#_U;
zL^E-Fc54-W`A0AHQjuP&*qF*ek={SQdZ^<{>)J!JP8pOv?l5QH*(UxK9<=nwDQk-)
zYq4l3#}aQT%Nr0QGYEtCYRfO>rOt2uq)|NPji4-V{>uQmMEhYP@Wx~n+-_xevEOZ4
zySgTo=`?^)aI{qSJ{s;ZrP-iegAIGm=9zk_WR>Pypcg5D?-5-ruSU^iU^TX6$vLp)
zRJ2b6`3RZ?Lp*fP^u-#egO5N1ffR~?5&+XlsGuH-ztxog>!xLIZ9N1<riSYaGiIbl
z$HsE7vl|~shUMN!JE(Un@BL>|Vz;`YCMTkH?!9+8(6b#nnsvj{?{Tl~Gb(Tnq7!<B
zMBva`R4T4sFF3etUz%Izt;`TTn?Abrga_I&n)^q8iSd0o39V$AA2dN2(!!xME5|mW
z(dXQd+f)m^aw{I=;#m<V%xt;;x>4i_?-kj+E9(FO_;iMya8wg%*7@6i`%!Nf8>mnd
zNgr7s64ed{`s%Yyra-8165}9+z#XHR8#*U2tI^tsUubqMRjwlGH!?DMT*lSi-EA>%
zE(2K92RcloFUroqyB#fWwyv`tNcE@=8jwW`+LXL(SE5PGo!U*E%-nQ?NytGtkF;I+
zNx?w*kiCSc5Qfw)!fPO(-K<VQV%f4R8o`t@d~|#0zY2vsGj1nP(q%|82Isgyg+4LM
zY$vQtx>BMyfi&U2M9JDv_sQ=P5#q0`Or89)2gg~Ye<OLCUJf>Y*n-eYnt|xmdOeY{
z%aOirA#VKh;C8VZyvZMegP;APtJ;MuvkdZswewl2B)C2(aDy};;j=w#&N8kUuW0xk
zxiZ5kmi~I7AYaomOMk0JpXMh<O~>FNMS21!Y8-Nn=1u<NX6OkpAqrInO_iBhSu-{r
z>3<K}wE#Mo`(f>LcgiEX|L4!oY`;Q{WSVrj6*7BGMZmMq!+1!6l<&XwqDP;#=Gy2@
zaO96M(cYJdQTs2EUDMhs&!1JK%R<afHwR@e;WSLX4Cm=lhp+(<U#ODX(UCJ};Gd<C
z!TlQmLXn$yO}b<IUATqvkg|s;1L)}wiFc$=PSiPH^xEr&>yP}x|F>4CRvOwHyezw7
zVc7;bBJ_oSB+$S6#YZwj=yCN(vr&HVcR0<2Etpn#M_iYeU<I?JG^1IP46VopNTJUu
z>7C!mi21bC=TxT~$@w9zu8Nx`Jz=#URpAYe)6MlEJH)F^Y42TZ2t~?m-`wN-CzV0s
z-H;adc(KFeem+PpOtn=bH*1WgBk;|;m#xsvf|2C$9d9>cNNEN@BsigK%`Cvt|LI7g
z>h`j`GW1cn6ZPgd_0qIfItq{KlV7fS;_B(6MT3Zh9#;zUgeL04B#BF)9v}hOILa(3
zqpPFlCrnf=o8IC5KPD_4{!Y_Raa)?11=B_Fy8OWeE~ZpxLUj@JzGuE2OmWWkDhstA
zrNa|3u^H~;{s$FwUi5!#b#h3aJ>ki!`54n>l<SQKH*K6ZYa?YjGCXo?=I3Q?QqhiY
zKDTeW`}<t^kN&wA=I@`+Z(^?&yjCPM+DBK~aWjS&k6Lvxnwo8&6yuc$W2_Emo)9|F
zf4t)>clLY|WP(GPLDxIFcaAha6shM|J@f9mwfbUlC)qhH2QuX9mhOd#Brh+Y%0v%?
zrGRXN@w6o;niWokMw>oumSV>8=RC_gZDx=^S+<)yC9vfr`tjq(X=)esk#4IQ&~rdz
zsct9!{FR$a0pdkRfgqg%RRKy&)eh>NM2G6Vvgrjf;6N$>k5=0TxVyc^@#3Mc(k^l8
zMiqy~cTNDW&zODh)*5uNyRYzqn=J-s{r?I%R&ctncfW%h`DbV7CMTBo9k<$ab3s>z
zcFMcViS$~zLcCvXS8R4o^(a@zcobz^y)Hxsb*vW`b&*%Is}V9b3#G9MRz8uXSI@l;
zO!L9r8WQ|cX_))65Y78($JW%!P3J$LE``9E*N)5g2A;6IY=D<m&lEQ@Q1;KmbvC&G
z_e4!20c+x*mjZAQI3H>Z@xxny-KmtA%EVu!^>`SRsUHc(8do*;&5rNi>1b(nB7AXZ
z7&`MNE5>Q`!4@D9@yoQx@tqCwK0iB4tNseblk}|aHvkq+gSD}77xZsPXeh8`T30}{
zdZ1XSgS!cP3v8xemYm<__A+@L13)I?=g&tX!1n<^g*=VP%GBzrJ#IVr30lqaEC)&Y
z+nmou-Uz&%U19#8B~eZSRjJj%a`$T#f4`TB^mc3W=ya=p$C5L~$udL_+1F3^hB&!i
zWC;Cwj821iEeo}-JG<QqSyAt&_Ie&vj?l{b#9!>&OxpgiR%Q}wWlg4qs@0SJLEU&)
z*<8ylf<LVGxl>j(jUG`c2A9KR30xiTdfS)C*RGm|$pQ&9cnW;*G|Id*VQJ6w@ehDv
zpX}|m!PV&4?^iiakZ_?O-ExBU;D$BnLO#z}<E<j-@gs%wziF3A0oW#R3bNBN8<qcM
zOtc}Tsq2woM@hE<-2^TO0H;b`3;5dF>{|;>O4)*go0|{j14z_gT=16fsg%wTl922x
zr$ALg1Q*)5l*>1l-1d6`Ed^l}5&^fUrY63=zP|Yedvey+e@~=iNZ5ps3a4qU4f(Uo
zL5jdUyGNYH+5UQW-H)$HMAX<x>4BLxJTEWyIIt(th36y&J;tR{IHPpOLJj|=opWSF
zkAG!g>Vj}gFZC3zEEY4Mk))kRW}PAqa(J_tJD@r}GxLLM?`Ri^c>(@xDA+OY0GE1)
zYe-O(Y9LP5><_VDe1x6zH!s@m1KE%lb3W?CW!kWfeTfs>W#O@|^-?BQKeRc%u6pJT
zc?G3P*PHFS2&I!2U4h5H%{(g5I;&+9+X(w>DmDohk?cb?7lf&P#&Ir|;vJsQeg|<G
zIQ<ut5>U!OI2-3eizNm9HFWapMP*Op0l2dB9XJgj#0GE*Fb?^^bPu3QK2TLQb1X&0
z#R?hOTU%y;Ifdti0b&Xe7n-A>=kZ&_vO77b^$%H1PEHAs0Yuzp@AbLwz)tY|f7ETS
zbp{F|mYl&#qZQ(ci;FM0Eq(%Sz==S|i9nH1943T*5d^Tn6Cflf=aG;duoh$-PBz`J
zI>_6nMq+%J8divc9`Cm_j1IH9JMg;8i)a1A1{YSizJG9e$-A&MJ%G`O7&DtJ4>)?B
zjtehendj!V{)Ay95dwiD`B|rm?Iqrv*5BgZKKH-4D!!~Ey^kopQkUV&r5}~lK1vdQ
zoY=G{<JoQLYsRhn`behG;ktcrvD`DMw@}0!VM<yk<+=3><#}lTTulCO<DF2KrOEa{
zB!URnn7U=wwj-&`2_``a(z5r-3axL6a)Av-YQN40vIDpvGvU-EU#8BX5J7;C4=%X@
zOfa$#EoH(idM^Jm>wgI}zw)FcJOJ%RPETvCM)sS_kBk)*6kglU9BuT2bq1pvnfE>$
zxK#%{+>0`w)tzkb10zezu);#>zW#o2sV)R_L0(5t!!ajuL>Uw!^-@RR?CN{`ex{hk
z-wQ$*084-{VKmRmKn$tC#R{U`LZ(oHJuI|zYWHiMIk1ueV8a!mFy@YrNO}CJ+J;d(
zPx#^D71-jBCbU`GP8jr>ohL+*=o?1d8AaAidRzWruTM*T<4%#m%VaD<du%Eg8*6tv
zeq(sfrhRaBlbAt9)wMN|Z*_OKZScGKH<Kt5r0KJHdC}!2P6K<>F73dxV&r_m_|?$-
zCB}EKi1!um@7U}b&z@Dk?ww`Ut5tVtjvHQ$byDFiA77eoXspu}FS!ZPh1F2u?{AMK
zh`W|g+w+aZkW^kT!1~-h6=FU$ECS;)qN<8*>fo87t1I64iM+V^;NaiAJ&r=&U6a)D
z9R>WP!|S5p@%?=!z^g#XNJFTXegvJ%h|mede(>?}|LpB;0F4B?5mW_i<P0^JBgri%
zKEK;Je_v)6f5-UJ@@9~xG~4dGX~5<<i|o{e1g^uc#@ySx<Ef-aGy$jpN5FQuIjOXw
zJsk-6<3d*7aC`sj#l~_Q`77?+Pj%YB$*8x58ROC&Sq^IMH#H0<-!|y2*KySUk2WL)
z?$~7CTE{=J8cQks<>svG!~Ri*xelE`LBlFj5&vF(E<<LIu|8Kezr;n3#MTe$9oiwo
zs7ZXejPvP(R)8flPnYs96w14+KcO%l$ER~470#>k$TNyr_tmxUxjCK_r<N4f^(hM<
z{K|3e7Ydz-51j2X;tYKiTkX0Yo0=m6;TftD_s=wDX>deMhSu*(IA+KUDsnES;PLqN
z=9BCnC+ZR5R09d_ak-TOt2T?|<9VO^;hi}G)K7qF!O5mcem~@y;KG8M5LYo(tVE1+
z0fPQ)eL2Fu>40$onQmzK*+<y^i5Qp#si>%0PbcMyQj9=wb2yp#=9u-K9aL?wR{^t$
zKBjpqw@rIcRV`O&M93~AG|Dk9*bSD}I?O8fgW<{csO>^+_-F<X7+<=0IxZF3FM3&t
zB!><<>7+bY5owv31)5AmtjXUr_6qANc7h{ux(X!zOTOBpLsfooY7CTM?vUUzOj794
zQaZ=)xB*8we6EESwjZmE7UJRKpS@RM$39ss+w}KGopHWeUK<rz4&6QR^OInXAXH9N
z(l7it->xRDLOVR*`B8{Z7PaS8!i~2yVz}2D<}%kAVFCQ{8&6R+vh;*>GBS{*`3MBd
z3;%Q+O(DqMqE1<ixraoAMEJHSo$Q}9IL@3CiTK%JG^`ec)f(p)_17ruCOoEoYu=Wi
zl4%J03newB?_YQ=p9d|NvU)|r`xCgOhGQRDQWUnKr{w#4dsHbp&c_vi`p!rX#es{b
zcb|<PX+RwRt%*LA0H6{63vU@AuX4@likGQx{QUA1jNxD_c7c>|nfkB<sqJceR;<q^
z6;NjnrD%RW^(b66Ge|*g9RlzKaEQRwEhUlL2Yc}$oD6^&i0_b<1KbIQA@EQ`$^BAU
z{V#`$c=Bt1#$Q8!RN8V6m+Q3vVFDN?P2%olzXfBQ<IE7t(ZWVO`r-c#OB8>s?R=Gz
zOFF{VNF*qx^(K`Ev_E(D(s5&5P9u$T<rtS%$duwPnje^W*#)Q6OzpRBak;%J4&(L+
z%$_BhbELC~bP-<`|DY><a_keLIz#bFUHIpRi7nbGxuv#UxXai+bm=1r^_(tc^mc#U
z#|7m9djBVLha7Jir_-fwk$go%zxlIjzt~IY<;gyAcrM(1>n@^%TQrA7G3&SDmzcLm
ze|$j7nxN=O40$ia#@C;Ou3T=*C4ujgApsi-%X-5SyA@*ANkCnsQc~I2Xz8N`CQx&W
zwY`dt2l8e;3&LI$y6mrHeRI;%aw{rgK^_1^WXFXNBtZY!!m5KoxIXpYuXkv#$;o#>
z^?8tlK()XdR>#1C18)-eJvG4D21KLds<dMWNEXnLO*}kOwad(Xbg^&WNJvdHD(~se
z6~b(fXDa+2E{|3^?<gjw`vI^0yhjJjeE<Hvdyq8-7hMDnB>f(P;D+vH+{10m*n`D}
z*t@$sGK1uij?wcCV(c-3VMFqi|Dog`0d<1psVXyG|FY0ZWG0$|oW*~N+C#nbY50<!
z)4nCgvvNE6BcoU4t$kQE`|DMUps=fe-$C)<pZpuOqQOs3z0sN`wQJ~AO;ydA{dTYX
zdVN}kERv}flnJ&j{(Otn_RaEF_W^mv?bWOcxgitxgyWZc38m`AVgiW!<DV@>;bMnB
z9AT&_vz4{e>wh_S_ztj9X1$o5Bw%5zt!^0vi)(fWLg-w9i#a1bc0k&H_sd%^)*~P8
z9sDo%2`bEkl?zhLMG6`6=W3-h>;lM(aPF85fRP4wP5=(o_uRs(@&4d;nI-6n{TXcQ
zSLDKA_Y;C}E|^4eq3hpE%hv%0=<DWUXWy{k1b=tne}5OC3tV9A0-+B+zZ+gK{<c}Q
zP(>t-gd^od_>a)AhQ1<Qe`#>>1PnAFm(O7*8$ocW30Sw528PgS98B}g99E%x)X8MN
zgO3l1rj5K1FK?sW(}R;+X~RTTPg|Z4#fSEbq!DMxmp(Melzq&bvpNu-p7MicX36N4
z%PS}QD+WqR&R>cjVI12z%Bj;C*;Med5*zs1q#MGj?1-fIOFp0a0@Z2B#!Yu8)8a9X
z%lOnejTi@%<=^`<McZeOvxkDvT2O<oJEUFRv~FNjv%o2=0?pmPWKyZS#}OzSU#Y
zZ1mj^&0Ht%y0!8!Bj}%<Ny&UC{N!;t9be!`A!59S|BNv^)OFTW!oKoD%W5|1ASUf2
zW7E|Re-Ut7SDt1nF&~sDbiPtGhKE}YPYQ0fd0&Ao4W13o@SC@v+}x6Zn0)Am3!6X<
z)-*R;yRnji=Ve|qSY^`Rh*<Fk;s$)FxrId&yngY|;USMiUrc7)a^MkiB&aID5;QXE
zOdXRe#0Cl9w|$Xfv~2_;h^Fx)*tp*;QgQ@8-ImCrJS*F)$r@#7*%_2xAW)91-&=Bl
ztm^;o1<ju>yzfmi21;`fjiOg;*j-XdDs;5n-QUUYM}&LgJ}YQs3D+bZj-p5`xL$1R
z;9qv%I$O9pHyJwR;?kD79zc8p;G7qQ!YM~fZ#tj#uvxC9-@#k$r}X_SuE`FFR>Tcp
z8xkf&!FjJ;8-)bEgnp!s6jPiP9<MvRVBKG8gEZ{vfAF1=tMfcP>F7KOtypldPp5hd
zi5gR3-*@2`+8jK;nD4V%L}T5FvwYrSYsq?^XFRj#+I~4|5B{XW>({R-k|4?elai2d
zwmhS5RP|AJx|7QPveyi5LJ_S_f~^dgE4Y&q@aC%~(?c>G+);b@Na@Gp%T}vIRXSWk
zGBQxU9QMi@oItp+yyMObj9SZf4Lj1URw^_}`T<#r{iiQRww!=#m@YL6u9|>=qW+!>
zzqx$Q)pW7PepP>z6E%8P@Di#K_GA*I5FlH&782*CC9O8f4OCOzYCR=5_EX;n&+qLc
zL!N&XpWPo3XB%yJMeMYbW4fx?L)clgsHd>}%n2|w?(tu|2&S~%UQu&Yc~59144~u1
z3RWK?=BR`~<6GSml4mQD|1i)^eI*(H9FHSf69?A`C#Y%iZOWSL3=_y_{68B<E7^`k
zbJYK%Bqhx*`TA^s{u))Mfg$fDOR#meuUB0B>7wjq7W#eYW{<+J`Xku6R+&XXJ^JxE
zi<s8PhX3Lxj{>a|Io38)w-edePB4A~s4Da%@QtCFSz&1E|D)-<<FRc2zateQo3e_K
zt!y$%$d=3ABV?CkuLy;ZO;(v*_TGDD@9e#@$==WBy1&2YkN&xHSFiItkI#6o15n3c
zN&%G7K~`4wgCG@3ASX|XeFFrB_o%5E02N{A0P@hHM5aW`#KU7crholFa74!rH*uu^
zJSyO0>rQNOH0cd@sshIO#BKVQ1<Xww<0ZF0s_VOxq!EKrEdzyKqJi?s?P6VjrG-p?
z!~C4pjzuQiDW5KrCa#-PP|)IBztHYo4OK!Zo&s@!U5lAiJJ)~oO18WG^_L&FuidU6
z`FLbAn!Ijds>e7daCss<u%si-R2;n1yQnZHjuXP&BRc_;E;o8&eD3_ppwDxk>l<SN
z^&u<T)$cM})k9R+W9TY^mm`rLr6ixtUZN31>v<^4^kY>`3JJFBSlU_#4e5HA?|tHq
z($jj{YF%JWON+@alvFb1&+jpNA&y+BSU)w77_c(_Ox?zySa@y1M_MJtPX4QW<;L$@
zC;l(_FdGO-7;#>pq&>zcoYoo6F>s-3SuN18pl|}Z0n<9H-(1_e8X5+`T0ze{Juu<C
z2hVy87*J5<;72{cluK%W9SKNhMq+FLt3XM->poLFVGc@S`5>oc@nM!onN~#-JpIk}
z^%MXrLFo>{Y)~7IjcKMd`)AT0F-^1h8|JUbiUjueD*>1DCHV{R{XHh(%ru_74J8*J
z4&PW^qBWm8{x`LJ?7e<P$aro0dM%ny{fO7hH5<{r@s_M7p~>8O4~JGl+IEMR{NG^9
zw=vDmb1qP&c>Z!w7++?5rL}jl#QrGoMnKw!-tVI79eX%TTE46ty*#l<hu#!kuN3Y!
z1>C)7yHKk8?RJw_a2byU#Jm$%9L}zb3(~$fwN>sUv+qf(5YpSEGRrxhRQLRz=f17m
zm3K4s{MrbXnS&O8fb5p^-?7_;3;gLgrjJS}P9CQ_v_Ekeull7jzY;b==h2|Im?rdR
z^#dqSKqSD)#r67Jao(9QrTjQX&vldM8m#s}c1!LO%&3FxvF3J0Mn(ldENI|PFJ=;+
zw(AuJA{>k;Fy^3Ydj7y_u~OB>nPu68WC5YKv1^?8la7@=NuIFsFHj_w<-YvL;q}ja
zlqOYAS0)?p^9}Ljs7?ORHkmHzQx5xhvs_1mzjq{bx~rrbzL57DmD^86kA#J04R2wW
zsH<}fYukr+FAosE2u)@T2+I~OENTApmAjOXV}XuKpof#X%{FX-D?D7f`uD?51<Lj)
zw)R4L%ZDb)QBq2TW2w>(MP!GyDkB3WDWn^CPTM{Af4QEa>(+$F$K%h~y{oLeCRBa&
zvGJmId;9F{Qm`DS?#9>ao)x80+V!p@HY7hVsxi(9c81p<X?$Bd(ENt>Zt2^~yS`a9
zY+qiq&l~L2|17z0_J#%XAFu1kMhhpYlUfRKDVsVeN6B^3Joe`G#ivWY<8QV%@5za2
zx3PgP>i+VMzMevE<Ij(86eLcboksHrIP+x7d*S>RC#2kGShaKggBfK!8qCS&o?6op
zghOvjnRGF8;hOph$95M#F#N5dk_<!76*rlTzUkS!*x10gODmTRc||nQ80hD-t;za1
zk}3TNLX~C!tvAy+NeGlr9uKp-6MjvKBv1M69YFfHMl@8)$8M#Ez+ar-VIzh*Eyc|M
zse(Oo1k>jmLqn7u>A&P4-410RRC{(3QUIWrI)o^&h2L>2{!ocY)nffm4r&3|+Oj#^
znv&%vAD#+eo>aD0r*afY8d>b*j?NmziZmw)l7x65<04H$|K%G!U0q*;BQbr&1cinu
zrBjE5O$7O{wArJ<0%4A)PDj{p(qUBK6iee#gUC4@+kX=)p7rUD;NKsgI(JlOldf%s
z$rJxZ+vda^b@WQQzICl}w|rsZaA_=5Mz@l*XPmy=xBm&^G@|KY*G-X@->gPd2Qo74
z#sv*6;$MYbPo-NqRMq~dw)oT+vUyJvzlo|LRaCR9iJ%mUTY5M))7N%vajz|I595Zy
z@qiZD?~hb-HrbLSa~VjL+#>Ra4TTh6lc<{uF}|cT>0^5Bd>$pg;8`oW^7@dPgfL_Y
zx%jJ4yL-WBXDHb69_ejzGWtigBj=;_4WDt`K2y!i1eF{<Z;8mWKYLoSQ&ym30AsQr
zE=5dCOdoxL25T#HXB2b>out|5Ny0#4yK%mC9LO2rge|>~1>G+QV6R`l4oiULb`)&@
z3{T_xwXhCAo<Pj^??xUTiSH}AK)K+2R5cT{;yh{FM&%NfM9U<GDgQCK|JN3dN@nDG
zC0j?NH1F_}mnll`E8?l=QqEtOXyK07`?nau8FsfJ4(06hX0|F;%jzE<P8(bcZ>RhD
zN-FM>1WRpkm~y~6eKpQ7pXx*3^IKH-H!rW1uWQY>tN!UPSt3WPe!Z!7gI&aFhAd2U
z&wt|NxXqF1y_SC%)gET<6Z6OSFj!@=%4W4K@RtqTL(dw=?3r|j3d&z)6Kw_0pN(Wh
zNi(f>lvg9H-)(F#Z&OGUNJ;-Djn~gE>O93f_;L9C>r#P_MD3H1uTs_yGmD3V8nxF7
z%nLPFD#b-Rqt5@{(v)<a@oldAx^P_RS4>Ca8S$w0tm^lJfO^yO4MMd=ZrW4bn>X+5
zD8FSh(57+ISKXz1ZZOwL9&#}o`iIW3M1zGsOZe9St3+(mCtv*WGGj^d80i(6ipqC$
zVz-%!SY;`O&(6-^Y>bF7^zMCl?}}6b&k*TEsiv-Onu<kf`S%xzk%h`x!@x*FH7x{<
zpY9(!-RPJ1-6O#MG+pCbz#1q{`<#(D8G*#V^SOjZPe%IZtNJ~mskxCYDrKKmx7Vo=
z;wef{MHyU}r9C`6^tKf%bpD|IGG+Rsk{4lA0%isGJF4%in9G>MOrl65`}L>fN;>Ge
z0=jC{_K>JEa_91Ct8V>EFVSjJbxq|gUfoY-Oq^=hu*;^Vm7V|ooJ;)nTK`a(#WIv7
zxLxU;MAxhhpR5#y;<)zl;*K5LvNaQZ(6fJSZR%tBHMb_}E^oh082;;Lwx51iANeHy
z4p-@4qNFp4(tWE`YVJR(U3$1yJs)24u&3P{je22q!rfv;+It)G=^+o6QQEk{?oKa}
zEoE-=Zwz#wW(HrEtkey$p{<ug)J7IcJH1J+@wprPDG$lM5q3mk2n%O8EiJxKQQN9o
zY_QP(sLJgnT|X|rZV{@+9E`ML>W!qelu>$uJ7w_{zzZ@xr{`}Bk=NuU(qTc9E6_--
zs)`f)`C|c(*J`x>KahXz`ETYNcbF1j-2>w^Sf8h-^|q?_6yPv{I2B-7q99584KHjo
zT}{oO8mtyK6bZQOsfVPf_jF3|N`u8|fzXA?clCo-?<*m@WgM`-|5+T||Haf)TT$}>
z6DDlBxbAspYWRVq*AiQlZT&%*>8Buw(@y5=vXw9$^7%Juk2)pQ4YSuDdFfZQE_hf{
zy&vvzQ@Eyoy%sP2CWGkNOVQo?S&^;Q&-17Gqb;{5tKFBB1V?T&(NIb6U2W8TzS?Ve
z{h^H`G<A@L+C8oi<Febt+WdA9epj@RQRjWbJX-a<If<&ECjUlrhewWt__LiKZz)$f
zWh<q<+W*_(lY6hchCi<5u=L?lknb<v`z)bO8oXB0b+=b<w<_lDKA?%MdDVnVj|?o3
z70_7zX2j+FoQe5#79=<6<b`_baj}=vX57WAl{)O}iw97ke!ln&#p*$%j5!YqS6=SA
zXHi0qklAcU&>2EBz=TZ|0NmNClxb$jYlH9jyIKOt?om;-6tcovm)ssh8_JBA^`GtO
zSJln#Yii*JO#g<>xcB?a!hc~r|AD2(N;((<KM;MA)x~XJVx49s5ZnY>kNvY`V!@?B
zMEZdgo%edBRspAMaAw3`=R}6E%y=z$;;k3|)uZZ$cWgKJEbB!r&vakQrwKkU8qB{n
zdgI)3mdSwd2f}8w8(V)Jed{bt_boO}<3Uz!{=$sbQ9?wRhFR$JPB^gz?{I6{-&Bi-
zvlYyvhuuW{6NFt~Ir=wcGRbDzbXQ%m*dMZnWzN*^t+UC5R{OI0WK+Cp(dN;68J3?S
zRrguOa5mEiSuri6Ks|lUV?N6NNAs=z2gb5urOk!)_OuU#p6UHFh+X`#L`R<1n1$=k
zi@O_OoTDzIg8hxzmPxT6K2vat12CFXs}a=#`W6aatJ%?l*ZYTu7M!p_uWoLdKpQqH
z_*EkhMCk{6`zHbd$Qpf+|H2Xt?4UiR&zgveF&%hU0qe7x_;blbB4yzqTG~sif+5+>
zq*A9Yo`*o19RN@_=S&?d=Ldh*ks_1(GESRIjRLHK?Z$}cbCreXXRK0oUrJY;$Lma(
zwFA1C-lzBE#r~UyulYltS8Ikj`6S65G|%hjPMa(KakLCH4L|Sme$%^~``()3ou!xD
zu;eGt>Z9A$uitQ>|KeeF!|Lr-;~&W%9gVI(dTWCdjQ^OYq@%60RU>AsQ+M^u>}~-O
zLIz<N#}QAx(8g7^prJfDtZPd5q-A<*%7^VJJ(%fRSV0R3ZyO&$cG~1EDTfS4TaS4T
z>BO1lbhghs?&}>Lalh=&V_%N(+?D({uJ%^U;wSImpKdzaA`G*rBJK-5mAiY60Ut_k
z1%JoZWC%Ar+M06PvP)X{9r|#8e}4z5=pFzi8v5@_Z_*?3AR|pu*w5_O)$ULW{lZ=7
z$b$g}96#^)vxl>#=o;teX#to%+pfF+5rF-*GgzIsjI!0Tg!a0)x_WvDv3=o`Jd<gk
zoqZ6;YYkGmV!Fw=T}+t@DF5Shb(NJJV?}RbV`6}tPz_fcJ{cVxY-$gG<S&`R);vgO
z*2fyjJP;wraInr4mZ_&5%53j3>=5DEO{ZmXKXUDlzZt5|!+?5q^SF(>(>>|uy_Q!l
zJiM&e|4vQLdJoT9;AFjIhza9UXNW)$7A|xVFQ}~Z68e{vk}5j&O;J>tzrXP+bLeFD
zTcB0ysObg9Tq*yfkDiVOelJe&USm;|2h}Cw4_!~{-@m`ReKhr4_}hp*rB>fhfje&s
zG+wK#^9?-sp)}F57uY4|V5k|!!e%K}KlCAXu4^h!paT6ngJYU^jIm5)s9d3OPl2+J
zERrvM+=P6x`(Fw+yDD~w^#sr@;C%do?+r{jVE0y5UqNCalST-9NxZOY6bNm>v{Rt*
z73NPca_s+(0cs0=`#Ju*3G1?g{QPI!<giL7hIs6!U_7tinyLa`Ju#7*Y6U-}O;bl_
zF;ZRcXGX@$H}Y((tnUEB0D?>a1(O4f0M8w--%Q0dW9Z~nbN`+YOir-NaD#X@qH}5a
ze!JFcLBAQW%3?oX{eWdhLfDd<TyR9@V%nCnva0I9IS^hEt4@vjnp@V%mt|-56}_}t
z7JIo!T9Q=+M)rSW>UD=~r1GcwuB;kc8Ux!7zxrAWFT}L>l?<lWL!nyRt&HP&tfVp=
zs@Wvb{4cM^e5h`B5@Iw2u^4yTH#f4YX=YdAz$!N_F6k(aO%ub)wwLinth)Lb&FH7v
zm1pxEI|~z*G#B45RJs|dqmqPdaVB~7ZqAXb^Vqs-9IH}pQ*CF@W38)Ja)0TeX*ksT
zzi6VHWJlL@>!zp{OUKeh%6G&Wr+cx4r3yW@XzNnox7T<56E<FokZ+uO3`%4zdFTCp
z)Y2FK)G=5qr?tXvMMGDYo;qZ`fNyGQic;{y7m$6L_x8aV(ACvt>lg=Y?(gcVp{c1K
zY|B7Pf)wf*m%A>L2#Q$i?cD=uy+-u)+rH?a){~<9yS4`B!7Sw*e6M5Wwhu$4#H0TX
z6^vN$0G}-I;qSM@fR+VYu!N{3!+)@ERpX~*pboM8RCmwvXQQ4fxm!wqUxjwSekNYh
z$=jz3e7$w!wY2>ff0Ak2HeXacMV*Z4dj1jz*Y0=1oi5Vc<&k&9nCKJb3LX|EmD%Xg
zFZ&!%`jZfTGVhVk%eQ}Q*TfuUg4Lryc8p}#s5w?Eb$FXvU}_Ovh=igTNjVtl>hG(y
zLUNl1)0k?)h2m=dWP5Mjv3GFcu$WPaO<X~PvM0&VhH18EQP=8g>qu$ArO4Uy1aTKE
zraU(xp5BL^^N-p!hextI3w#Zb5*h@=_V3a&4VQ(mX!dnPidbP<upjGUO^O|p&0Xi@
z<V;qjSYqlh=7iA$1fuO^wL3RlTwK6aj^bm$f;2EOk!6?JrRioN*AvC24fa4fnB70Z
zW?(h3451L|Sy{4=njVFtP7}Nxl&i3f+EtDy8k&kwucg1$C_Hjt3<gaa5XjWX2X>ft
zB1tQKKVQhn%Rd254fu&WboRRHL~Mg(ky3PE$`K5Hey@jQ6|C;?ZvuNFJN@@G?E^0!
zX{)mno1+dMa@>p4Od_S4x6l|+v(Dj;dpi|g)2P16509noY3h3>zF8SH`k^{#XojV9
zcsmo@l^-A7(040nRj}m{x<8llo<o1PR>7&<?S)T{+L7}E_T3trt@g>hc9M@AvgaLp
zxTKcbO$FysJik@`JbSF~`gDf;OxeTG5%0jB*}fL1Jil71usUd>%FpcgP0##)kDPuA
zUU8q|g(gQv+(2*=lN)&0e9f8ovpo=jC@g1e7dRpqqR1<7p?@~+6Txi6J~?cm%JO70
z-9b4Bkw`7kCAUzl*@~b?mepNsq8>Wf3lY78ETn-bR~YaCC@p@mIT3u@uq>fOyqvMk
zrPc}FwNNwIqX(roL1dYm%f3~2<)P<sVgk?<sLiWagtuwPgHRg*P|WaRzMn%b$G@Dn
z?tw70%Ku&CfBEtyV7GGxZPW|Ly9@lmO|UJ1K^Zi#5DgG2WuQq4^$^&i=(=;0o_K*z
z))SK+XPW@N^dt()!a||Jx>op0+9)zPRYf@bw^mdO$dTDeGIvz%N`H*|AqKnNrwI_Y
z$@!^wq@E<^@I9{l-$p%RwpwB^=Z3MSp0rNb;2XDt#ZxyZT-o^Yc<iZO>%*=gJJ*x8
zpj-a<trmVIgSzbn%(+PhSTvY21Oc6zVe|6M@1M|FPNKDY3Q@iptGeG-=~9~<@lmLu
z$}gNd_np&TiOHQDJ!3`wKn{vyFGV4$wkeue;m0rECTB!`BY5ZNa_jc+Js<vhmlAav
zjJ?;c=1xzn)_LgE5wDIchl<WP><PYiMBdXfmt>ETJU-O)$?P&+HbjO+XbtI(EG@Q&
z3q};bVuXaWn53jP;6=lw;Dru=Vqjp5DtgoV4~S9NvjjXZtM6Rt=;);8<tfHZJbLt~
z;@uoNSX(*^0h&fhC*Xy_R^Tu0ZN?oCfe2?I(HkEb`Hb%kvup~<^u?BA5xlwD%Y!^X
zJ~Z_77>KdwDs$DgRk$0YYxn7Ul0@;I@y3||XT332^ciftIj@-hE%OH}1go!D*xDA*
z^A{tK-(Y@QUmsj{wx=GEovQnwx>$zKk){$$%_foE<s8meKfjGejgg4fbFDKWuj4<w
zgW?bGuaOIe{_>8Iw>LseN6&F>^hft!>dgWba))MK=n4^*bTA{U8A0&t8b`x(+kj0q
z9Au4#y8i9VAVF8`?b*=8vUz2`p)A2V&+5j4rI!bP<FoQuyv=^eIdT`-IO7}8#;kPd
zDpLx3es|2RC3$M7G_6Cw_&MJ<X!alCC;q#64L-a&JNNLad^ZJZat4-MN_H2P?aUQt
z?s%q~Z{yyqyAN93^(RC;)F3wJB-?gKsYF)Bv^o?`>7S$wEiCMot7V{cCt>Y$C><!!
z4Ckv!mq_Z@DL2^_6%@kc7{DO3e{c{gMW-dk8TrP{++5nVR%I2iCek+ByX55HDL_5>
zYxBG1!ER2MCo6D9enN@czJfS1K!Th0gtjWz*=a*DVazfD4S-LR;bu5uObfb7+~%+y
zboUP!pYXk5=`PS1GLdBhlm+L%5LFO685L**UP(vlfb?M3DPA(=r*tZkJJz`KYwP2b
zyW=kxez8k%^>LF2fkAJ-tJ!jv-hkK%Ee2ih=a}YRbV6R&f3Ij=|7&#MjdMS`4<wGQ
z)jHRnd?3P(Qsl3`N7L#wKgxWc$t_Ig*W|K5*y1_PqdOdfT==x%cOGsWaFnNEu+YVd
z#?SX#zfB`o;l3AM*zlHGXYRwCF#n%}6(%+J+r1ofS_Pd_QEaU2TC&EI#9tYlf4X53
z5SAN~&Fr0Y4!_;|<}T_wsmq<Jxke;T6gXP?!@cnv754G&5pO8FJ9f5uh>kQJEv8a7
zaryeg-lFX}e*fYJ8&;squQh+qXBR0qybw0Y$m4~`jJ!ve`94~OF@si@j6=^DdN}}k
zTeW94007%_o?c#@p^gm1!hm!a<!*-!8=!d&R!M%!KMem~F^VfH;z8j9cfW}yOm`)8
zjVybo0NVpDEVNU99b}+=Q&U0fl{(c<p>GnM{otHgPgqTq27M6h2FZaKCbiQnnv1I|
z>ivSc46Mt5V)4vzy}3k&EdlEP96#T%=_ZxhI9MzJtcQ|HM~DRy6VL^Y*dO$nry0ch
zNGGUw7XIDZ`a{>@7VjcTY*Y}lCBjWXIJBT5{%Jglc{tO@Ic+1BDqy6Z%!|0Ybctp*
z!LrU1=UC|7wU{&k^V<KFL|%>3)b}1=^4$+ue?4Hd<6>~6!63udJ#QURKDY0!dred9
z7OO`j)y~0ofpekoz2M>P^QBEsuR+~9+8{<Y&GY%$UZvZaE=g{yd!HW(1?N>K<=a0z
zBQwy(+j;u(D&EsQ`g6UxrvN$awXt}<zpd;Ek65fiofHGMsn5`!%1-3Jh;qT!L;6&m
z;yG^=T(90sAe^adHe{9Mv=xyEA$cCK##Z;)1zW!4d9SeIzhy_=+1!ck)+1F5Zwxw`
zVM9_3v|)>b!VTs@yR<Li2oDv_&J9+Tn~o#)C`=blNcE=m+Y}XMC00^+h;F-q`o?Z<
z@xV#L_6R=#nIyJ{8=8O%ihKbP&&^x6fESTTYedZ&?-#$Qg-1y1F>%A|a`{U$zsnCx
z&<8<zR?^o17NZ(vIJSf_yhZT%0saM%@PK!@i6{c`1!t8mTpA=%X$a}Ms4<gRDY{XZ
z#C^q{-|`o^JYI0Yw5KBk1VI7MT#dnfqEwn899;gZ7wV}z)FE8_{J#blRdXY0YfaTN
zm{&?eCY6?+?|G>H0TL4a*qk*qzVC~dV<o+NZz{Oi=iM#KR1I*PsE-MEKJ|>4|F^O}
zw{d&!)=7ErLvpmT5Rc^z-mA)XN>&;p%DtIx(KuIc{ncq%4e{{~b8jnt$sj?6(MB?n
z9@Q_+=>vK$gqq$Zx82(OW`qSB0&k1HSiC88+`Z|`IQ+)>M_xdXT}7wdvqnu9Jiq_1
z1;8t!t?p!YMt}YaFMnx?<9k76yZC3#n2=A5C6ty(orlGzjH6}KPiBN}-p*4pUKE~L
zsJTx;x3OVKkj_n`a=Q${7a{tRZHNKwv9QSI>c%DkIvh5Lp8Hh2#P;?l?C;xy%8J4*
za4rBshSdplR5P{ipi~10K8vX5bwLbo$wX;)UmyOP-ozGrU0r7G6ig5bLI~CmXhiMC
z#T_bnBPXqVwLgVbGOIJf{-pFX!f<ajn*aRD6Vy9^1pw^0>@s3qXeJ@1TJadh|KR6s
zwlKUWgK7o(>};h?V84FsV$)o{TAAFx6hCc<wY}iCuUkbT#LINp%i`_qbQI6_B1EmA
zzrvo!PszK$-uUrF(QKZFgPB9DpIK?asAUpa;4;SPRYK?OF{ar&v-kTc?<r+yN>=0j
zem^`p|6hf@+MPQ?JE{>@U{tD-tLW!5poytn)$~Ds=5vfaadP0Rj;XC~`dn4W=!xr;
z=_4A9r7+_R#}Lv%E5qIWeL{{E&#-Uh1@GhVNrPq|h}P|KRoj`}z7)(`?fj_68PTEM
zV;^)(U_^5?aC(c*+Mn*3PN`zpm_VT1UmNrj3u6)N%uC9qw`bJ%&xSBdNh;@sX1+1!
zAcY<s<Ovihux+Fd7%sDwME{CFYK4A;8y6Jf`Jb~#zXm7_foz5mjMqNPtIMKV9so)W
zR-kER=m_TkasYV)3VdAjVy5yFqnR2jgW?8+d|n>jh&>#*UiU4#af7#}_P$MqjdS@7
zZ!*O*zC>`wC@Lzhv)3MONw?X;N+76N^NJC~cUp)U5?cg@Dgp3cJ3@+{U(zA~4T2-g
z5W+0dphD37KKEY}axd3r^LiDZ)#x)%Pfyv&PXblDu$Uct-t?dvo|>TaCeNPwbCz)G
ze&dV2lDZ9Hree;N>Bp4%qI$;>m0X2J)G6BW^ESa^=pEUy<i$Ws)AI6(d_DX<`R@IU
z&S~*vA@L@e8CpF2a*xD`6x$r3oQzi10Fn+BTi+>%`?F3sL3NQ6yw=HUBbb^r!5mbx
zb%{<K_`09FbYw5E-eq^|uzqIzqyFxNzx@$XbV#D{yJi`I?1NuEnZ#LXQD1&C@EsIb
zq|^t-?2>j8xniQ<URQgAC*z;mkRW=(xqXggQ|V9-am8ekkaV^u#(GS`Vy7_YQbC&j
zQ;VR$j9W$|=}~lNKY%RIl0b@sg5QHA$BNCBJDtE_+6<Y7fP0<A!omp(6RYX!SonWf
zNngkytM-VM%(%%nCd%+;jw-G$&kC&LoP>Ht>Mrd;><rUmKfOrNghGt~v$8Bzpx`0E
zX`N<k9y3R<Knuzde>dHgNlWfbhxr0p;{rdMrs1_nnZd+Gm>z!+n+J%~!qpkHX>cXS
z=5sE8%d?Nxc8q1_HtO_)V37dmqqsM-rV4*H^d<GlLY~_a3r99m{U`6invCd6a|hDa
zyl1)M6IWFq%W{}B6&u$dXP{uAzq55*qcr@NpZFc{myC==ks)6T$e(dW8uKR@{qwhr
zr}V_i#vVGg8TCt(@(*eG<lwKEe$-bk#-BmYx*S%--*2xcE@cs}>t=RLrE)<ohDn;~
zd^s5Ej#+Sz>8%i^DyA1*OZecGpYM3mo~0wh?l|F+FMhLszEGXGL;2ai$JTWHezLU0
z8O)Z^mFqtRoGRWwLWtyHMMb>Q9rLNLTsl&~43ylbbez23l65k&(TEp&PdY$~k@mg6
zt^RkeAw6d9*LMNARJzq6TAZI>529x;(|?r<{IGt)Qrm7EX3Xgy5e64ygo@L8MTCLK
z*KOueTsLjn;@|!hA2AwKwhl_Pg;dVak~}WY?bLOgY5|-3+`AYZ_BOaq#0gRm?^01&
zODDcn6j0awvg~delSwO(zLpX}lg7<-L)QK&*ay0At1P|X8LfbNDU-JEteg<P>!jnz
z#|)9lpPq<*G`$(6*~+tU7t{eD6n;qPuahpA$(drJdA^o+xi;gz#G-%FgX{6<(kuvO
zN|}r7OHXR2@215~MfF$3lG%w{Zru*BFTWM%UVFFllJq$6RKIW~Smb=%Pj*e}ntKT8
z=v{x#y(PvIXJ=))^toOli{`%}SOL!y%4=#zWNH1SqaS=5Rb!F9-jF*>Eqt?$lyrvt
zia>Unl&~j`V19XdR0fM?m-hZ>Gl^Y3`eNm@LCbTCP9CY`rw^~+>!a{j)=PL~G1Gdi
z;*Wb_nBy$3m?}N_{(Y5!%_{oUx~HJD;z;^BIS(=9EmvMwjEt9CDA&~H_V-7qcJ+vX
zxBcsiGZ2(0eCEz@M@IlP+TieT3jjhukt~CjlsK@Zvyltnvrq&gM-4T2I62)NvA6#d
zDWjV=Y$0^M(}dD7g3&3tpOK7RXZFbR>IhmsP+!2J7)vtQuj%t-^xDQp{rtUb%fe~&
zy>oayZY^vrS*2~(IZ-69)EZbeEy0VS^<N7pKnW;m46mL|PUSKA0ZWF$vft5`KE3w3
z@N$YfOl{ieye!Y0Q<NG+F{X{aWy&Vir?+3=jbM&83U)+Iodj&6Fr#1jtPMjE#_ufo
z4g#$FiM}BS3GdY2(tde+amC%HxiV93v8G9M*<Dn3K2*`u$7fUT$y@9P(%I6a*?-*Q
z6ZV`!+_ml8$w<j$Z@YOO3b+Yj9J0|#V;TK)w&^|8wVk!Y43w!53y@n!e~95L6Y73Y
z!|ju&p7iWy^X6k*P0-Sd9au|M{Af3~H}~;z-P?|hAVd!o-hSHr?8tP@{;HCbj?JW2
zZ7ni(C&wU@FCd1orI`}lI*L3xIvP-8*r5Q<2lx~k%izdJ#=>22FMhZB(0!Knb!F`U
ze#pK8A;AW@dVif@sr|ZMDcFVt#tax8M@R#l>NbnIAnr$;X=V+B5#;o^t1yK#HOuTq
z8a5M0gaB`w%I7YpKv;$;u63DS1Jk#zdi16WQvqPC;ldL6<B>cP#5{w1zjLK2Z{584
z!B4{#y?TMaJU1({ghKB9Rp#?tq3;zMp4ZNC`tYpKE}H}wvA=sZCX61Vsuk{SCi=0j
zX6;<uSUzGoE(`1d^<`h0lUA;d6P%}aZIjeXYrkUV%3;uQb{V^)$(Ba-qWAxJV}ti;
zms<F0`%2`W^T6EQ(!0;5y_;U37xb;EzPXv+gynZZxxRTia7>8zp~hdb&rINLa6Tqh
zV0~F^k~YNz+2P4_nZv-+q@(3{Wd1MuJ3Bog%eiqnXd0W(BLiMBeR8dO|JtDii5JMc
zf|L+RP$&W_01yIj0vWFKB!Fz6vTY89GQ%(g$}SjP0rUhs2t>5O`md<mGH-K)O8xcf
zg`ZsrPdZO!ceNhP_W~jk6ws&}Ji5Ae7Q+=`|J9ZXlJ@TE=UnrU;O^#jH4#!5;Thah
zW#@Z|1FLM#e(dXaZZjbNG9yg)GsxPMX0q3+a+v}m(+3}Z-}<Ba8FKowLtUKok0BA5
z|4<OMB|s)Jlj`DXk$idUh1c&~7ro&|ar~gKO0v`g>zK0}Cp*hy`ma+SiY(@P?DJ7;
zxNbRZc{KJEH%ndDo#=e|{{2Vcc1SYTwUGGNW7?uo`O70G?z^fAGq%^-%;_~&+GNi+
zT0`asi?3abjn%ad3QL#04`Q{=rcD}WGW*TKaiftCgUIAhiM)h_cD)(V*t@UUL@#9G
zh5<zV9;+SHsK#o$_i&Dm(D+Ltn{4w!C<SG3jbvsQ@w>>w9L^YGrYHAGU8=Q4DS~bS
zv)V=^r>JPR|5pj+_y+F^bW*7JfQh0;(XukKHd_ry?lcC?DXcDV2Kvn+K>P*rXfU~e
z830vdfJ*>Ur>)c)pf^Y!A<JDagtrOq!n%{VI^+sQCSHPO7+3o@S{Be)K)d!2M*cx2
zsXF)48E3lot9%~ng*69(93ss6A*7kOVxF9WX+j&?=artRgPaa=n29(MDc*hy7+hJI
zde&=GbT>e(3H^(m28KlRZ&9ONj*<7DUfW?aFp+Pm^sp{}k^kS5=WR*WDae0usM<cz
zGodWy%*20c5pv^s31ZgT^NJB^mG2OWM=?XkM6WjDep#qLZ+uAHZq!<?KfvhcthmF4
z>BWxcVK@J|Y|?9Fas>1GmU?+r)vbc%9~RTiOK0UhBC<?lHH9^7k(UGWv}rGY2lQ#w
z;L@N+MOk?d+`+$;$8#hdFlBNI4Y}%kJb#mlJqnwgNA^>?V*4ah(Ze3r&(8<zb0ZR<
zRG4U1Vnp_?KyE2~Upgbh0_ql|hJeApenZ)2{-hDs4DIQEq$pGrbiVKaQ30^v*nkfm
zB6a_hG4{GGbHRy$o?e7ba`MX5V=ypHDojQjLSaEnS)(^{m5KPtEWm7(PMs$eYVLyP
z{SkmpK%x`5eyRL|LjxVArw<WI3hab9Ex~3GXU)4!rZdy9S{II?b<-lvgzKgZJ6Sl7
z;RqEXneFth&*)c8VGClb{Wr6!=gZJLYkqcKe~qzf6=`<9Q8{(C8FFCZUFLbr<nd3r
zh)v$yqq5P+$YW~7@hQ{g@O*F4S(mO;&@#EHFI_{yW;MMdhZoKlKb=~hLvCH0f2x8G
zA-n%>oop0veI;v)Vebvk>n?hg&``oLriy{yiKQ!<E7rJtINxI*^2i8_`M^%2F`Zk~
z8$Yb2;;z!RNNV|s_=CJH+oM}O^WQ2BNxBft(n#_Ev)C!0&8Ec2uw*F-d<>Tjkc&cq
zBA2KrnA1_*$_K$5yOJTy1Y=HVu>6O&Lj=Dq<!2xyxiW-$A;7#3%D#Ui3_YOCq6Ak4
z)OrA3@Ics(<tv8r)##MJn;U4?WBG*q_k)@2dq)V;EH>@AgcvY3C>FBCNhnv9@xU$e
zO&Vq#Vh;>Z#br4OlMn_()&4{75S~4M@L)1~>@;)lUNO0;U-kD?4NTZ|25m{#%1O+v
zL#E6gwdY+Iq|W}Ys4M2L9elLR8NN=zDRk;Z?`;5`MD<~+v2{$rkeFiTI{MXYbkr|Z
zZHaGIJ%(;OtUv7|qQB+8BfsOxbzbJvL>h%lR3sr%ad9zZa$Q#A{GgfZ{N#9supmMG
zEdjnj?b8F%y*lBtK+hA$>*xQ9n&8*N^Fi3^cjM^RRBXLZa&WsOB|1y_U1x<-3O3hm
zMo38wWpiwK8^d(?r*Oq`vz&DK8p32Julh*W4q9>ZqN4(oU%gEXV0#r<=zkz14bviw
z;KHYnW4^|SK720&X9*;zR8Jiqb8)?>(aTU_e!|V|69FYW<@I$!f1g9)nw`emlmhQD
zdiVI<jvoVH1=iHhqUVMV+tb}ON7J&f`2o+u%F6opLakZMz5|IDj~t4{4yBberH-lO
zpuDsGMD*vYFCgH8VHFM%$iC#WhjI`|xJDt3gqBNcsMH!R3oB_%7gKa~arr<RC6NA0
zSa@c#OxE+A{Cqks=aBp6_44jzjFRq=dsH`2nmMN?qU^{>%MdHzuH{NUzZCuU5+R96
zL1PvFF7)Yrr}zi?X#O~2ZpUPatwEa6EUajh_XTAbF?Ai1UVITPrG7;CdFh4HvM1%6
zmzU4y4asgd_|9|B#JhdZj;l6$wc7nmMtU?VwaL-QpO2t)s}*<C<mVyFd28p&$>}8C
z$yT2<;ipy7o`U?!-!n(kxavE~EgN^^Qf$%b3X8*E=lN(-yB<g7tv0orq^edR4+an&
zU$nmd95YE@9)`u&9l`GUZ`@!7wrUtOOT^MTERiZQD?(rzFNF)j0D%Na4;TiZCM9@U
z2z~A7V<yH1gE(pweHbFt2q_>RO8)=@51;P%F<?v~d;5cm>laFAS*@cny5AU{m><!^
zUa2>WJ7=R5+$?$3jlll{#tA|I=rlw`L|Rk`=7EikY-*XDFNuQs@J+ah<h5JzQ>K@d
zMQgC8cYaFuHvRq&Qu+|fT?<NWmZyufp<fvKI{xO`Fc7%b)N21<5t8(a_$KF@cTzJ^
zUGIk*Hz2g%6OHA)z#qEsb^7A<miwZR*IS}v0^W^t^6isYgQ-B%dzdz7F>M8>H#SNO
zp7N}nXZ5u??^odMY}e>7swXHuCj9yUi;<W4{IGxn5gnv{VRSxFN+-tsHtNIgz^lNZ
zZT`k{UGz44tt@V-$pynL>Fu91BS*?utds(18qJ$Qav`|*PP%2)b;1-FyNliO&bOIV
zU$8BA+)tA+3NR?aqgUfI!QtCW>I!`fV#KdOL2%LIhe}s)lWJKj6pH|3jvDpnTTrYR
z@kHs#Xq!l@{>8BV6E3bMfaXI)FZl~J<XBAL6@9veSLlA`07TdrbO>OvzzBAJUadB(
zvRYKLQL0Cl$o%NLF(aWr5G1fQqcn^B!7m_yh|vErH+L!teNj{Io)BDBk<<H0pzT#?
z9TKQneYjx&+U;4cO1rAP%6C9~sfu==35!sNlr4tEUNUp7&2`*A4)z{K308KrSTwS-
zaK(>LskhJ2;>vIbo@!W`GPo$+<Q1^0+pDVH-@h|`walq!f3a=LME|0K?ri2tp4a1y
zEWc?$N`x%JB=U@N0#m&HsAi$sd8b>&J)}4Bn3-qHN%iaIh$Di|gY5;cZi<>U$<x#b
zDgA#8>+Ea)I!r3Bwb$<7OiZekVwhAs=@O--DSO9ql8hdgcFMy4b1qJMN5xz&a$txp
zQl=zgq^qUE{ij-p@&w}3S9NuD@LWSk5Q;?v;ldAnO916SPIa=k;@`Faw4`oEoH{cJ
zK&^ll0Zt21l@LFMvQeYp-KHaVm{lqGY!*B9FFAn<={>hp_a-hd=g25I%Z(T?jp`p7
zjr$L{P07j%)U9#=3jYqk`qFfS=Jz6|%(%gKJ>tP+Vxw+;AchG$8k9bui`zCqab(_@
z26MjFglybqLeiRJrroPhhLV~?rKjoS$S;fQzYOd#|JQfb@(KOd)y2P)4HAyPt)fak
zs-_cLjCGe|(zWhuu5N!<$el`9L`RMu96j^jE*Gur3LK);z5XU^a33(Lq2Qga4H^Zq
zNBDUYC*O{7<6Ng5K1xpMt?N~PNV1+_{mPa<GUk4;`C!%Fb=MmmA0MmqwSOPmX54nu
zr|qBzHt+wIrV-=YQGOLr?`36+HT-T`Ax%KcMDH&y=9oQJDh~n@KuoLub+A<Kq4KXF
z76&;*`G;RbBzXG_6)FhnCAJXK0kb3EqA=N`0B;l@0<=UMlpdh6P~>p0$6*n8gDs6x
zr9x@VH<vgEv-2WoA>_5I4zcorpcwWmFk6D{ta-t<MtHD@1Op9HwAfO)DeqX8NF!e}
ze^E`yu;QIa<}7(>5F=o`c7d2sIm`5ZF$^00>v`N3ZIfH;*Q)++WE8Y0iM!9W&`R<9
z+WAc<)!X-UYOo0~!fF%3B139@FvVr*yd%QWTtX|T8@47fiCQ!-cZT!{3O87+l>0Cl
z^dl=g*yt38G~>J9pH6B9-<L7?%<dc#ZOO8cFcCyiS&IJ+hyK^CydD<@lR?VunZ{*S
z3ECkIvZW*Q-Wm&MrH7pLbRZ;#6-O->zBjNO!pUCKbhWV6*V0OcekdxM*DeRc3djqf
zU;x1Ys+_9rS_&q5c!bwb!X7yJfLVXWyWeWBD=$fG_{v$U)LK7E=q<CLY1<D$50l?~
zSkW>fifITzQ?bZ>m5wTzL=Ix??;A|XDa7nNc6~?g009B!Ek&1Hb=jpPkv?BK%}-e*
z^eb5Y-kPd^al)%u6S1OZr-UcC|7(DMXH`G0Pft4L;`90r*p_<@=-Sng;?!kymDQ}o
zX65Nr&XbV7nKY5SRi<n`biD5``iM+k|A0lO@LC>igGH{pNE%mN!(!1@P-VQBly(+k
zbr)m1r&{-{A=giXQKEd8HLq2imxeW9T4MfQQry9@cJ21M5>L8!GUtyj3BAyp2Soum
zuNG16WLRS?C8P=PC1z1e;)idDwHQBeM1#vsMu-|UPCR+i0O%wt$roaW`Wj*tg?2GO
zo)4Ou+8tjG{J<6zwPlNikeQOm_*arSyk?T%W$KuCk1_Lo0f;b{-dFhgXhxxsE;y|)
z%t@y*AE2RZ@LD<P(n*NCh?A+V4(3Q1fXHC4r2FN5H4e|PE#^}xNd=<&+FAiafl7#S
zoAI4(52`+Y^Z$f2eE^EI1d`aaP7nP$qeJ|$_Y9Q5n#-p%K6Bgabv?=XOi{hCm2qNm
zVM0jT7$W=VxAk2*BY|^+Q~G7g&JhhfEAoHW^47v=p3T0>>{i0pH$3a1tIh0j5~-4z
z*sZAj@yfrsKrC-O_oUFvWa_-ZFXDa5P>5`y(GRjWJlUfk4$6AV*SELd7#X3PN%8@u
zMED>5{wFQ#Kl3<EsMNR4CDi^0VlbH2K&$r~nh&%Xlnxbyd0-!MKA!swEL7Wlu(22X
zM)_#T>W(FH>ftGXu!7M4uN+h4vu$T;u&K4gnWU%0C4|H1R$QDa9TFh>m@5xL0zYIV
z2Kc->s{>Nf_cqzSP}3parv-i>kg2E&ii^b|J1QgNR-3KXxKcNJ!y(P+dJA{ViAMTc
zo0=7UCal^Nga3(u4*^mN;dNJyqPO|$>>$<m>#c;w4daE5v<BHH!ePB03=QqOBDAYG
zyHY+*7T>O1dO0T&1tiU0lQr#%D6og@Mc($5?Xu02HaMNX$X3&1nA>nt&5iaYCic#N
zw9qR*9d#J7c`Q_yJ0~01(eLSyHDulr*<Pbs8bf?^?_}Wgy8oSi!hP?a0nJhwf>|o1
z{{iJ4qC20p1^bBmIaf%SBQpUIGUM(5bB2Pddf*jEeIZ3mipgJ`_J871Fvx;QoT~lZ
zcFoaysEx2eqhf?XY5{lzp+5!~)mFM<xllHIApO!1{hFGPzYH4+xCsnYy17)uJLwqK
zG`i>7K)?+m37~JqFgHUGKQmzFl4E#*j{z1%xF$o2A)F?<6SAWqpm?YarI5k^xl5%E
zvea;b7Xn%;m6bO?jrk*k+xHzyJnOp<x#V6cm^!ZhBP!LGo~#58m6#|Q7$1LwckUKu
zXBvghmhZC%yUez=`>O&Tr%BE1d$+pfPH?R9AM7}vN`JPVaQDe3q8jzqam|0*&|}_^
zUUh_Tm)xcH4>es8+D%%v9>Qbr9Z=c0B0I*}VLnd4XwJDccC|@|Y<T0mv)QRp+P%>s
zbiZPUdQ#43R?t6QoUlljywH!$<_V=L_M~x;ilFTtMpg}FnvC{lK$NB$=e_rD-*Nz!
z4cQ&Rpj?zM2C`ajWA8)=U<+W?bJQNj%Dq27wiWhN#B6}>ynh8F|Ht$h3n;swefJ8Y
z;46I7OO3jn2E;-Pjuw(X5MIP{=pW1>QP?6m4bvC#MI}FzTu7HaT3bjLP!VwEvqyuL
zy0Ws8F}f3c<vLlCko*cz)ZbiTk|PKa`cTnFrvM#^KvQ{_^9@}oo|76i+W~b<x*JLU
zxQ&&5-g>@p{?OM?fALW(-j7csQ3jf{8fWZv1%VUKPMaq;DNmmIn{?0HAqnSvZp@rm
z`0|O8FBClT)4=Bax9^4S`L@%eNA9^<PX64(3g!K9hxt*pFIR5+y_n+H!qz*N#R%`@
z5FTe~0IJU?vXU41+ZW$Q3glXW@z>OTjjT=bIXrq8+r56rp_W`b>x7zypyD7+663ft
zuHxMxed+d>8OjDu)r3P|O}5!|a-&RYr~|JUn+Fv|<pH0A3Jj2Q_<xi%YKs#P@LzET
zT{Y@^b9Ce@wGNj_i}{}?A6B~X8OP0WiT`9(oN#$3ifo1241N5(J?UIpHda0eiS@Y&
zT(nbDZ#1KLH&}dgft95-I!%`IPVd@S6R>#7{2u)~7a7SI93Sx3f)s3tG%R(f77!2K
zxZwabA?lId8-S3o+1^>JFlFl7S7TC)KmE8ST3J%;9Emei%aLBzX(gk<@$SI>fd&=c
z|7I!%jefId`NYruwqiyFSdF@RQoU|(y@Pc}RDCh+q>SjKs_u1WeyFkf*PFJ9{JTWW
zgF&)`&oOXLZP91TLeB5mO%LD(HHISQ{nTReF35wOSP0fHFdK`zN-CClF~s#T`X3OB
zD$I)XLO9yoPY+g>fV#hS(bX3>+lT2%)`!LLx$(>98_>C>m2T(B;PQ2T)Jo>Aeww)Q
zgonp>`wSAOm2%WTRp8>0#OV$0BC1lP;4<XTK}-`w#(`O8FkRl|+CHpg5Ox5nYY=Xs
zVr}&=*3_Z4XrXa_5^|q;L>g#rlsxoM;uXMkOYb#_+ReCK#%yC2TZPCntII1ZGyOF|
za%Zm#d^RwtP%_r6kvd40*x}(5)e7yzm6Bh|n;|?&ZzW|J=d*B5Q3L9e;r3C^<09)r
zqblYT!5<3^Ybc`Z8=qs^*!gs{_`IxqShC0c`o87IrSt-&|474hEjB}&2ct6?4~#;!
zkY662odoX9EbMfc(^|fW%8UMuNrUPC(Q8p)t-)Na<?a`1+9kyY*_Pp~`f6WrugDHx
z6Wvw(|5|`)ZMq){tmL;U<SABizDb;lL7s#3yHB0I5r9NU*xUO%F}inls!k<7OsH`l
ztOhqkzP0w*SCuFdZrN8Be~ZSQ?Ie?8(UasDViOZv_k>WpaKls~D`+VouNw7DcP>H2
zj2E!;gN_&0ZE7zWVLbtIPTR|K=l_^7R4jQ>uSh3|cn)eAjo_jVZq3JBDa4__RG2Hp
zQBO4GF6eC;zR9#BdbtK?sM^3nZeEz;#{@dk&9U$C@m4fHm^D;_a1k;w|J;~7$$rQN
zm^S;2jb$WeoNK1PH*v`%mgY*Dn2z`ntUC3kCw{}}KaBH^UPhK2TH>SJvqm-d6<3}R
zi?pueiX!96O;SHnjIgcC$3Mp@OWD|o#C`A|V3aJW60-1kJhG2yV?V#vc1O4J$}caH
zjrCFU#rE&L*{Ys}m!8i)nWygO*$Ub7hH%?aq=aH{;(yn{qPZJaUVrPMsBP1t3=uzi
za~A&KN8;uIT*T_4d~Gj!=Z3g);9EFwuhqMrw{BL*1$>1C^XIDvq=AqVCY2G_WedU+
ziaJmz!-fx%XITEhQ86CE1J_={#${jZO^X1RJG3m6%idzIlJ=p6KFJd$nQ-O`rtm8A
zSshkVo&v;Tfhso6eknXY(!f=lHWc~{doSc?fw4|7=n~RA*gV+mMO2s+L)Dj@5hhY;
zIHg21%Bsp|W`w^FFH|R<5FK}G+)OJ+UMQK!@0ASM)&*m$Wr&ZWWVJuB{#f+a(X{;D
zL+?pCB?v6)4Wr=ir1Io=fXVJ2DHuFkrQc1odav@I1MBnrVzI4#$^!LP%&LCvl?QtD
z8$&_`W)ZH_rAc$H9)X#n(z8_$HtoM8`+i`MK9bAvJSQdYA(nor%tvy6Q8p>a2G93&
zNbA+F8<P2JK}bXV&YW+0an0{*Fnz5W=kRaemi3A7P9gQ+;gujD;o6<}YV#^YN!-ZU
zIrf8K(DvDQbq<j05C}X#-vEwk*hHgDKrK1#A(HgQ*cjlkn7FtF@USmZ%p6MwBCAde
z-)l*>#h5RG5Jhk`!dw@#;lDBKg?$$Y^+okslr*9wIAe!WAsu#gAbG|4^ph=U`T=ue
zkY*l~iJcVgZL*&tO^U3rwp;#j&($zfi@}Q%cW2Wp&tE(_jm|s@ZTKV%p|`xixguG6
z=PzA_BZ_!6<F?Bu-^bBQf}1y6kPkiEmUDQY8z>)<izgG-8MZPv?37}bG@<p7p!cTT
z$Db*P^j*Eae`1)eZg^MWcVVoMaL-}?X6gVb^22?a-4zvPT+?15j7UC1tmu!lq))0{
zf~KeR>B!p)r|oMZ!f5(KJQE6Vq7$@fG|VHgw3#rxC4vdCD^jc3qH{NgO=@S?()#Ng
zSA~OcZsOCH`!zD=voeyh2whD+QJ{>Q-9RGKZ?O<#!_h7F1fIj(dq9M`&3HKd^Gjg|
z(3$`<w}PF?f5XhrGkwt92cJ^N=^L0RP`PYv3mcOa;V_TUhW=1tHq;~)J)grna8TVB
zt1{Q#gj10L)~;ZF5}{0~u<@0>mEJ*6m{*~zKF(kcM;224C@P9Tq|)9^jX*i+A<$hd
zOPn?|iUF;6Z{K0anL8=Zh^EV{hv9{qk1PSbeO+w3N${SMxRP_83e&627iW_agA=Mb
zYH$y>hcHR@o&O5&bUsMLOLKVlSa2KXEYDn1Bj^yesjNzQ&uwJiX;Rf6(N}E;D4mJ<
ziZi&iY>tLL4sx7s>l#Tk=8sCT4Hs`cmyEs<5?i5p>f#cLC^U)l#V6nm7V2&ui$<y!
z(N!|U#NB;BV4iRILM5D~g@b|YQA}LY%TdyhKQl6N;#fPo-lw>mN%gebi%k2GjBH8C
zm-<opVl>|RIYfF=$}Ak=+15rh9x)<s?>lgND{dCv3%cez%bIlDKIN3?Dev1~h%s|D
zElcS($66wDosW_sPMV1G7UgXkNE?7gUpSV1co{Pn-@yk$IAR*O@xcHNh8cIJtX*cR
zsmo75S&;5h(EotR0Lfj6p*tWuTyJ`2D|30fvNo6{yupS6O)O$oj??foR^^_Ya|()Q
zzlSifXfb+1g`DLNK1^YHQN4ZU;uZq_tbLoPo2vH|A~_?;O$k{f&Q%0dF0!R^#?v*i
zd1Q+>J@`?Y*RKFpdgkd((3zz_fjG#=qN2S}*qrBl$$%7STzA-&{J%0o)2G5au1d&6
zvxdDmaNtgYjTXk=;{8%+tFUmRzvtF>%ZO?V!Rql_+TO>%pT?Yx<NdB^*M2I-BrP$|
zD%aCLGBLru-I<=jOd9mKGUH~~wL99S?AgMd)&w@p!_1%0Gi*zFd)uPRsF(M8&k2$+
z-M#VBXI}iMyr0+fg~P{yUg^i~m`S#3l9?x&l*VaL3Pw^s@=7l;@ZkOcPhk*T8>kzW
zG}i1OBKIsMmXYHIg8L2r%{c+i*iqHaPZmkRR1^zi86gCADdLSi-(?uEkiNQXjm%L$
zv~m#a(IxZ7%*+ATj%)LpxQ+nDK^a@X_^Fn)D#UCesh*Y#Kr;xHq4L8_{8&V!NBC&6
zsgt9w3+5F7FfeIQf<<u50XhuziF@p}uNdeRfM@?e_ER84(f$(oJ^^75LkqMD7$M(W
zV<xJm{RV46U`z4t44s`ACM-;~+eP{LKW1JPEv7Hu%^Z0p6E?1>+M_}*JaYsUvvE%d
zl+L`PStHi>L+n-q(xk0~dm-FO<8v}|WS;wj;(w{~3JR!vIz6{t3E<rJ1=bif{xe?W
zQ75`}FvO8BOrP;8<=tYVp78R*ASxQE%lSg_@@66O?(hqZsQR~6YPjeP6WdRly!>51
zvV;@RU|^tK`q{+fHL@{ktnoypll|f&_hCINJj*>4VxqU9^WLnG(lo!5ocrl)nK>!z
z@{VGcXJA=wb%LYwcKV}!3ASIl;l{q`aqoDg85zD=PCh6t`uwkKxqY$pffdC)3>^0j
z`H18_opZ7?<Y6z7yJv9g9*^G}>gQT$yHXg2s(Ghc{boL(L&I0R+9WWt+2<s@8z7m|
zEp;j*5uOgiGK2^eU6)stLh1(mh5>W8_NXRg=B#GsH4OApOmOu?B}PVKiqs1j@xd7h
zP$mk21|<qAs}}?yXJ-O{RRbcm0GUm|z3l9Uc*f_sZ(5>KKH*fBDd&ywI3tXc-Q|L5
zr9qqJD@7;997zjzxpgja|5TzE8+1(f$mrUWiG^fo7GR&@uh<=s<=9QRe<77+B_n`X
zcCb%eJE&wZXEx@h4p}&>Whj(s0c96kW(@h%>)=825Hl2@6N)E;M0`q9HGc=Gzu+UO
z-AmYh_iwR%+1_UzuEN#Zru@J1c=@?4y6(VMO+D&(X~1(K^yMi}dl=!(Q#K6hSsi7X
zFvP5C@Zj*2iU)Zb;~)|Bl)Hkqy1@D3XEmq8r2F)|g=DMxHN_>>k6-dJVZ5Ir96JbS
z-O0kW7$z?z@cK}LAuofa)^}RvM>p53NgXq%{Yih`GEjg1;hRe`k?&{tXbmL8tS`si
zMBVSQjNV9Btk@#zUivP|_kFnh$9TOvov&N&mvuUPtfVo|6WTbAON^|8C*<4i0pIrO
zNoL4Cw$)zXywsz7DV2r%Sh5*j1K|>$SsiL?{vC{oatzXG2#{^-jw2{u85=;|>-D_c
zLrKFym<Rbf;`kV#V=1Vv&aPU6^cFar4TY@0^n(crsAr(QRxhqbpKd>{%Kf)?bttfM
zC|F}AIbbHKRxlhFWL056@fFvGm<R{!=Iimc=n!s9d?`$Nd_r_AnwV!F`Td?@&SL01
zC?zPRhoHjtu;ETKm!yh|#KJn!t&J3}6n0tK(jJ#YXVcV*qhi;U4f~O*>4QHbMMf>y
zhZ|!rKtLGRO&HP!p3Mi;A%oM?!RFsVx<}|Q4qkr{-Bj7($`5{cdN$;k$DAT#&Fq<#
zp6;i@ycG6Cg0`lXJ?)h7Mo%HDqog4>Ien;9J_HJ|S8Z^@3=ajRzrP=Dr2-ATyL*jg
z@lljO)7)ctZYCc-5VzS*;v*z}!>tKVv89#*ne6+Y#*dd;vseD51r$*`>VTm%y3+(e
zYKRQ7nk)~6^k7mN4ERd@{BEZGnW}RIo9oJ4;+~qp)BNGOi_@LCg|e@vv$*x534uYq
z701FbA`wLkS9*|s-)Lby7#JeP_|V!WyD^?Y!@c0+ld8gM`>uB|HLD}HFBR+m@%5Hb
zRd!w5up&rEgLHREmw+HhcZbs5-6>tt-Q7q@H-c=sL%O>`y1vDAKkwh~kHg_W_C8~+
zHRn+?B4acWE{O0eub(NN<$`?nNwkcV2dabX_vH*+N!9@GbXh@K%C+00ADEC(<Q*Nw
z_g^GK895F~rlzLTXD--f@=Ws6yyrqg{xXiI@vqB7!DptN;nyP_>c`XtB%^)W=JaRn
z?8qnTZFHvkod{8$EPN<83;7;220h^@-d31!*vs#pnkDRE5mv`4vx=ivRiKz6gYQyh
z_DJNYMR38vF7pebUT>L_roh3^e@v$YuqPmE$HT`jihkpyU~XZ7i;uqn>{FBgVtG%m
z`VxwF?N{8@FPQ@bRMwu|adQ0qFPW%7Xp;QVf$Ep>FqjP<GBP3y(ipy(TbP^YR#Zd+
zk$bG$#dbBgv9zy?C;hqFy7EfX6r9G~f|==!B%OHNgDE{6>RZXp<NdaVD*)UI&iR?2
z-yrYRG`n8XR>358DHj(Pk;EUOS}GE4uL{kY5fT+=(L*H(`9Oikn6!k*qyoi&6>hDX
zJlk3kl1Qed?a=CHqyd)38Q}A)KueKC(yPzX;IB0Ng%9P_0MGO0)K2g7cY|-rvm>k6
zQ7b|1?a~MSQRfiiV`@B(Rf@OPGP56RUASb}5V#{v(6Hb8OkPn_$UV$*BV*Oo!RNnW
zvc?x?qSP2;?wrw_gij#14DO@6z+tp5GRt=CEf6Aw7VXR~%-ODeBUKF(<^05TnIs%D
z0<X*{hK_A_HR|CjM|j7*Qq?Yzaz4J{1-*Nfk`-()&z*%!npt8XJ3{J;xW31YEO6f+
z=XgX+rJLR;Ay`&YqIbO5R9ja!J~1K0li{2|Qn2VMWgBJ-Ryr^pj5}qBgkv?W<w(Nd
z0rlYCM8ONAK$o!S>Q`7u1!ym@hTk0>9hs(z0?+y=|2;!QM1-=6iUJKL=!L-^1AYfk
z>Vjz#wjlMenNr%Os`&bEe9YoI4k57^7LGLeKNvKB#2(8-Ldir+Yxj;2*kv3f<xBK%
z$~Bt+7JKze(+P~=aSr>VPbui4m~b8|AOc4o1J|V+4b~|Ix}8chpwIY-Tq3UMC{#*<
z>#ac9d=PoB0Lcl_2h=1$QiF-y`v{wV8jJOA*fo$V-S&>Wc|gxWA?9a9_1;3dW3S49
zyGL=6r-KIs0A@Hf<C;<LMgGJW@e)FjuXIAbqE=q<L3dLW<KLsTBJ_3x$ll(*X+afz
zC96MCzdcG}(^@;)R%ph`&+0utkYo(E4M|cQ_)5()L>>wQC(7I0#CEFswmc_KGA(zL
z*<Om{H&6FPOg6PY^f&bUyYt=NaSiQ4v+>zathmeWu`qIz?11EZdE>!e$Ii(KV0KNv
zH6$#R&3=4sx3fRk!1neR3|2TGWR-~n(rDHc!{Jnpa;=sbj)bVlNJS;3FHTMzz=D;V
zlj92-576@gtPCdzOhC?WZcYU(7?1;rQ=kL8qF)yY_#CpN!KP)3PL-rq3MpEd<&QpL
z1a6NCrP9UTm2hE6(xpcw4q38Qt++~{biHmn@p>7L1R4-#AfPW%PK^~PfQy}P$9vAZ
zdXFquxF|{n=r!O*0c11yax|D^1jsGI>Lm(vBr3FauYlWf5e!j#37j6v{FG+Mx;GMy
z7~E~Gnxu!bbCS7%!9dLKsrf3IA43jKu!PTlO)qU`hPi7sRBcCyUyi0ct1*1IB)^f^
z%un%Sh+o1x$@qKZE_n2fKvZZ92~lH}oH_I(#4jH?NSUR!&F(8LA;xFrMhqWxq?}tK
znYquNU`z{T-DaSDupiM2(^C&{P~WQFP%_9LpGo~2sm|+5(A#aTx_q_e=OsBSp=5jt
za9RTca_~BXjf3m!=Y*+N()4ADOAw^N#=|qv(-W|;hm9NzqR<~6b_9w*d>PQMgE>kk
z&PN|~+S38(5GPGF*W^GAO4bT^FH=)ET1qgwl_Spv2oNnQ+kla$er#;)bG5qTpo#Ut
z9)+2rp<u#904mV)mX?)i<>o!<P!8e;333xS_fN#g!HGc$CzKY%)7X?Kx@9S|eP+hl
ztY5I+0A6F~EKnn98kMLsG_MtuI_^pS%<NQ3XeB~?tMwa10*3t2xBhPW?HfKwtCu7x
zw^{(?!Hab)^OJ|Bk-QZMVQVT?XU5^nq7)J^odTpF_76V}!f{~^;>St5eSEU`i@Xqw
zXkrMF*Z^CLmL%XIkhaBKbhjL+Dc*ouy6w%TQ?OQ(HoM*3FEfrDJ*+_!$8z7#-)DkU
zmm3ddMX#}CWhaEU=B+_`?XPVtD1BBkiQR|Y)zRW_Jg@|AqO*uBbIF_vh9VY&3bSTp
zaT^c;O#zWU?zI=A6gZ7_7dXMocX#KGhljWL=PT{U{gt?yq?T+HB&4SOc(c?B*4zPk
zU~dDCPY5MLTU#5j4d80c0wA*jo0b3>Ak6^z1t4d(86IFrz!I1!kOD*$a3s?5(^XPS
zB=of!ShRjrc5-s67|35o0f%^wylL3{)>5L&_Zr1khC%yZRmwYAWMB7L1|ZI0-gmY0
z(bxY)1b%>gjQqly;@F>g{ndk<dMb&|!#s665#u#j1Rp+p2<Tp;T`T)<_*-9>KdeO2
z<kZYeGhr-PAFh1-lwS_`T+r@_{(J3h5fZ=lbnE%*!TF7E!VkJ6;$}y4jn_`9B}BrZ
zKvDBLuY0s^pmumtY9WytvO4!Z1~ge7|CmLG*rDQ0;r}f1tO)&_K05V?+dX$SFb|8!
z5zj2tHQBDMC|R4`)$?wNoo{k<>64!Eka_w!PWRaDUev78Hx;LWh1r)>VZe<S0w71Y
z_~*}p($WDA1$rlOG`f>RzOAhN!$U(;(_(ZyRhq{ANHRR2#sL?BjvNedcW_2*k7kX6
zO9Kix@Px7zAu1XEV8G#}{&4!oxEFt60=*hY!vsHj?A<yJGdbu)$DmvyAQ=h{WI)`3
z6CFP|J3DfL7g;>vIzD)iS9xA4tW{!fzq#$j9@pN!e1xD(RBFijF)-duh3)cJl4r?E
zNng-sl2vC~I`rbHG9c3ecq%R)UT$^uxYv?jSN4Kai9TbifBxuTFpo~x`C2ywrc{?u
zqx_FVFd5gVC23NY2&oRg!Fp_n5NsOrFvf~=U`5>5w(<!#>XAcDh?wZ;!&i9066L~C
ze%FB=IzowDmOh_-0a5YS$eS}1o$*CpZ2_o1+Q<IDA_uy9v<B9jqUTsB!TnubrFruy
zg%<iMA6g16-@fUPak1ZUBBO<?&1(|hExY+0K0dxy9b9&4DZ)a%)=GTE0XQhZXo|%N
z18{bN*AHy0(h8$f%)7wbnXQj*V`BrcBLM50e}8}9)YSBqZOiEt%qdkPeH7S^DPqJM
zpXgMyl?R6ZF?={F7h!6-ils|{Iw6cddOSz30|X*0|LN<b&xwOZT%_88mCffA1RTm9
zJmcph%;jZe<Oz~8XA$7_WlVZWo!)IeE}RVR)@xuWsc70K<BRIf^nw}uZ!-vTI-Fk2
zk*`9fRuE*DK}CjT`N8)5>QG!lNs6G!hrIqnR9Pz=Y1N<+S!!1bvRFlF3I7E>C1scV
z_rl#F13DiFuko6xW%)e^F16G+-9!qJ2(SP9-fBxMLIzeJGuRSg%Yj=d+Hs@|EnXCZ
zlo2y6{GZBnAn|O=`L0A{u7-iDqko~KNYb%XIsr`R+_JAgk#K^@NK3y0l1kr!&K)qA
zj*ke!j0gy|#q^)jR2%4p_4S-wyUI>em;e4T^YOi;QdL*SfY4SW?OerdObrF$w{%AA
z1^dMl*<}8bDd(k1=J?ti{Zh$Ow;owsUID|HDYG@nye^DMTF$tjO7-MFN&P!*M^IU(
zax}^7B9D||ZEbA~>Lz71H95Bmv(%RWAaEpto#2yU7~jr?n+r}IS<=qjru99S_RFaT
zhc`3+Oir8Tlpp(=;J#--IY(ZUM+@4o7Z;7FE;0oWcni?JzE(<9Q3O(fbd{;2rq2Ir
z-;)ehi%+Yy<!^xW7`i_(MNLAE5T3nS6a$IyYW3<|;|zX`f^j&5xe*0~1XMr8>YnTN
zN4Vy~jG4Oesn8M%mjtJRAhZ@~yVTfRxQL1aMl;3mCRWEU1=3id!lss%6j?8v1sved
zR8&~Fx%<GZS<Ma=aBPDP3!ELXhN*>R#k5(oV117}Ey?I-$p*E4$y6$tisy!|6kPH6
z5mxSYQQ2w)+1rmL`@9S?Cx!dVOtoGLwp);gp7fUnB}I+!-|=8sdWpmS?rwHnT>?1u
zV@0>xQbFVn5K0|$s^H4w-hqy7n$HQb9ropg;~xq3qwG)G@w4Sl`6e{UUX#KgbQ*g+
zt|;lJmm%+#{M+^IT8WT~bH0C}0-tTjl)*xAXcoUhtgX<wb14OVUIdve+XZ_j@<rBM
z7Fjn>Z|>K8ue<lS!!biYOFnL&C4qUrcTXFRwRM?Sw`S%c)#sFMCOd8pnXTQpZ4Loo
z;|BD(OUtD)ybPf@$P(!Cg8|hNVmD}$btTOt3lf0P`sSw5!kri&-(k=bsE8QAaSw`N
zAo*{4I3FyErrQcUM;23MK{n@;1r_S_M-lXduspa3wF0G`qL6T|n%^FgpxcPbs4#TP
zYA~Os17H`BnSxHS;bEWBVcA)U>7$Z)I;tP|m-&T-l8OrcFoKlA{<ZF4ErIlXQ{@d8
z<yv88ob+`TC4_Z1GWGx6H*AVFs8F*aRB-}}bPU`~p|w{=Y`3ujaitvkT?D~W|ER0&
zHCzapOdS5UWM%X5K>NPxreB!T!4c+FA&#dOV8!zYMN&e~b@#N_{b!b*m3Jm>PmcGN
z42zpBmKyrX4t?(W@Ob^SYb`JSr;^qwh(pw5<Pu3x6in34U0d?ylCLc)f*eUJ_1bU2
zq;gnb1?e%fvaU0puG#)2w6$E|1d7`KoR}ov@x^o(>JRYIg{rKVcwAb*4^YBN!YjVc
zEB%ouO=V_oZa;ivuA@zyGAn$1$>;W-LZqe;B`7hIc4e9AlD9GyD?+ph$f#Ll;)Kc8
zBLAJGZnur9K(;u>7V2fJtd8OZXSAQEdwM{S-6y4*c1*MnFT?C88<dM9Rgku?-oG5T
zH{2fF_5E7tkuz03E4vc@lYyu|kdV@u2i6?L6XRnI2S~L3esSFt`B5P=6=}1s@{Wz&
zWH!##ocO~ik+CSV*8rCpDYV-!aZBqZvxxqWoE@<YAr+m+??`34#VPYY4-H=Y%G6Pu
z8fGJN`&|nki5V#|(hIRSS=1<@egCE8!+tJhoP>*>TELgv@+ng+4tua+&B%yfi&C(2
zagi4o0=J4aSC%5Ab(vhnIQ)<POUe9-;amL>Vbcqf1+wsd{`~p3ei($OCx7zb1P88)
zuTaidyev<`b-s8K>Te{POT6>87)0$$@dVg5w&y|&`f0u5X5_X_3-^!X>@qLIo?SL0
zk{;>Ljpgu1_6T)<uHz{e7l(PsM}hka&L1GRfqUzJoWvt{f@*qcGszB8LzPltMs7br
z)T&+@T|pZ4^W=t3eedM@D>_-q^?l2EF2EjAI#>j&%TEsHreM@I%qQ22T4}+N(}e_)
z?WTI~(?!SkC3jJVV>P!QvVqz}@tc<a5FVTiKqY>=hlnkBi|~6IW9gJwASQ-GI}yTX
zX4^%yd9;i)PHHO!2+_)YKTtbIW>h^zhh3-#gbK93V3Ov?$PKDEDO!+MAUMhHQZ7q_
zsSkYgw%muYcrug4B_%J=m50aP;b9LmqRSczNMzll7FYw?BN$%i!_`<R#vHIKMMfeg
zGHCn;qM0Ri_$_(>8d?`+C$$#W<XlG_WK<WA45n$Hc>UngoVw?>fmbMz)m9m(VMt9&
z+uPl>btqF-`tDwS-7<So7!pq=vwWW8?hnF{!G@rvldR^Xd_nTQOR=)dNCP;9$YT0#
zm8(4cwz`9(-YIs)P83!fajZ1Euw5eQ5x9B7c#%R2r9Iex5QvjFjA1Bbmka+Z(HR0e
zI$te8psK7Rc$V#}gj$4&OB(Pw&d)boN;tKY1ln}H4?8`bSrqC)7ns|R=X-YZgEc3i
zySq0C7I|&Ap`}V>K&}>JjsiklgJz+45FHC`TqF1cDz`64u~{ewHPI-h-*8be%SZz>
za4sR$c;1$&R3Nv73M>5OT&q?$1kqquSI*#yL0$dAYJsl$C4Q=j_7nhT+uO!mF7$X2
zvbdyq@ti7T%95#yldWv<)7N{35@@joxZ^5+cO5BFQ7}|XjzBnz!+Qpq$;kPJ*_G`}
ze_GbktAfm;z73!l1G}kGNDi4lore_d*8#O6?E_c7<Nx}?8}1zhEO7`G&J->#$5@Ze
zGdCni+8re^ltW|p{2$N2hXwtS^!i^|mtg97gp5<G@wSy_=~CZV7h<AS)?yqb%-s}|
zrojKx0{lfyd6R+I?%+36&)8dkLi(%iU5lpBy8~*KF2y>D#v1<*8B2^JBRk<td!C3|
zvqco~^pv8*t-PvDQYGX-q9RSTRKZ>ep^Fhmg@b;HN5Mf<q`_oH^cTu`wd(w<Wb+!x
zsQ8Fpyv`F7zrlrp^@@my2-@ZWAk}{93PDSw$tXsGN|H1PN<a`&m#j$kLzs-QNfE-w
z&c2zt_&!G=Up_`eWj~GovwO4O4o6WLyK`mgR>ZGlZGO#qOI*-SE5W4(A2rW9HB(C`
zaI`cxYsP+4QQ1O{RW0CGsZnOZ0;dW90cKkYf?VnCX#+scd}d|G7I}9KSiox=P)bz+
zsPTgurRJr(W0uI54xUVj@d0GxlMbF)jPUX3dvF6DlhAP48OCO^;V_|Ul%u7;-=Z=@
z<ztw6bKUy5IiDCK&k6*i@l=Q94-#PbOhYU$<Zy5QiJD&~81_2HY1G1^(P6azLKK9}
zqYWWKCbcptZP5JwDc;DMeJ^um<#E_&OmlX%qQB@TO>u0@K85lo8R$i~Xex7AGYm_9
zP>S>_>lSZt160>QBhi9Et^W`N0|C>Sxw$!jqQR&x7FJde4@ZI?S_Nd&$}V}RelLn5
zPR`!>Itz_9_d4EGu~{ekrh)VVO~#$rdeCzj<-Zhzz$^$*BX&He_dFtBK_*Fq(!{r*
zi_BjQt%ZDF_oZU+MKTN79KM5H0h9xv{sXc)wGyR*PbY_BGLwsppA-DD`VC!=n|46l
znq2~O0cxpfxFsiVLAX-F>-T4h2#M$Rm)F<#2#KJY@k}>>k&=`k!P#D2YZ^*M6C5x|
zs?3i!Z5DN~SonzKha3Qtm*H28K~D3fh~ONaIc8ANM}q#$fq76h23ILjgWZG$dyA<+
zlT*&V>w^%vinv}(BvW9BM6MZfoquJ)_NfI6f5%(?w&>~(2Igi9YIo0?jj^#RKfJHp
z;@8()_6i<GZ`bZ14T^EnE$O>Iu=ut<DQ<rLgHkM_jXU2)m&jEi_=kxG^kOgZW?S~w
zxy|nW#jm@;RReoHm^Q;J|8>i{uC5L(3Ni<Xmls1pzaeNlHCZR-=JxpcaS)r0m^tn!
z0o0Q(*Cr3fbAWq|UcT4PuqAVjj(>7$Sed%%HR33c%B8KxgR7F0siFehS+9`^d4z%V
z4oC}M5@Uf?5KI|cY>L)K$Ap7ktSJ-%ut%sKNN{j<{rjPzRQU-o=S`Yapy@GBGfTA&
zon@B+Gqy%uuyf@8JC3^O)P=;|@5!fTf^doO1h&^d1n54JAJ8U^YB}Xt&ht6>;Fhr|
zuq1?IL|vzpx=WD#D%(w7avq`f6({&=VN53Wg*$Q6RRncuj5a9iv#x89;h^MpvM~mF
zU5{x|^^OdMz!ep8hZ7U4joyPkHP@octBX$5vi-Cwt>5%fsfG^C%&lFiV7@2eTCSwo
zq*#M%e{jWtHFGf;JE(6|ez!gmaj0v=kR|*hLI?!&GQd>kAW$;`Fua;!P$qZ<+}R+#
z#;$y@s4LDD8L*Z%I1yh`eE<UnF1bfmQUYk>{-ck#-#XB-wY#&)U8|LF#fB-JCf*lz
z%we*q7@FSIPmQJ)qJ{qf^UOey7C;gK-~nhf8ch3+vaC10ug+W-RjL5_2_6!_=VWA%
zQVW4P92W;i7aW-}F=*h9f@9S>ATYJ7SUV2%*pY(&v+`t}WREAE&$@m*72}t<SljRm
zc<qVG{m*{9Dspck(a+T|et#csE);DaA01K-r8;XYP9-JoW_>)`l|7O>3)h^<oP!DQ
zj;45Y8PiRqIF&M0CC=fC8t(uoz4eXDO|mxVILKVqRJJjFK?Ms`%&A?&Rv&AQ;Rmbp
zN5kh;SkHgm@3KGjd9`G1X#KQGj^oS@J>xr_7#Wr?mFCZKfe)3GP%|7{Al+g8EAbo1
z)fP3Wxb@#+3jz7d(hA^ZKy(K$EOXj$vwu(^D(KY|w6tC_Ssb>9>|n7P8XDL*IYq(!
z1>8-AWTALWLp~NlAVAFkd>!8aM$A?N?<&AK<ajbN^~ti9Lw8h)r$)7(<Ndh)PTi|-
z8C?V=(J3`uSF;%p*-n4j(i+^kIR4QIZr|NJ7<6W4ZS7oHq?iYUrD<Pg=H?_oLK85E
zSHyUM0dJsA2}+|gl1cEIP8StH93}5^rw#JM(wbei9c@|m+ql7S@w{o2j~>KsJh^`c
z?hRs(k7nwS2c2QPYu}WI1FtBh*}txu;VPXpXvnbI=VU%)PuO7y<+^B+p>gQ+q;z*P
zwsw9TTv|`y6qbG8TyfDeLrtSSC4#!qn70K4ANZ4wRvk$nV^GLHI!jcaYlcn@1Pn}z
zLtgsxlc90D4Md7q>P**66Rsn2BHjr-8{SC#PWRIM3Hc3!DMY~pTQJDDFhIrK05RT1
zqM#h$m7HmmSm8}}ujr&F@kIdW2K4Z;cX$!a89~hzAZul=&7*-A=Poy#F&7}x;H3jw
zY(_=~@YrxIY2h}$hga;q1D!gh$YY90CMhc1*woZ!VIw{I$F=~j$KSNEh2qnWDgzdt
z@1UWSf&@LpZd_ZCtroV|!<n2_Xho_9wrY4KL2o;R^!5{&=Q=z*+<G@%X$-o!V^m9#
zufah<FZNi_8i1BSzZwUi<DiNJ%dp5%o{p^a$5D#(HchS@(J?nue>E-~^i8nO9euQs
z&-58MFT2D~z)a6^|M)>B>a)AQ2#JoC2;2IiH7(+5B2fotXhH&9Vw9ArRLmdJj?!wT
zel+-7XGQm!2{Ffv##)xF)ocwe()HiUS4zUf<xKGl2~{dTME;T^h34jQ-*_huNs+$g
z8+WRL4`w7yS`hW}N=8|7{9qdU{BRc6T6Ifurrx8~0%4P{w;>^L4RI9G=e4CPvSw#z
z2abe78aYV~1@D4A-|YQo<y!_+IC&aOO%>7GviOu%01Lm#SAmx)`LrC_AWYT;oHGEf
z1I$~rY>`%;hnE+)8=tHhlakm0Q6M7|vGA$bQyDPz28M<K1yV_|;uMR>f+FsbYWJD^
zf4_cG)gpF6T(0%l`jWQnyL%+9d=C(EM@L6yX699*j`Q!d`h~q%05(TM4hGfni?)vj
z6QEb0XPp4D%Ik7$2>d=^{y=GIsU~{Sc|RbSHUG#tS%FC{4RP`@2TAX`N~@O0(uYoW
zL5Zw}81^S=1GaPn?RA9yF>_4c*>rN)#Cv{536GM^(lR!=?OTfJ8my89NdF-ZYe_z8
zZHm(^)b>#7k*e(oY*VZvdT3AxGN(Sr)pty43eHHRm_$<6g=-AC5x#yFgaKoLAQrcZ
z=FM=OMaiA6vXU^LxG@T1nlu>c9j%^qslIJvCA=p_r|BY9es(}$g4<@wzK{Nvqnj~C
zoPeWq93Q!Kigvoc$Wv{3&N(c)ct<2pLG6!XCW+SUlC9!azx&mkiAJL>DX3cjE@kZK
zW<p-7Sq5qk5>(q@V&Dv~v%r42B`*|QeLYb@FRhbgjHpO7X4g5@U%4!YV(r?!r*cJb
zsSL}4W6SZbmnY#&?aAH2C@sEalCRVF8O7X|l&`Z`_rYxXm#J}iQTzk{IWLAj+=Fwr
zX;!9<vrBjEC|eo#YuI6?$xGh>0RC5zVv*26EWU+VH_oQ1CIZu<a9zTM!cR?j=wxA-
zkX>D2)!tF@zHgL<>UmygB15JbcD;5mOk1v23&yu1gLCs1ispEtJm&#)iEvnUWWlKV
zGZH$)GG!(QGuTxO{wAEm*~;!H)6;m5Uic{(OQGZ(y;P;}kZrY*rCZwR61&x@8R<NA
zhFCyOKsJu`D?L9?I#gl$DrOSaotAaVt3;kt1(nG*01S16zLp_8h0UbK<JuZv?m%pk
zhR!>-xHtqFe9(D-x&f?WU@T?)s8AXc#t4Mm0C}4I`!vKhI<~Md_;7oX!eNovq9qd9
zYH02rUe1HUM4QUL`Zb1LguSh2FpX{N1=p(i`t>W7K+1b9`~3JSTW;wl%|fs_fin^?
zi#u?f9311bvk(A-fgm^tNUU*cp#mdurl+T0K5Mo>T$eS40@OdlZ%X;fJneSipBRK*
zK)<nn2<*SR+MuEC;K?fx+z6{i?(E(pV7>N(P^ihL9H}ABC$mIE*1NHfnDUwP^xX!K
z6PB(|*Y03f*Zrgau__^1lqw=<P5wtEy_i^1GPHKG+~;Ate^yXzGECeo8#kObv7R+1
z0xirNa|{>v4vsC2G8L3}7J5Gb6R4&ZTxbS>s|L*-6<f?0`GB6Xs$y5zN^D8o`RRHv
zS9}3YO-t*@nMgEGq2V2f6(bMG2A2#h5CDxDH=W{W1YmH@+##ktc>oD$b3%ndaR{hy
zklpkJ%-+4+pMIe-49r@?qItZh$BP-`$%M373*a&u_)0?xx!zc~XBVF*mZ0%Z4hDc1
z5p=$m3(oI9)yY&|OF6wOHF(KbEh)-e2l`XT1p4V8^8ceQh2Vfp9Z(F`JGtlcB?m>4
z%?SdHem-3P0QeR4tE4^sb5+A3&P2bO_+Mam7q1mAusdgn<)<~vL^E)(4f*;Do|aD|
zDvkxI&`5tuv}r$eaOrR1%Dz*2EB|ygUOLO0eL`+QlY&Bv=;dG<9PS!N{ZZAxw=%T6
zs)qt~hUXHJGX?VAM8Ezi&6~hmeoMPrjI(V69S4vvf%JT*Zk{%8!b&|&G|YnC;&WXa
zo5~h-k3sJ+OmK#%Eh!395Jk15sF`<wvI`*4Nr=F+pU9+hd>4WlqoASIT*1c2Hv~Y|
z-NQp+$$rWt&EK(<Ql<d!dQ@Moo!JNsmE}659ZgY8h)2&Ck50053+D4^I3h$pis%NG
zCFI~2XpHx^mCDiRPOfOudPoS+1|aSPD3gfRz<mVix4`@pDqL_6>{MWHhkxr2T-a<t
zgOO+)*PqX8Ggf~@Tj`J^&z%K5_&*TGmqs1ZGOZ!KS!rx|(?ciDKJ#NE{SiH<7G|pl
zCq*$+XwI-axupD?V)U=Dp?3Y(^g4ADKSdua8xtcpn_{TpXFjf`r!T5kWA_=Mko4@i
z^hudvL=97+KGCR5+mK{Eeq`I1VK_jHT?F{rj~_3y{fvKslU|Kf(ZI6R*yfUqmNkcG
z<rN@p$`)zZ8t8O1UKGZLwzgx0!-^^@BK%oma9x0}18SdD;MpHa|2Qebbppcofe_*N
zj3&1b@dAkaK%a)isM!Yyz?VB@fcKrY;G>A4a;TlVr{A^Au1cG-$yPj933dJNZ=#ol
zP4q$=NTio^2Jf&KY^9vdznLSxz_b{9AbJl#zb0S~jJ0<nnNHYeI*jMOKDd{+aZawp
zb)Q3lFq|&l@s9b#;VEb21*v94ggIy|^I%`@?F(@@3?X=3ppQ%YD?L0c-AanFr6zNp
zUbLi1Ke>r3kvW144o+3X%*^a#*7&C)=MNlRVa<bfZ@aVm<FwsAC}*2qW-3q@u0u5w
z2}%Zk-gQ40al5Mw{LH2--FbRig-<W9r6mLUpA1>>Jo!cHpcd^_LB$GaPo?;ZJv(GD
zt}H_YCsY`OvDh0RA_PK#!UwEY&<kk#D1Z|lRQyj*-e75ib3KjqTCPM5v~-|@ZW0!T
z@sg!hTi&=zgl{Vlbczgn<(9zc#29AkUis%=D?@>C32RPsVOXtul-Y8to3&y)oe|w7
zGb&u;z3QxF1fk+ir8d6NF)^61#`O_i(`^@&b0x~=lF-6A^0pE=4X@Yu&=~&5!qF4`
z`K^mx^@h6?W`g0p_|#SkjsOr19`KA>FN%8H)7&mL-Am&|n2C2fTP9DA_FeH{F=_V#
zBjNKvn&anpt3U9PSq}2)633hn?CJjqLF_K7WZ!qKiu^+wE+}!vSIlHRn+hv9NB$D?
zIzi!m6l4jp442IqXxbc<`-;D&TP;xGMNG`j0?@@aG|s(TF%6U;V8ouYKfiw$2{0Uh
zE`O$6qZ@?Tlh@m<$ZBa3f^!ADwqx`21E6No_+^X)W-@&TTO{bL0KvV#Zw`n#pr)|S
zjQ23kv@hU6R@uHtolN+>>DAi6zDpS<>2#ht-<Ci0&HEoG4A?NhC-*1qcS;v1R^;g1
zRw+ODH%Ch9DZ>8)_EONNfmY_CC*EEiG%&Mt=%Fv_0J;PaDBX;(_0J1rv8Se{zC`c-
zkC)>gL3jRiyZzX?K`>at2<?$U0tqK!hkDzRDd5u}CG$q_U(G=WPt0?;02FK;2_FXc
z?VEzvy<Y#$fByPq8%hG(@}{T&#Z+h#snU0$h5<Au-?qbK@Yi=gguXsruRP!ZT7?Xb
z;@o#Y2V0gtH9M=sfTa~EG06_J(qJ`ax3y(}vv+=ep4)cyRf=@4O8?upZ{eV!080G>
zh$>%A?EQ<c5L2PLh~;HkOOefNl}u18h*fMJh#`~Ip0jN}-=Z`R3vu(8C@RxjHif%P
z2fB)j-Mpyg4d*HE!L6FKFNF^}=HKgPCkqam@eOi%4}kdqbgUd49M+;Qb0bVm`!bx0
z^*?||P#7=pfA>V2=ildb$yG1N<70njW!4Ef)^67r{qB3X9J{L#F#-S55<Lx7g;^({
zU;>eWc2eGehuJ;{yaZr$q>wRVUhte2M9dudUze`{0uHK)mmG`n@t3BC0QqGrMb=#g
ziYK57EU2s;J(5&jO9fr|<Nft7pd@@rSY~Y6L<dO9B6<oLMVKb-IoqeQZeCu6gdXMN
z`0Moeb?5Sre#@?~HmAjsNox6h_?V^5bZtFK3D1};mO$Co+&n+rq}Y1{UQOEe4+U7I
z2RxuR1q(l}ALz}~G>jMO7nHLH;sH<!`q-u>_0l}H1W7c6K%j-9O>gMXB^LAft4IvZ
zNBlNAZizVo_F?90zRtc|M1)C*=<fibF10BORhmQ^EUA2@P$h3I`zEMDBhv4#t?0+;
zgV@CtlD9t4mC`pCt&FxI;Ni+zDJm+!?`gjmPswXD5=zRDG1iw+3n)a9zDiwh6U>yf
z<568Lx<1p6YfE~Ne)g_Odz@2eZr%S|a%^*F(Dtk<xm2Ag;H7YRymIh{cloI;D$9C1
z>!D4XF^Nk~T6*yxn0cO5>GOE7bNA@{%9ZW(A1?oMq%F6j4{6KM`?yWU0%2&xDy#O3
zwyg8CqD&)0LxSk38CAQv)PlvcLVP%C40(6A0*xp+L$jmCASUlCl%}0W8;`C0lLMZt
z`-j=ve>=}HhhuD})Vs>1o;Qv+H(5S`S^TH;cW;A21HSNGjdh$JtnhVv2>jboPqBKE
zbVLUc6chEUJOf4!@Ku)#R?d{ROjU=P1P)3{lS$uE?kkrA-;txj)yh)K?CcOZ)jsVl
zHC$ic-Ofrf5d?h@Cb#A9dw+v4GrlPampS*3PfhFDWAhva>h+tYOK8dVvr&P~jFVfO
zYoDF))~yJx)^cTrJ#p4Kf?FZ99{Sw^?DyyDH1V=qC*eALt(hHn6I@N_=LrvUS?y7_
zG{P2A7;~CL3(gM$(8T=OCD4N%(EKAO6Pq4Yi{WU9RZnLT<t<l)`LTBGk8)XSK?h7l
zId!<~ID3SKj~HpU+YQhDu*A1WHf|e8kmcw?U#d>$XlmTfuPIFhYD6}3)alYp0mV3i
z&_%UJv=IF{3Umc$zI!%2Jo6k)e%=LvZ#}BBbPzhqi5e|dM>4V>Io*5P=w^5aS$`2b
zzdSoJ<+vi|mml6^W`<LB+C|r2)X<~dpRWA}-uqpC6J$x-L2B^%u1nH6*>UT#?JPYq
zl1f27twL+vzK5s%++4PMK+JQMi_&`fo6+6<twyse(Irxy65lsP$$;SRp`X8F>--f_
z8-11ZQ$vW4h&Sx7>nA-qn|UFsID^}X!2GgF@u-e*N(*_?*NQb(3$j7MqHfa#gK&X<
z42`ag-1Y<fQ!I0d!@WiI(Zf3$%s$9J*-S{NO(HU*9s9@^((JuQSe|s0;RfQl^g35h
zeHopP`9FJkAxR~ENEhOaiWWd32z_PW#Fv(}Ajp$pcoqu-q1I2ejY1Ri%Zck(wukys
zLPvv5Wm23TFa+Nf&U-aBHw}3RpHH5ifvTr=!uMsaHORWCwP5Kdii$o>%E=lHj!GhD
ze@4GKU9s6AR+J?2bp3>p!cggP^xm;{l&_)^NC8vR@C#92LR}8AJA7l*hjJ4dPMg1-
z<mg@TFlxEL@@g--<oi!EyUu<h77M_tdi>MIUf&uPIJnYz$<>x7yhQi3X`mwn{e$+t
z{zKaHHN+<{gu(Fgy^HNp#7SZ4Evt3&7Ew?T<a3(lP#(+geo6*)d-7}V`(WYH#9xNf
zKQ$e?wGJ12lE<szg-u^)Js)_Wtwl>k2L*|_L0X=m$*V)^D!uV5TegUdN4z-tQ@27^
z&v?J0xLl%DYL6t$w!cAxmm~TT-*SA5Lst8Jw)M$cxR&)qIp9m{<&n>KERDYKmMgUp
zCLG=F!ks49vxuQ|mk(Tvo+whYi=s8|Fe`V*QD|{)-a;6bE7h=QUppcEw?@m?IRez$
z4Po?^&sxim?Q$eP4Tgd-{4^^%Gfo~h20r?%X+P~$Bf)KuwLc9LOY<9Yf{(_vQu8DF
z$O)tm$IsVUliUXgSd2bzxR&qujYm5ouiR$O%EY)D2a-@pgx)nCWJZza#z-*fe4kyi
zM*;u)+WS-qPEKj$`95!yROlVIYw-fL8laHhWk#vdHJmaD77*0kHXU(CZl;zoinxmM
z5C+G{XascpOE@{LmU|-XxW4*^v+wQ~lZnK4KNaQbJcz9Q+>7EGj|?Xpn|NUD&h>n{
zCC9wn5i_1yq~2zargM|#v)z9=+8E9f)OBgon8x({=yJQj<QXv3Ir-z-X7eogIZ)?Q
zeT-D4Cw(9JR;@Vu(9@e>w^7*V^yizWB;13A&nsP*L+Nvyc*jsgHb$g6t14<85hsre
zZ#>U_`G^~@nu#`N<vx#9y6x~<ZwQm}v85jD?I!nUdIjSU<7D!^?RtEEa(G4bw1uII
zg(f6Fu=)345qaLYVT{w_tq|PR2~X#|#c_UFQPvLEV{&zS1fp}rPHxjL*I5sNu8xNb
zl*I!Qs?ebo@5>gxyMv;+hDTCS53bV<V$ZD*^But^<?Vrkl`gN9XB(o2Ij);N;`TK9
znF_;JAB+x{oev2P#9oOp$3=h74o5p_i7p6;AG44imOY;0ZMl1vk5E|VFEAIlp5xn}
zNGsaTwEj6>%FsESrdrINS6)9O77Gs=ibCdEDlxw6@IBxsQ;aA|X3pR6At7-LWeSWw
zZ(OW(Q>*hoTrbe$FMaZVHI&BdflzHrAhENi?R@fS*PD=U>Zp`HP%|^i^TDR^ar@wA
zyP-o!#wk^pZ=fb2eD0wKWzJ`ndF6SHDN@mY+2{IO(_>_sc#lE;qy*7zr47{Smd_0l
zG(E<;&*RNoqWjCq)}0Ej^Lyg<aCHVDJ}jf9grsxqtQ&zt0o!D77L}E3=31WbGPVDS
z;L;IeMJvi_PjzK*A98q#%HkVZ_R8?weRMew_OVB)Ru3>_mG#m>^0`LI5=xypTj9$g
zQPXY@TZ>`2I;unsJsPzJsLI*)g3i0nO5zoDg%;%N2?6Gw8p?NYMRXTD0@t(GAf7j?
z<H-z*DOHXEE2q9bmDuN*o&v*;|9SRWaY#;vAyl;caPf)6VY@a~ve(Sax;TEGZ-2f0
zO>mRx*#Wn0aaz&r^Y__B$6>?5OhMKR3e3`0<Tb~z(o%{N<@ID?#a$kqN60-*k_^$q
z9+&A!vIHNI(ORyXrT4kmtJX&?VqaSv3Qe~^03Kcs93J)k4j$vxk-(6S_xTq`qGP?r
z7UCQ;^_QKx;k?CV8EWF+^C_S!vH6JG{VoSY-6kyFr;7#vVY7|TteAT9wsT@dQ^d5|
zV%Jf}#eH%iOi-6g`?b<5{`<WZSj&(ev~da?J*=N!r5zkTt$@|<&*IcV?7DLDxrE&a
zdh#%3xV6ziAe)rE%O=wki<Z*o9M;93NKhCf!3s97^(Yem_1|Wz>-boG3n}t&?BOq1
zQq-4npUwxVU*GH!FClH6PeHI{&pj8`uD(?kX>!s-7@(snSD-sjd`^#Wg1X2%+|k*)
zaz?DFSU(=SP9o{fVIqiRt~AH#GIt`)>bxc*o*0VM7&=N8(OE(L^yyK_`L_341?(hG
z#cQ9z3to-7tKrReBK4LxEc9xOLReuYszH++PxoCAjcQBMizIEoTB;2*>2M9UH1&2%
zQaMZ4$lrZ9uLhGo2*Y)a5_=M`XJ~mKuX7_4H}1HA;_i&;`7R3;`JxO$mw!>wu;Q%z
z%lCz%1gM}-N85Tf6csG2Sm8qlP(t5dTkpjA+<8CkPZAq@@^$wJ&rs4X)Sr~l=!Dw{
zhor!<Se)NW2bdNWnS8IvY|mqKQSpk9vZ4QIj*1XWQzD`UhfK8pI{R^bmPGg<-fGme
z1MhqFSt#v-3?&9MRByioAYz;^p~kpd-ncarx<z(3Tg*b8u96=JV(^_!IVcyzX*y3k
z)efh2^~!wfis9pv*tBlD{aR(kFlcPa=dZWhW4<4G-tr-RMV2b9QdrV;Y*F(>yV5{!
zuAm`q5SzR9B55f7lL3MquNT~aQQVdkHk1;zAT-q5%erZCF`}&c?*FF+0J#zfov|Vp
zA;Lm@R|iFcvo~jQM0%H{fASAX)wz9;R~{Q+zr(HEjrc^1_n4-Adb0-+0{WNbRiNGF
z&N^*@`=k!aiC1#go|#;2w@$i)5`IRfR}CwQcC4_9p1UXlTp2~LPP^})5@&ED#0tVC
zXhZ+x%G0IWJxz6^EK0HB3T&U?;hPu1D?!<Xp_w!nK1f}eVW81cLUqY-Y4*zz-4I8g
z4sByMUi_?tm2ABrY&dVwc|=Y$IwKbF3P>1a5{>#3b4w5+5litpl1i{pYfn4fo9Iqr
z+hFm&u4DE1`Wd&<XXF0%R<2#PlUchabdfJ8RC}O5z(=q^3C%QpN$B7Q71;TpjpD#G
zmmSaTxM%DD-Qj&o&m;V6HWsi)@tq2^-tBXkHgDrEdflA9has+9y-t&>qN=`{FbeQ{
zP4qO4Vc5&c7+QRaVlokBNLa95{<Rve_{M)mvpIg69=qCjT7Rbe>(G8ec%$PbI+hk6
zwprG8KC&3A4J%!)CMRgEOIoz*jyTG&*er}MQZlL>lLab~zGFA=2?Rc53@E=Pg^nM5
z82M%mq=^?pli%jru2D}sAwPAuZkKp%jBQQhmpqWGZJzbOB0K0TjVJZ@eFYwLa^jz{
zwThB^CxT2LZ)%l20FGJCvt-@GFMG}6-#!!x3{?^>b$&a3cr##xA*(mhZO4c@Jx%L<
zv1!~Ix1S#$FB5m1`5IBHsHtSxi>dYU16PfW4|RvTC3&DwXjz3OQsE!4iUo7z*NIOa
z`4f$*-T#mw3RKMnqh#sco`RusXJJyVkOgaFNcg?$D69a?YfuxZ{pNB)LPLZa-#dS6
z^PDkjfWvcgKc=Hx^HZoRER0fkYsVH0IrkAo7@t<t3aha2Jgz+X*+*rrFe6N9PKr*M
z#gv|czTOx141aS|)D9J7N8wDO`<cT1-0Cl<1=BZ#POWRtbaC#U8Y67-L8{?j#$$iw
z2h)zrrKy{arz`y=COv5gHdT@sfu!KJE6V$I*8m4b*Q?P_Z4qyRn^WfIg$I8GP4ZGM
zY|HUqc^^NIVBhw?=~$nxYzZF{z8b2Cvbq3Ul*vhZH{Z>#U=o5-0t5Emje{(``*4q@
zqr`)8{c!|eHko60$j?`vyWLqi?G??rjFSCjm0s^(={(MNK-uo47CaHBgN+A>v+32b
z-x`(-aXsB$v|cq2-aeuD$nv$tcijA9y67W6xE;PTxw#T}=Cx2&@stY{9(wP|uC)WZ
za*rO|a#x(ult}FT+_B5E2&vGzznxd>_+qsrnYno5N$K{mS#BOCRQoVd(0Z+9vld#E
zY0zK#4Wv+AI$nd8@_>>5I>jfu-WgYpug+-TLQbr_>Fjd0%K8Yw`(aY7aqrRvq*Q*}
z+dtm%&3cRt-_}+k4WSmKOIQVU&Ro3#%B-drZJ)%OW}}Pw111wlxJSykrrZ~=VNA;X
z$*c!}m;XTVzF~W~#C{5`==ORz6d*QUy6ljuR$o8Rxy4MidrK!*-9u>89%JM6$mC|<
zjb+`qMHKdDq9jaLjL3xdA?vy=LYZ~s$mY?j{rZS_!l^N;@@cqj(e=PXf)c}XXL<B<
zj*xA^(bzmoc=Nou^WmZxIHMwv>f%y#^eWo6RlsEe_7${C>(5=c+zr=k9KY1z@+-(y
zXii}~Wjs-UIHhAaO);krlqnXTCo4p&yw{Hb9XE%wRn&11r0!Mlj;xjG+~IUMKq8=_
z5-L9LaW6U_Ts|C%`HZf3Z?&}Ej*wp%T9{9?-jzzxKQS)6mKmE;R|~~X?hOc_+BMAz
zj7MH=gj`(fJhNM<GHsXXbVR*$0oP|A*zg*UwhtOz^;#xuk_Bv#WsvGj=f}e$3N3~#
zt#S)_ybM<ohc}`;lSaMQ@I5!a34pf6CWTKe<DjBxNQPbjEjE-}t@2w$j>V$qB!?Kz
zkB!s2S*_v1qJ|pR@zw$FPh8Ks+&#xX1$I#Q&tdq_Lp#=@p8i$7ZZro?(Rm!P{`CBo
zHMJoPPQI5cy;GeE#C1G7r#)!9&|R2PnQ)K0^&+!DQ7L1)oA{Rg_;$7E=1xsOf`X`^
zsrqU6fQgc0+HBvvv=DVsDv@zD2ljYN$>;XEbpsHiJ}w9N6<hDJZmW0?mkF<|zqVc+
zHJoRmJj5>8G_~hKf8|UzT6Veq_P5~#_1$)Z()3iw&<=POQ4S8ozdi_)AvrMqE-ejo
z<gmAd_PS6qowLg?)Qa{u6#?)Aj2Ls#JvrlMTBMOi8Z0cVT4vEw1k}NIE0dcC-rs9u
zkAa7rG-;K^*BkxI{*%Oc%iTp6tr(wn9D9rU%t~kXGB-l6oC;^x3o<A7ET3JU`DdOY
zT^{BBHK&i<rtep`E<!f`|G|$A{aJ-ef2+I?ntpcON0?&4G3bVj{#91#%Zxuvy#0++
z#7P;sW1R$P)y;drOi{3_qanWCJ3cN{XSmBc9Kp!BX6j;L9H4EzE)>{SUa4G$WbA${
zx_A<Z>qz+JMo{GbVdS69o#BGOaz|8M$416=Xa!_1eYAuq+{fp_E)iQvxFm$l2Ca9X
z4w_Np?G~#n)th=PTn)<4yKiT9nSCqXglKO2k!$n@UU(Ou9yuM%*a5T7_{vI@S?Y^v
zbJZV3Kf|)GJnEN7)qk%)zh^7Ktioqts~2}W@POlLScUf6bZ?>Wk$BY{DV!@2iWsc-
zL^d?f110w)x7CC_#&aJ=r9R*YhVGWBRc@zRRFDFD`sfbu^#D=>80yMYf9eDJ;3@w0
z;>(LF30O}CBa}ZPQ%wH-H4k6eg8WnO?d_8`39L8;l#J>7Ed;_nabfu{z32({dR}bq
zKC>`&h^_1JstyipSvNh@09ppYSo*2khOFnCo>jKVz={c=K370OK6Dg1yz2-bdto88
zWwXKxgw644y!h>xEJB7Yy9c&tkda8N@O`0WDi`Pf7$#ehgj9W~8X9<YX$xIcf%dZl
z!i3hoD3M^KF*P;yBEz<B-lGCKP#^>VSHdpyW!5RZe@?%-LP<%fyIkX!dbK_ba25V$
ze%KyxxNO+}KS}2G0m;Ni`-S%;Jtak~5}&z|SusTm%j_ze(0&&BIvQZycD<IoWiWN$
zGG>GpnnX?2zSWrRs_ud#=kUdZV!)W2@>}e9htTvc$gnLpIlyWGIPI((*WRjDY7dvH
zl>G%V$v1Bj7;vWN%iF<ZxkRHJGd#d0GN@O52XSp9lsfz)@Gt%za1(aI)IVa8p_0zG
zqvx1$7x1ZC|4<y24M5|g9gkHc10OXJ)MklP(~kQAv(WvDfXcmvd6?DS9$yG<EJBjG
z#`=3j9`5UWFGHw6TY`M!&olwj`D5pE?|OrcTFI8>KLb%QFk~)zUJL`LEXZyFoP9ns
z?!eB6ccolGdN}dauuE^DUIQi?n7$mHo$q`~b|Q6e|4;wtUnd_$5i;D7-gj64zS=~6
zOeD+%=|ugHhl(O*h<j8t`N;)!KWYy1NR3o=QiUKJlrVmpL|rXwl^4FVF;t1g_aIaf
zN5UL<VP<7g)x^*-_Rp2^C=s6nb#X<6nSWq}=jnVFy<VXMO6PZjHPMRtIB%OP^2~~`
z8EqCg>D4L>r%P2{48BuS0cNSdz6-+pA~&)RK;|Ft+79{z>4BM+YqmIAvq1R`qHV$B
z1_759h^m)%jBx)c7;m^2Ht7hc>$aVL2qVeT2G>gGYia0Y*rVwMzz=Kq@|`kVQS_$@
zRWM+fTxq6nB?(%+St~X7+cW?1`FrSy|KC4+_B9XGey;-xb6}v}M<rM*hp3P+Yh=cY
zLrPF-Nm08<6ld)wI7pb%NB{QPY(#)#3%CVdtl^G(ih#39<FYXZyp^@*?`oWS7yJPM
zCkUa-7(nGoP_y&${M$T0Q`^HXlT;4lFUT-R2DC<_NP$Ahg`rjp;<O-i9nj!_I?;>*
zyR1-^=YOsy{*m?6Far$pCLFXdGAoQdZgI$OyTdREt@%ECG&2nxC#$~(Ug8}3=86jT
zSU-ETy<omFOIOurimPxyNqkHq=N*e;9)w%^4E-hKGb$RK@`GxEe`htF|44Plfa4nn
z`_m>P)#JNV&W{UI)apBj5vYi7-Ht4D1}6#<*786sD9F_Z(uAhn+;E^Ce%Uh!0FRQq
zAm3whG+zh&yoq2?BB1TwVKW+nD^KXZ0E~!n`2i1*P{Bc?3V~l&lut;nLXQdu{<@^N
zw!2$=(#>;!YN7+81-c<~?bSj0&?ZB$3qlBlv>9Q3hH2c9{<Q<*3%{$w5@b8%F*Sx%
z`%?6DE8Eelb<w&Uk=FL{{HnQM6;vyn`#-M!6_@#f5{Q~(YxD?Rbu{1tc@4>7b(8;Y
zL`sNHC6UA~6OHa=iPIDK3o)I`uRyurCFT>j@td}Q2EWJ?Rx=kkwIcDZ0jb}e;j%2j
zlL4qVMI9YEZUW=yr$-PU(zFjuzt(dKvzN5bjKE_rnrO9P-R!}PLxF<`R3U(zIRoUf
zJN->k=gSs1f*gRl+utm?vur3v9O(!HUz851md@8Gb*&kI5pI2pfDxjLbFWev6nP81
zj~6mEg(g#o3MGo-qKHETJt@qF7+^-`C)l&eC(ZC~fkDkUg(smxL_wo?i8Uu0dBT})
z1x?em+kT|XPX*AdQc~{<&2kTT>aAKf5wbNITMF91u<Dn=BWh-JNrNxjPnu18N^>(0
zh>bu(LIPXp7?8pmYtn+f3Q?9+HFBfTP;%A@U=)!Xx*Cc1&ygDh97SW0xyS4+9jO28
zCoN+bb>Q`xd=qD395|*tv<X#|++8h`uhCmQ9QJ#P2iD`Lj8rRvNGdy+98LzpA|pA6
zxHSZwyd{BNlrzh*j`@x|Q0UKsQOf&iN!2O(<R6B~1i|Db@kAmd-x)Sh1Nr9H9<irZ
zcP>PM#v5>Qhzs>m;<$08iAEYg_y)dNpwR0{c6gwnYn4~i@!~j``I%Qz5;`-Z3N|so
zZT?U50A@{8#C#%($~$Oo@gfb7%EgSCU>{y;xbyjQ1A#!iF82xAkWT;4QPYy8nP(P&
z#vLmIlR&_*r%Z;yj-*jzTh5wnR8A>XZBk335C6|LUY9J82+_|G=j5b-k3UR_lZA}s
zpXrbDzcg37%#lW<RPTIaB@!e}#buV3OI&{vz>;V793S_XrpUM%!abhBqW$p`pxa45
z>|+L}W5tdH8UZPEq7hKBYQ?J1C5*90=rV)SqpY$$6-d`!UOB)Sg3wJ}=01=c35!es
z^!7~wbScu!dU=Ir=g)H<7c?!UtS8UEu5nK^9gIjYB>Kcd;Duv#QNc`x9^VrDuh2;(
z@4ktc=4f0=9hBXyPwmssFWq_AeKs<!O~Evbrul-?Elu)<6kbJQ*o9}~>Cx_(Uzzf?
zpCsHizr|95@{-cdvd1yMhtKu(E&mEN929WA(isyI-kCh6KegC5Ekdi?5O$i2vZapy
zFn+*A6XCf`D%VM*wAJ*tw4g6UBa48`{T@~X*B1R3C0ZdjTv0D0IYhV1bKP!9oZ8=!
z?XT65csB(ydFwoh_X?*nroK8>kQGLJaHz$5X6Yf(={EK_^DbtD&;g_TXgIF88(j$r
zf}mn<QRx)H(^KMV)t>=hjVkpko#{Jc6@l<rNhV&Au1+54O&}~Fgogy~ln-U}!qmoQ
zCjdhhgpj=?m;v)?)+h#q1=tHfk+pX3J>Y0I7zm`Ce4|nyNE}lY-q2a`lq&gw=TbH;
zFL?$g>=b^jGH#_DH>`TON_GP^YUzTITxiYVlnq6i<U_8(oS)M@%#?~W`Iqu!OxS)j
z%EFS_`XWtKZl2xTnak1CoCJPvjKAbW-zd#-s+PbE4zFS%AhzB>#psBM`?3?PbN?pB
z?d{t(O<@Vo@%0|ke^+?mr>KsU5~Ws&;d+^`CA*68hXsdfe{LG*dKr1z-Vg4Y)wb^2
ztHig8;K+`bEQn7C+eqe)g!dzPpYe{gc-lNbuMZ>@eFWnQKz;?VYM7W%15G@LC;zGX
z5_#mB8Em9c+t8p?q7254;gid9;Zf$}X`0y<Y0wYnrxZ$HiYz)#%JBUg`?BX@*~^Qf
zdue@?r&%%aaCfx{^X4jp)2yLQTkeFjbo%SX?~{HuKTPQ~PTKb-TqiNy5*<e0%xu`y
ztA&}pf+DQm8TvDeE6aU#Li07R)pU+U=-^hw{1paoW%1Ecwr55!Tr7_vb4e&`H+#Qc
zzZqA0p{sHlu8h2xhKko3O0Q0cGl~3%KG`J2&~}WSbO1^5eYSonT@Z~VJ>13`X%9UU
zxwdwc>%KzI*sEmY{lGe|UL8W+2-$|c{Z=5nV|Cb;CV;#qfrA3`kF+b57?U8o*&yZ`
z+yjGwc-rJb3C1Ka#&oUOy0m0}Z%@KEEj||HU#`0m38IcIFUtadM%r-UYN;k88!Ib_
zPcfs+!Rmn(?0kuGeSR2y9$Q(-YKIuNzWIW1+_*y<$+O*4YnzMfp$g#B43{Mt7d8AQ
zP3wl%ZLhMKYR^9kGRE#GL>OET*FG2hkA=PsE-MX^v>`%6^>+UsO<x&R<@&shl!zdW
zlr%_7g9_5!AtIeBB_Q2MmvpDJbVy65v~)=bA_5}aAn)w+d;cF?>nskSd++<1d*+%e
z8hDn6pAj&~D0`61sLL|r<1!%8L^p@&#>+dZMY6}JQ=501O>N_K(fL_Q;c9Iu&N_>5
z#Rjguv$ns#AL7FMJ_TbTf5)VOW3G<h&zjgg*_Ezq<#vES%BLV#;TVISP$UCpO~Xg4
zp~0M)9+<`E0lN&-SOtwU1Vq0KO+~xyA)p5ujaS2|QXU>pLG$nOSI>Gud5$F`HhfL+
zINj31!r0D^Q)vu@-dtW6N0VRRE{Pd_1(YD;!9;er6QM7AwKjO@l_j1n`ug<-iHq}f
zQ*PjHuG34fX)et2PWAZP7+h;LWAbu3%{j)lH?z-(nf^h>+BmW7@a!S6hCrOcIK~%?
zhC3GHOJj?u+8$f9W~s@EgG=R<s`4~>-5FH@LlIKR1uH9|E^Eo5bWO&X0jjxQKcd>)
zVdhe*R1DoFo2`3qLE5c0v=ZI7q8bt%Cf?sYOx9dkAU~==_tU=+V_htC(4EBO=K=M{
zC%lrsFT10CS(wQpbR;YZ0y(*{$Y*G#g`@T-b8#=2QXCbtt0MX1Fnn&ckN<(1Vb&8y
z1if*Ag;1G2*|3Qma}4$kR$1V9&H-34Xo?U8(I9+F^78}5vwgv%;`7|3mNA$tL%o3~
z4*a<SjAbeMm{-O(rYvM^MprJM$unz}>GXh;c;to~Bp9=!c?`e#Um0z-S;71|<SwAK
zVW!HZ@Px9^3?~;id2>m`-r$9=1^srh`JEgRMUCkGWV6r2J?cRRvKg6~WK))sRYvM%
z(mx&0*LvqlUurN!)^qDG4Ilp9|N2#?2h*7B%s8IgB$<&#i%Aptqk&Gi>ra)CVRB<T
za`B>W!AZ_E->nL?%6N4+2KPj+>_DBFc%igeFmQ3pbYbyL!Js{b^@6l~=<X#<qxl9e
zz&;-+M_}lf${PxH{oF6bF@U&5z^Dj*X?gLRRhb+4r9{#0*3bXTmuNg)@@U*`y;NUJ
zjAXQ$f-0IzI_$b|e{H1ytEW{L=cMGP5J5)14Qn%WF04>%P8~99=KN2Zrr#B-|E(W#
zNJ#K`iu6V!PvB2lLSYYBlMj<3SdL1$U;3-4HD5atyAF-p_%r8FKv9^9=8BjbN#nu2
z?$0XJIh;l4gw=8y*R{8_a8ZXco+dFUB*Jt33|-b{xzz2*q3O92M-m<I3#C5-ncjM_
zDFMvl{!3Yfy#urr2!~NfNdgiTCjD^sgFZycJ5*7UAPcmdc<$qdakg29QQ<z*C}D(Z
z3u4HzYi#-2j_ch%PdB6f_knhNs#P~3oDcDT@~Rvac|Kh%mQa8ssJ)IMh?z5hS_`vF
zwtMY86#;6PJG+VK+0R3M`_jGZWKQ*(x;e)h-Lux!F&UZPe=3<0@O{^R>a(>erf<ql
zZwtx~iFb9!$1h=jy?x_5TSKW<!sxBzVa&Jk_LjCi1rL-xgt>N2Vb=6xi&s`2(*xC|
zgnAbDooG-YUmH%Pp`l^w>I8Pu=Q%d`0<gjm^q(hBo+R^H;MeM~qrEzI7e;#lwi>8%
zqxoQfh0xq0A~8ET73X;Mg=hcw3BKv2BvQ{&yh|6S!bTX-$&K&!!5ov~g^4b{1wFE}
z1ZKP+Q&^GIorbUk6IVea*ROPG%y9D>2<PPYWNX*FVZQC?q+>m=*H=ydpbUM~;}Z7?
z`Z#ZF&~TNX4p9&~?&)Ia{`a_wuqkq@Hw1buQb>b!Hm`<~#q!m`+5JP6$Tu=pX7h{B
zbz<LEPuyFcNjoqE!0Chj5QIozs|HMpx}#*v+A#I7_wV1w&y7FfusNm<_-Ih^89Y!R
zexe@&e#GG(GyFoTgkD05_`FW+?%lO$lzk|}##XQ{pi#pr_%cbcmpYEC(i&wU@sps1
z*xDstW}zyVYo>`AJ$+If7WVn9G7-{D0tsVsV&0_5*!0^fCtXoDCO^E}20Eidp)U=z
zPu@!MN6jA5mU?3}{<4r<$z?0K^U3*k`#8s@8X;8CK2GrqXtVR!-`2Fs`*ZAkGpLw>
zdm9cKW1O7+G<MfV$X9W18R^0i8LF|EK;+v`xThA4=Q!E4RoX1@|7}>iN9pmWN>h7T
z5rx0w`G4J5b6kP6#bXTS&<L!-CMxOVd*!VC8|)M^CdueQG$A?3)FB!Vv|1?fg7AXS
zj@=b&$+AtPN(qfo6PZbVpYC0U=uNjfp)L%lF22DJ39;NA3@MAlOXH{262WK<R>d?Y
zQL?N3f<!FJpQfbYmik~E*`=&x97$n;t=nFx_i(^TA980A+ET18O_}_T=kQFaGF4hk
zFoKLCMUk-#NmHICKToR!;YNc9M9K8{``B~2k2fxAoEKwp?UojS86T90H@&}c@i%|r
zimO*zF<Hb7OE;9iUn9XtLCR+;5PKrZj>Q}mkeo=bg-pbZt6C(fX+=|zPa?qKZy|5_
zJj;?DuYKbHUAp%+x%ea6(nzE4hq~R<>_$38TyGp07Kf2oF_7Oi>J|I@Nt&)VaWv|2
z)ara&VGX0*I=&!@UU#DeCN$(9S8D*^5Z}qG1a=012f`i6ch;%`*W{zND{kn&fMiJ>
zu+mFz`!n#qfh;+%mEHfZZ?ip%$`VW28(WK2kEhbiGHgOz(x~P$?^Pmoi6#?AmVFh(
zBb{j<PR6)XV1diZX=dU~6%-nxFHiF4?j3*0Z;^gaxDs@}e)VS{&U*7B_?CZA=OIxD
z29bXRJ&Wbp^QqzxOKkHwCDmhRWHuYdCpuOzwyN`}Vmo4WwVz%e1x+dMG`;?au9Moq
z0v-Z`<~d9`I%OsZ_$2S6HCQNjal+@b3iNBZs04`y1=b!A#I_`NsLLZ*^0=sQ+pRXe
zb&zZx<^tmP$_h0Yr9EvJ*p@dNtWM@M7B9Y>R4dX%IIr8-*brg{3ZT8Ip*dK<=73B~
zE5l2&^6hs(_uM6>)1pC*dUnAi@ZApOVTL0<paRy&i-Nt@q`}5yvxzIdk*}G5G9<Fw
z>k|yR6Gii*tJ5ic+Cz~d>~-~yP?wuYWl}FSv{)0_s60qW_KG@w>60bp?B1Rj{D79O
zv(`j9z7rR}D7Evr{@vKZdcQHRf_W-qolT*I5O!0|Ke-$7AR4%en4i(08k2B?n@X8h
z7bn)M(M(XIK^_H!N2n3zKM$+3A&|mD{43#}vK{c`TeEd3y=M0-0Ri&8r++vR@_K~J
z<gYigD61z6wy*nR89Xl+mX|sVent6Q(S;MUn}A5`>5u@7Wgl!k{lOLX07M*r4m|1e
z5eQbg+piQM=*B&|`5Ly0OU}lkylZqfCJ{b>tmO@UQzb71ZX5emM)xQ;Rr^ysK35sC
zUW`u-A`5DFVyEcquEww9iqynxu$&5^R=tB(5bHhYZ>3cCEHzsvboNMbi0WTO&%6hA
z=DBUs4{S!=7wdccW?a~z?AX@Ew0NHi+}C!{wBe+lwMMN87X2OI9j@tHKIz;3=PQPr
zm&&>rT&YtAEw8#f1YSn2slptkrHJt!zV%}Z<TW9c_3O3JN&NYMmB`WbjXedtX=e;P
zc9J8}hh05PE{-<AulO%S({TY%<hNkkCcI<pe!A#VNG_eRdU0IgqeK!!@$M_lmmx_x
zW&{lxF|m{LThH<k`}X#_)+?hv<;q@46EH+lwtDIiKFV2S9uo0j<q3LYvyzNVraD)+
z4M^F7vWJC<YIRU)xztZvFh;`(jDBf%w~A6n%XSuv-q{Pn>NF-;;I5=2CqQ|j^An5w
zMY599A6?up>9i3!;J!Gb`L(=OqiJ{f_Z?Ys?=AEZI1_Qv%J-+YodR%C#oC*kw#SRx
znZlyg3r2Q=jPrKhf%RY7^6XRcq1}Y~%J^8srw;}T!E2FEBO&4~7II=hb>l3kRc~`3
zWUIJlum77SU^cr}gQ5>rjPq5;_@CXA@Z{E;)Nt~lED@@~>e)(T6d=5Bp5me20;D|$
zhEcDRx!|mSM}-vj-FTbF6*JpH``rW-&a=jwwM#GKTo;wZKV!QeV&qAQJ%QOYY9WQD
z`sAE^=kX^)hs&x}H-mwiouTEId$0bF3vjbrd*S$2q+B|xacs@<#ZuNYt=CVb4dY%}
z`~3J^^Qzb}$uc=He2v7+fpM<Xe|tVT@x8cO5#xu-eB6NSPHuunY&6tZ9fu|d4031h
zC>MA+14Ki*k6VZ@0$6&vury0C^5sa}m<S6v-z2;YGL3q*xU$f0#GDU;5XBT$Y<KGg
z;Q2-rQBU3t`zxYdrlnDqu*9&1+2!5>VOU3}r;sQtQ>LX*s|yW>HGf)b?+IDxt7g-_
zc>}cSy!h`f%U-31ZJFSonR540Gp_F=pYuTrGCi4m^&j6IM%UqwIW%98KmfjaT2Qu|
zXh<n2epU~1m~KrL$ug&KnWEnu8{X7*<-DwzIqwx{?f9~M{nmN`LZT8m^asl7m*&kz
z$l$hhJN>de3_`T8-@d6nnZd@vIfUPBf6)IB!GZ<^KPcWFuw-)EW<1|G<tNNLQ6Rze
zPC&sC;+%?V-4NdWk*NEY6ouFOFl~7+p~4f*7u=oG2~>n)2$Tiit9Wf`SY+eSGuPmF
zu0Vpdl-|rb5W|%xEG|fx88$LpD6?*^Hb^Ncf2BK&WW&4w@(+qWnn+Xc;}S#r_FK2<
zhhcAzC?Ah*4jwThkq<L{fC$dUQ7?oj2joQ%jLHHwaZk}!aYU{-cYSUy!Mzwf@stIB
zpk{g=^jAH}?sbE&E`K)aKnCNVe^=*-QGR&8UbvN|rV3{?ZusS_7OnN@D{HWfgsDAd
zf9=Wdzi<2O*>gC1hqXTmj_ESzAMb~nORPxPzj|;xEj^=yi%rqOh@7#GcW)I}g5;+n
zxdw@(Xr2l8c#c$q?Y2wq%j+E~5484i8+(dAxm)VbvI@1?#^>kd!M3l!B69A*wtnqN
z#W)%3&SbgE%Xvp$48Ix`W9S^->t?GOU7hU@Ki#+W@0j04nnvVETQB?yp^N+o{etKD
zyVjfQU4SWi*^F9u9S(;D4j~9R0O<Cgy|3LzL~cBvops>k#G8PUh)u5{`jOYBLb_e&
z9nWpuWs~!Rb<b6Ttd(v86~)@fM00z1``>|bsq)PjZF$2gCL$1P#f|b|x*alv;TTV_
zmUC5PIL})k2|?s3^8!Ne1U`p?J2ex6E50az>iz&xuUy-x;lG0<Bt%JL%0C-%Y3+gp
z=l+h{R6_`JR)j7Jj9MT62p~B7CD4}QS7=ftD<g9o2~@(0DcpgOtjcZN8Q{2fo^^Ac
z#immef=IeW$RZJ^=y-d^VdeepcU{;55u}gmm)uzemR#wADc)fqW|6DI(Tz{Jk;p4|
z4bPAZGn~o5oY50-`8D0;3*ulZZ^0QuA7L0{Ov@!;=!91<bnul<&-F(jh0oRg)$0AW
z6-cjCdC7*IjBpKxheBRO1t-bCo!Jonr`yG5;mevgcrHtB4-?JvP4^I?e|;1VSmGG1
z`yXugVxl|)KwOfGa5NpB<CMw<M+(+*^0lVWO5ruKn%es$0_7PIX*5a0H2L&vsoHgN
z6s(F0CaK8G1knr(vMDvgp%K2Ek@)p!sOt2a;tFKios!7nV!f3Ua&x$T&OG5XOBaq+
zVWz8sw3sM4oYYvJSM8=$bgxqJxaZXWm?|*N#VTv4hZ7l;Vl2P+wMx<m2o$GT%=pg&
zuFIa{;M^0lN&4xR?K_I$^VnVOheGg`0bg_O6v9%{ATeB2Ni~SrlFL{=&iT>AvDW(m
z0fQ+$n11TWV*F^*&;};I%fDa1h4dcGP1Hz{PyY)3JL?m92Rm5%t)jxn78)gp_UUZ<
zx83#*{O${w8M`>On_amOQzKYnLP}`C@WjS&S|1FILkR!h{)btLxTvMtRkXFTV?}8S
z9E|~|;5z=Fe-C_+ua~_}7fWuN;i7u}q@((=Ms*`Nqwn&-=jPNW3fIqLx8=SsK)?lk
z&(2o-6|<f_7(LW(e8+J4=QGZO2M;9Gy3y|IwXY``X3EnA;Ck+d8~%hM^9#;}_<(Ae
zc9kC(csg9})H{HJ_we6oo6nF3qF5ph&|VkL2E00IjM(^ZQup<{VUcqVo<jXHA{S%4
z!><<jM@4RgvabKUZiG}r4YwXR%#G`by2;$;m~EGfyBL1dN;{iw&PV7g@G$&nJ{(!|
zSiS_b4AyXnpE`#_-_J8D)CIoR9IJxA_WP(dwheEd$UJZ&tY382O^s<dTMd5+?;LEw
z^itm=Bthq=J~u9}M$-GCiAW<Bhn<U?Vx~~zV{-nlo+yDkFZ$7mW2$42Df;rFUHZ+!
zl3zU)cO|@w$Ikxz;2bMW;FvDIoi{G=aUy#RH@b-sPQd!wovg5%Vmo1VE+#vyxDXP&
zyIc06BX%TOPR)X%)W4%&y60eGBq-1Nl+#U<@N1BT^xkpEAgJVVlV8;6m$j>Qqcx7+
zmnLD0crCv>F*JLGfmgkM!AZhA?Z!1aLmNcSSe%<O?!xPD{*o9@l7zh6vn-XC3b3Xi
zc&Kdly}{nm#RUXM8oly`A~iFi#w10@e#G?~?;4Lh!K!HZ)q6<|Ko_8EW)=<42?3O{
zps!i~?yujDh2Fd^dP|eq3s$1w<Angt5Cs98S_;UL;3~S~3zw38VQM^_Qsu|%@tR*q
zr0{$$CbG1@IJ1LYC1cz}6=eZN_gmn;k@mKU*@hv%3<ax;I*b<7O^EU}siA@hT!ddL
zkW}C6lnVaTZU2rdCU*bKTUD3?!o`%YU2S$p90L(k8LuvMb9Fem?Zl{<Eb8ImF`&5l
zAs^zIu&@x)<j{wi814Eli?#l{ap<iVuvJDB5~Q}Xo=4+uFMRrD^c=9E-GP)4pINaN
z$8b+FOo>{EmK)~kaTH3)>Xg)9^MT{guGg*yvo7RWCh5JCyI69fnE=;h9ym7wZRbkk
zbhesVb9Xr!d?7QM>kNt)5BU7Sp#^oG?8XQPpWq12rDNa9M<u1BeJxrvH5s5a{^>DB
z(UkOFvq8{ci?&u!T^(`t<0Nq|e{(r34U?DngM?-d=<1j`xRU(z3#2hs6bH`e3ze!&
z&4QBIoY5rWCYb&tfpceVH1uc|6K|dIODHf2-ru5c{HkBXtvG#^EQrOan5J0df70)e
z?AY#+QumYO726Gx1~FAfK%+vxi=t?qa=Qzrq;e#uZq5|B+2EPm*;_mXUk0=wUw(~B
zCM!FWk{&K_P>U3k#ymMna7iL51<Y`}X;X4V@E(Bl2-V;Cc<hBPls16vQBY92?bE}o
z*$3r!h*kv)pdt6VbIhd^p9a(32Ve7-&||>heCh@IatOv+n|qzLD5iQolWe_=ftMuD
z@M|RF=`U9rRB*TcLvriiNUDK==`KN|ZekC-#lwWCn`qXc8)P$RPF}wL>!MjQ^}sV8
zeid9TroBG_+X_|>>!b9Dfd3NM*)1{z|N3g<uAunE=kf@C%6`hIXAp#(?(8iUPg%yP
z+wguA+4_{L4Q)QovomUiM7AFYn_t9TVoYN4l30e)jUKvR+A!KopMP$E)ZC`>w6ml!
zO=T2;3_8R{*>yePg|ll0fbf6E_3RdG>cL>D0lwZ#hz*F~lqC{kj6yWY%zK&8^1-A>
z>n5!709t5IXnG344h_+GU9E}yqk}8awK(d5@QcmIN-`9EMJql=M$|1@U`K~jVDVoS
zPZhcTXW+icfCS@+uQ3Ut|3*S|<IEHJhAz-KDcL2s`i-0P{nQuJyHmw3M<JGvuUh6>
zs{19>dOH7eNw|2qtVf#vwzk=Ktk(PLB|+NG0R(bFB8+K#_}sLFKN&A1NBwM4w1edM
zN(q9ws??LkaZ5=0IE#9{2UMh33Kj>%tvPYmF-&|(n1uqE{0~%7pDIZ=>yWb8A22*A
z*ik8%4Kr?}7M17u(Me{ZKtuLDnNSurhBx`*Ktu75S7Nb7JEozRA-)(JA*NU_t|G4A
z+rf|cJwjR^T+1nGG_DN~$aYD5^LI#LR#9hvAOB8qDlGH2LcV(H$po_8@SDc+wAApe
zjOUGd$|d*sLdMxr{>||!!N(6_Y`naZaQpzghr2|hOl#uK+UPAFcbTG&41R331bOhx
zsDv|>w`BOlrm=;fN<9p~G;Y4w?==?C3-6->98Qbj_kgj6(ZBYTZ3$XdL>mf~(E*IW
zqnHoAZXCkg__ldpdgRG=DHpAiem>uYOM*O_?2#BWI}1OK?+G8N7N}%pMy?NXe2GAG
zb`%g4=?V3&psefzguo);uo);|1>dwEv8Z*25wV_uBCn6V{(J0F$!$tkqXGVmcLU7n
zPSY=u5UVUC3c$!9{b5tZCw15R^b3X8(HNpV0`m%}$8aoH7nc8>aJ(1+`zw@6`?cN&
z+AK7vh#qLIFN*4L(Za4(_{{}a?+@O(P0_9-u-HA2z}oolVtc}J`jPj3xM?YAUHz<E
z*|{b7We-+ERo=Om)z66)De}*iQ8^t$v{4ToT+II3Wft_mHAP`G)yW{4iM7_%c(puN
zCg~>CNbzmqsd&}zkMZr4WYl5eVS%1qWc?p&%S3B`X_N%JI=zqg6I5oon~yXag{Kx^
z^D6GKn6~T>f?hltu53)Me6cav5CRRdGy8ywk7Q3#?E?ruMl+e`XcD8zBI9-?6O%}H
zhL~gSXg4CKh~Ew@_cG`~#hxCFCZ>)KLZvYbd@W8xNR1Zfk@?DG*P(lyemg;otlM<x
zd0bm8t#J)<ekB^)AURt<sgTZxxfPiR$uQxf{!wz|Qo+a+uyF)<Ve^B)ytml4P+ojZ
zO$~%@Lz(`lRRSPUZ+wQ;$Gny*tS(twQDz3{nV9m`Gqk|r9;%tOHbo!IFR*XuJy;(E
zoaFUvgFT&>mzOQ(P(-=%tZmyh_@x00fNhHD_qfRp9#3k_IKjk7vmRLHzOk_}Z@WH0
zn7v%?wq@~IO-Ib>^OKgg`1e$^g;&5(g#_l=du6s>R)u}~j$=}uS8G&0UHc%RMLOFH
z78!ze?JCt`Q!sLKdwY%5@Er5MK3dN~sfiR!>VN$B@oS)NQIFHK7Ym*WL~m?N0x>Iy
zZYY`%@Nj>spQWEiRyPZ?l#gG^P<ipf+7NC>@R{%X-ssSHJgwW8w*qitn#m|c$$SN;
zYCu1ZmDz=kJU_QE^o@+%52B&Eob3}_i{QcV^F<^LKp*bVL-fLKuj97E)qY>wrq$Hq
z;xKF%8qa^FX-M0=-hC8l3hs;=m4dcQ&0HTVD=I1;@A<}({z$&FMw_sbaJe-kG(t;X
z;9`X)V5(j;iO0=Fp1>U(eXq`j+t^c}HnmUGC&7B;qu#wxI;qn3VOeAe-*}R7*<w`o
zG@ajgMI<VL$(<i5(af;Ynm5(&3C~;K*N~+o!y<Ra#;&B5)n+Bi4G2`XIm<K|=$wmh
zdOB_$8%DP(TlguiE$)Mb$e2;Tqo=$?&T3F!fL}EL$gSagaX2{+dBJG9lpXgO@j@Gf
zLrhHCRjH-a#h7m?V6fd%IAlDe?WY=~d2}g9_eVJMdB|EaSb!m*{ueLqpL;%Q>0nL3
zgc1Vs%DdKc)eJ;6F3!%}-dC=SN~ydBI;sL8)6;jK=;H^`z&8tXVEAb|<fHAX4@{!x
zKui@ZL*S?gp3Is)-bi;iLv*_@UuX2dX-k0)2=Nn91><}lxBveA<%@D}1<^_W`8@h7
zFHU(C+7qhs?bRj@08Mi)$3Du7NMC-7$kc?*5pp$5@41mr!Zeg02SSt+SSUIxu!gxH
zwa>E>?|{4<K?a1s7Z&Ub7U%n~MJ`{05l7?o$-*+O5&O$(RAvS!XEH4AREyQ}FP68^
z*dk4#)rzhLpOx$58G#pdJ4hXX_Cp{TB9jF%qal2X%m?D*^PTmR<l-3ZkEdQ7BEmf2
zEjfBTdI&9^@ksh(Q*DH`<xbs#Gs*q=+X_6Qxn2b$&n~U&-YyORF>zec_g`jh)nU2y
z<dIO`2E&X~$}_JgI6`|!?aEGiUpkL0c2;U<4AMpha|QA)T+|5dVIoD|o5=b{IKkUj
zc>|n0>&U0sUsJhayJrTtVTa3pRuZmBS;^GE(bB>fBx8AWii6VcM4T*Z@ojoqjDIb*
zU&qn8sT-3`ftCSPEP|OqTx{`l4@IQPGMe%6L=Q1zkRtDkYj3nZPJ66=Bl`1eqgZi5
zafZ?u{LRubNeSL%*+cA>@u+W#pVZ~LEG?+93A$?VMH3=t=pUx~vZBynV-0pzT4PO^
zRGJHw{QCERZqC^t-4$t6Q9sB^9y3vHiTA56V|n~+P&*fWqAUv)J9AZ@B<5GSju@#p
z(#Y?obqBV|bqM?jlF^8o;A8<73Wu#!w>}bRA&8F<m;qvO%Joi;j(Of%WSpC!Qg^wX
zo`5GS<cO=w1Bfh!t6HYjzy3}!T@bI&9_kfbbcMut7X>!3AP0a0;j#=Jd??R|pnsIt
zA5#kV-`VgjpMfF>ojl>QfBR8~**IPyh?EMz3$skB1beAfREV2Kf1FrP{l25<F&J^;
zJ-1}>3<DYd&fo{mB>$BS*!+x#WwzKWyB#ZEE<_<l21KSWKtDRVx)brIzx(iHDBfG7
zUiJB0^^s@VLaPV3)8p?jU=|q}yf~V#8-`kygjrjtl!o4J>vO3KXLMq%ylIQ&E!0&P
zEZIf<F>sK8Tu|*cZ{bI+XS9P!k5neKFT4%F#q))}6Ol2=I~b4)PTc|Z0Y2l?-R3xK
za{j!Ti-W;O8%^8g3qN?+wJVW)fz|Y6s>~953Ng9{+#xvo7W}LmCgo&z>?e^DwL0qr
zz#8V5vw{I*&-UiTHhysZ=nDEYsD((CuKTM3`pJONQ~lcuwg5g4d^4)cqSFiaw+nTf
zySp3kz$#CqBXIcD?$E21+agAQx^Ft2!~F3TZt6+N6Z#KnPQ{8eOg|kQ9qYJlvgBMJ
zA@YYz?JAXVyWz-k^9n_%qsym=2JoUbru#=D4-(GSwO27DT6}A=4d5j>|3HI%#n~Pm
zy7^|Lre8KZCqWKL1B-l>IJ3NDXD?d(&eYctzHvAD6_!ZTCyY@KF}u(Q6NCa{IG>Bl
z=NZ%H#VB8kV=6wqrJ1;HG{;f5X7U9BXesJcF$lSjTw86>*H;DEXzY~YB1;uYOk)JF
z<rC#P;*;@F`<H*8K48&K=hHTuMP1#gcZ#-+|An}K>TKxp^I$Rf*x?uF+wTpN(qha#
zfSiKkik+ohA!#9C5>U7RC21jbk}_tDBaORo(7!)et*-R`?LA>ZL2$r@dm3zQQV`pm
zV0n@5PZ{y$SgC?wiACwolyxy#D7YFzm&|1<%`!a`xAumbPiK&ijHxAYKczeevW0#g
zo^u&lIX<r8cXQNFx;|abLLBeEfB$mPuD6&!b9-P{iz|rq^`<QE_A&AWObWmF?thlJ
z_c#E0Hy9mzAR@_^KB2V*FYn*=^B4L(q_5LEV4GqZfQv|2fjJ1)YQVly9_D5rT<sJ2
z#|0ptDLhO=X#!D@!k(XGFmQ8Wcw>d-U;U6r;gQYvq2k~yGu>q*ZN}tYUF6ny5U@#5
z$@@&B=j1LAf_3f0(Q;KEW!k?s0N+)-58}X^mB(pI6XA~dENcUrKmPlAb{)x95Bj(u
zOYtX+%Yb~ml_O;?n?r7Z+ZW8Zh^gRk0kaVCkV2>v^wno`qu%_lK5w`!DeKRz3H}Ne
zxuSvD9`SSo&Ip_c#9#~~6dTo9`|a`58=Z}BT&Yf3LT=o!MPY|{Da11m3}-H%f1a>_
zClPo{0Z5Sk#bbER5mR)X&1bWI{9i*U5&$D2I7jJQb0^L(PGv&_-l*Y9<2|yO?`&Ve
z?Zl0+Rvfod$L!?nw5LEv4<aR*4WpX;q!mnN^?@3bR-PfwdW1R-EbRS0F>=byZ-=N2
zr=uC&L0RCe3>51a+7oZ-Ow{?nIlZlqoQT!c!OKGu6vz9R5V?q6{7zWL&d+&WPOQ5Y
zByV_bPifBtko3Ie*Jhb;=G?(bTr-)!H_j>3Z%ZGU*7Hset4kcoMak<?X0#dBd(+f-
zZVtD;0`|4>)!BGOf`w?opl&>T4aQv><bo!@u1+Bi?prao6Y7^uWNbU*<+tK+{r>)a
zi@+QJoCKl5D?nJFK_4F*i_s#m{FZwQvSEP)3T285i;kUP#EiSh7L$Ht0Fafzb+`d(
zuq2A7SO9t;guq0P7G-v1W?8afS_|L2_|8)=%UT^wcj*@#^2{!>3-C3CxZ^6Q-6d1w
zhcny9M=h3h@>IElIorqXI?7aiT$vkVb94R6*Ne;cfY<HAf!I-+FP+(zI^*7BHNvbU
zJG*1o1?eUcv?EV<h5r4Z`u857>9$|uvHfNZuu7be7iTta`m)@bZty5b1dwV3_ysDi
zU^f8fV8f4iW`Lc`V=<n0)Bq9Vh`1_y1U(!K+x*b(YBxDD!?JwQ?$0I`7$)5|ysF>b
z-%k5$*75JN;5uF3yIzZj(Kfrj?Gk+DV}|ZHXkzUag&DmF=nd&)+2`iLuRz2T30!pQ
ze8s?DO_=)*S(g-uv)BfxI>I>>mX;KrdmZ3ts4RS}S)iB_Tm3V01CdG?f*87DVqzQ}
ziVSTLO9a3xg#dp<VB-j$4D@FRq?A!5Gd=QdJ|SQUw&Ra&Jf#1YfBB1`gd?zLU<%py
zlD>xigc~-l83nOjoQk7W*RJ7f$&jqVZNJ)m^mwP@<eLNz;;ps6xjyBAsR3F(7~ZH6
zVAclgYQY(h6tS5_G~Wnp5`0gM0o#W3mJw*@G&kyQOY53ulDTg8)*E%-uX)0yjGCQI
zLcyZ(^&aCmG3^0ExI9B%evi3X-q0j7!S7ACz!*H>lYi1I?lu-B?1383Nwz7ipq~11
z&l=6!zZ4fAMeY-=B8?h`V%JAjjXa7T>wARH^H~qc=b3I5Y*M3kS1I96s$rMhHkQqy
z?sVqyr9H92PGpfIcrqu10pJkR9+%SHF6H+6FjK_|@nm$dGGz)GjFg&VF19qzy*nsT
zsu<JG!}&_6e*|hcHfiZl%QJkNO4ok`b76jx!INf+#q4Lwp}ZR=UoALLIq>O6_?Zn}
z+it9qKe4-&Dh#T?v4SYlcca3mU=|CANHfF<0GG4S3zdfXF<4pk@+!d|*PdeS(i>Bp
zSJY2ZwfH4_*~?h<YT#PH;4lZ9HV8gb7|`52E!d%wesQ{KT4SGCPBQ6^<W-6uc=35N
zOosY6a_qTbfrWa>l)6O6e+ilo!)O8GRF?<TMjWGh&Uz~@e~vTb9j{Ux95AFx;6(@!
zmi_zCy<L^R;EQvs#Fu~^$9KW^2B(2)7883RPWVcPIUgMVPsk|?@5)N|HM<=;4~$1}
z6emOf31GEhaqgH-*H~Y-JwnBiEMRMZiRzeyRu(lkLCUsJ4Z`Nup3Oqfu&kSdSAQE_
zh&BxD9^bt`^sAIdWpy@pNy0-nbh?h1tfW25CI%6OdD?`i8}suQsmrI(zs&mFTu)e)
z38X@kF#p40bJy;&>SJ3yTqNLQ*ZZcWlk>Z=;_s?zVF{Lev++DRfS^;lVENBCir2Sq
z6)8Z?*ZaUrDVHe+_EONcEgfc_Hveiot^%%6y7$Fq5mg$c|AJwXFxv%rTC>;t8|1$h
zLn1y?m@z-T<q-$W;$szkniiUFYfh<~raf0oOgC1o2)16*V~jMFmbP?tqN1UR<J!MQ
zdV~?WO41emK_=JDky*UqR@H0TRYCJAwiL|+b3K3db?nJbOoDfp`(`#6>SABUDxb@$
z+cxS|8CTF__>p~z-de#>xq8dsuw0)m?r~Qwo+N=*3c0uwX`$a}{vK0cM!&HR@v&tf
z6P7CMbR7g!*f7Kh(dCvUnJY8Qx_-=*2nk!<p*75@ywG_^%ZTCGyUtyM`kYOIF-s&Y
zwlzGUTRui3w#n8+UhG*S+dX|seycN+(-$PI9PI~QSxjMY8XL+1UNVN|%AYU0t;q!`
ze)nu@5o#9|6_u?I`K(I39(f%h#KxAD1-6O{brWr4k3AMAc`=Zopld{PH{Ee$jLTLp
zf^K?ZG4YwQKMQ+0)iCge)OWZmCfsytic(5|)$JuhG(Tfmt0^hM%Nswh@dlUAy>p^8
z3Ek428(>Hai)+~SHVJ%Nujn%xn`NiO_>TPM?v5#06w4m%3Uhu)F<*q|Eo70+cTGuw
z1|jt-rY&k=8yOg@&ufRKZqA9~{j9v5c#xvXyc6#}J8ip_+`3DDlExLd7Wp_HC(Ds{
zQKLXi6Gh~WHKxn5$F<e{Dr=usfy3)f)0<d_E54DIYi+OFA3Au*R<_!ZsI>12p<LnE
zJ>Wb@(wC4bMHP0syMPhHEj{U(M|gyTrB<f7pJ^BJ#rh=U6%X%kn*4SrB`k?mw`VMo
z``00ys3IY1m=rQA4d}<u!Yi$5Lb&n(O)QqIuVr@+bwX)5GnSpgnbj#F=w8F?wEZ>_
z|9@NnF=}!ir%sb|Z=1=ZGJ^N9f?-^2;m0^4{iUaD;yrk>1QV5&WSb&gocmTpSadHH
zB$oR)m8|=`p4aCSCR)&akc`i&d_^oSZ<&aOne%P!^ia9GMqlravrPz&h=&*L?STeK
zdAx}Ds{D^{7>=XNjp^9~0;qFQ9cY95hTJd6JdjI#cP|m6Iv;x3lqFmQT%4TkD=S8r
z?QlD{?n2HX@QSUMnp42o5Ao68P?i<Vx4x|Tcm2jx{Q-m&jBd=q1FK=goto|UqE;2L
z`2rM9<<|Ib-UxI*vY&76V(70Y55Jn=eO;y9ZBG$^3nvdo5LKqUsi^X8b(RE}!kB9-
zhBch{Uci}x(FFQ61Y~4o1&r_fRQ78!NIrBsFpGq%YUaDtzYkp6EautHlpuw|=Hz=)
z-R`*}I(nG!QoB8S_ndR)>Po<Rkzce-Yvwbtl0bwo6$NI@k<1!4r8uL>%P!ut^G7$2
z3Nt;w-`nu8-;sWMy(m9&^CyW^zzL0~;L+Ol%|@{G{3Ogdbdk?L*J|<nW3$FdN4t7x
zg!0)J2U*0D^Cb?-29mz$uhBVM%xtD$bd=1Glq{CnCZD#{%FeN(;$Pl;<G&EloZC}R
zh<PWPzse=3ZSML>zJ*x5C}471Z4Hkuu#=Y|l8(t)l<|>$hUASd3*nqgRzFstFxI5|
z-jy#?1XlokL<xyFJ^z8YQL1x}KR!N@%$Whvas@u>aj%yJb$%c&x<O+n>u3dCBpr1x
zC;LaWMQcW)sPUTE8bcD8y9ev~>uXl{wb>V(`ZqAm4$-hC0Lo_@?6v3lu<^MeG1N-7
zx)FIw@ve*F6GB5B%~i>)1P6UlBl6g0^KY?H1p36tp?OlIL-%ZGtE?iAz65rFZUI4p
zg0*<T6oQEg(X)xr{{`Rz#jeGm@ANv}3HUV;OY$%7Z-u^f-0c{9N>g?`PUH^UJYcQD
zMF#`Ul$C{m+XU~#vPIf$U|-keR}}H7=9W<&EpON4Y`DQS5Iee|(JS})#hf+2F#q{#
z8R5?<OFk$S7kU7FlN)Sr^P~!s4<~}#cg>Io(6yVPDK}pqNcgfiKogduypn~q!Jb(<
zPfjp}CVGCngdh5ED+|f$DLUKrDGn>uFZk2TiFI$*hxl?dtv@FCk@-iqG=(loibu_R
z$TwbH*g|zI!P^agiZ3*olb?D#$#`I6y*Pc7+Sk(HOn2Y=hdjqsXTl!=2Nt-`a#a7A
z474651C~kRKMZn{`>W}_)WWa7bTna{8L#c5?g{Q)qN6NdZZ$lwK>3SS<@aKXU;WI{
zj4()5mN1gp{z343DjB)ErP=aKpanW%IHQ5E_$}j4pPtj@$l^(b2?psc8D)7@3puXg
zj+gSdbYRVYf4k)rHT6hXQRSN9ikfyZh2mC$I$;krUPk{f88J5fyzxZI{ZC3^?VR{8
zPM9z`^^y|zAB2_<D69La%d-iDhtcI3<IB(_?l5{As?+AnQAlIAX#0^;2uScJ<*IY=
z@P3maL9eN)BM78v-NlpP6366j6q?g}&n~_<SeH_BA>p}KWH=T($gO@za<C=XDIS3T
zB003ngJOkM>b3PaxeV6V#Tlawr(3WU1&FlL{n!#LQ<Jw6Yo7bR%7Za7vEcT?{QNje
z8#RXdUjC(#g<;}Y%jD5bE|dJi&hMIZ^{^yE@5kn?yp`kr0%5d(gr!>MZXLT9=Z_aP
zLURKg;9F5AiDN)9v(x9XZ(6l?OyvTKF`nPnWZrfY2F)j}5+D#jpq8CoG_1SwTa9&w
zDnwob2^7F=;1R-N&Y*(FepM3Yw63kG5*OA4N+~r~I}Y2ji6%>5*1Li&w{i@2wZ}fL
zm+8CS2-RJ@q}07;L-Fq#O8xw?k^WCdmHnDDpgE@v{2%cD3VhzUJ$kr|n!M0H??j%$
zZO(H0k#0vkv&Pc2LBX)AKL(wr$n2(u815TaQ+`UHO;l2WU-zr-Zd^*5qPX{335ih6
zQk{7SANjK8Fb#Z|ezcoDru1xj@OiL2$#}+{{+CARcy)#fr5Kib-&Ce~E`QrcM7S^$
zD*B14RWOJuu%aEf#i>sl{kX;Bw$i6^sD5*|dG?aRX}Qm876q=)=U5niOtf!3_q4_&
z$XhVb3TWI?<|c==R+;F-@L6-UrDQE|&yKre^m|AWL*oWbwC~M{-z`YcQy;7@dLaXi
zhx2FZc5cNy<BzJECFWRN^oC5Uud3s9beI!$zHZzVj8w24ebq_*CS;j_|BEUcxBWO_
zo%OsdN1Buti)LQesLpa1*Y<8+yc3sF{}vyO8hs_XbP+K!xxZ+yVwhr3pb9;qcxb(l
z{#g*M3k|A@@G&89CYZnE374=9ngFV`_V04H5K>ftPz78@V7pI|yJgBiA>dPZB2}pS
zErMJi5^}Nq;(_-L17oTmuM*F3$+%C+MEJ#36La~!87B?jbvFY{Z{C%0Y9(DM?ZkwF
z3*NzeV8IR5oPyE-*tWpfV{MC9x4<0dRVq-<7y!~>ClJmsC&8GKDQS?|ccKD_s$#f6
zK!S;ebOO`d#g}cNF6-7(3wgmPitiV40=0GQ$QFu*KYYGP$dXmjYtiQ6+O3Xw!EA;@
z9*OMYW_PbvH|hoM7Yjoy@|7KMbExbmY1U|T&1SX>dA>VdJEgz-6=>`FlSbiwmFFDo
zy_Y8(_g>b8+@@OWR_pWr?X-iDx_=_w>vR^(^JrB-VEKdbpN5(6=Mh=fcDwHW*z;+?
zLiYal<?gKa3U(%yu*@*hKT}xE!LXPxy3w;ey58?_-742+8_P$llaXAyX7t)pkSK5)
zL+j>*D33YiohtdcbDDT?xGLtRgf3TKb#C9`0&9uq&Bg248Vho|b(X27q)PoJRk8Or
zWDSY$P35m-@p{xunC3Zgr8*?n%%=yZyK`u8anCWPkix5fG|<t<on-kNKm8eI6M7;~
zrd{_zovUnTQO;O`!sIE(ob^3s>>N&dEv~5pS)G2v5-mJ+Zu2ePUX^rNnb-GqJq$Yz
zg}0_y)F|SLG&jjf6{{6vI%}-`I6u}-_Ffz#kE*0^DvWm$;5iZfC_xQK#SRa3RxFU@
zvig8qOD?(l1+@|m3n2El5u7^>XAD`=bQs+ymX}8Xa2uF?)1GsQvsH(uE)T0dS<sDT
zjq(60M`+hjo>n(IW>G(8ps6E~FBPZrIX^J|LR&xkQ^}#DRJbyqrBQsp8Bnq1Z=?(H
z_%sA`><pO*nTW%opD{F~s%SpIlxGI&`9$xOPISfS4V)y^BF%Bk0R*dcw7|9j!kVnC
zSdB>lwFYi=USZ(|pRQUnn(?24nXzBWa{@@(hG-m=e8!ukvPjyA{lfKz!|!)f!{d8r
z`)5o394|fbaEIDft7YDxm(AoH!&|HPZ#l}RSHydDd`9T!w^NF*=UGA-Ys)?)1&^Dy
z{c^NQ+AK_mU;SD7G$ri)?)Zl0(+&lyt(!S3IUGR~5M$@)Q3d4!s&sT)z{8TQEi
zSWQ6{50q(r-pFgxy;{Hr9ejPrcFNA?xBryPALy8=R>;XWQp};yMzG`kZNi;Fnt0K?
zSNP1@g2()ZO(u&gC9QQvLWRAHN1$zpRe0gkv}J&rBCd3DkQhl{GWLM7n7T!hy%@8l
zw<ra9fLeM1a)Nq;bi#B*9>r^1aeUP3s>q;jdtY_r`siG9z1u;?rcd7<oN_ZCRi|Tt
z6H>>&XYbBwdrU&DpW$E@Ct%9=G^Ai<Q$-K2BZ-lqS}z@O@9aDXJsGgCv`!O@_=I&r
zaDA#2e@Kd6pd?OoNV1OUO1T~*_HK^Cjpw{noY62akp3P~t9+^+9pSy$zq4j(SfoqX
zayk0d7T<JA$lo@+KR?{yF&`jAfGzyKl-Qz(`^(Rt6RS@!={_bJ03}MWcTZ772g=vy
zS4sn%7Glkm5ip6v9<`=<>_awR^n|Z$;c4CdE-PJ(;l$0Fj5Wak<79Vz3!4OGnm2Zz
z$c3#syf~5+t}DRHcbqr>?KzkL!x8`_F|edUh%zvcLGBZ%NiD+38SPy@_B>(Ens|aY
zMa1VBur^kfQ#oB;d<tXr4VS+Dep?7#2<bbfA~)7rYJM;!))yS#zh`37q*Ggdn2RY!
zN0P<21oje-b0-t&=B;*Y`wt|T1MMD^pZ9+-ZBN*qWfL;G7rw+7oQoq8kQ=2FsbT3~
zc~9&tD10vK(yJ}gTVa&%!eDy&{I5xl%ZYa#mB$xTp?~JO!lC9no^oVU(N8W|xwVZ=
z4cuD}=h}DZiAF4+_J*Nf9t(yh43)Ke<%ZU09G7}s*PcxES5s<z#l-#0c14V`?H#2m
z#+a0yTSj1|`(Eb8Fj1$%?CUl}Ws8a8N_SVukDC4-@L;yLzB`J8B-X3M{aCC^nF)iy
zMyr&N>l@?6UWRj?teH^jZa#V*OXOxc_np3B<Hw}P%U!&2nBs%ZUmG4<&(Nd|8j~>U
zuQ5{Rx8ESN*l+ra4<asIerX)K7G}QwC>1j6C;6=7me`nYnIz#F#v9(3(yJsfAp}f#
zAJ%m44Kj7V;SDf9JCeXvMa?Or{U&->Nu~ZW^0|{$IZxJZl!k}LqmR^nB#n<svWuQL
zjXpm8Gk>`_df7a><{qnKc5mnKqUG7*+cfGu(jyNxCVARTP7+o5*J=aio#$@%7WY}*
zQM*O%g$zi=3rbtm+7kt)Ic1c+jALJ?tur2@JlgZaUECz_Kv`w_c>m$Md(6+xg(f$d
z{t+0`>ZO|%6%Q=BEDuB6`d2e<Fx`+F1Q?Qz-3c{^E;<)7i-WMqVM(9Mw);TPPBana
zKDp>!6ZdS@XraO$$%CG-?c!npK<SxVXSU^1P^)I|XclX;$tSo@@G@js!K&W7?NJ<K
zM=d_XU=Qkbhmj~QDu@o@#+^LFH<yi31!8dj9fI|~u`#t@8ygn2I_eLHPQ9LFL^j*L
zZ?>7Q36%)GXaD<K_Co^ZQ;qfo*@hQ}4>y+@Z<8STCEIIp_dK&7$1?HUY8pA7F&srW
zIFJOvBpL{sT_?H2AoG{-g6!ul3ipAOwocO{aq~^}c>*lNl0e0msPL#I9u{O9qhO=L
zoL1p=F5<T6c7qDI>a(;q;a#>&rz@`6#L$fDpO%EDPp?|T=XQBcvC&aNbwnnU;a&%5
zx|8$g^9pw&1a^Pyetm~kK_{^X!XgH~f2Tl@oYot>CeoI=XIfAA`akGL27lOo|5}*3
ziVVq+F2=4QNm828^^^VK?}ub)R#Lu}EWa6<()hkp`C0mvlW4M<AIz}@7rgo;FyE~f
zK_%6tweyqe`EzPNyX3R&VtK{{+9c%(YnQJDHh!udWCvM%`gE8kei(kZ@?(Uzua3{k
zpT@JRs|7YAv*d1T8S5~oi%~sI5aYqBo)g-%P84gCp^%n)n4J^b{|e&@X+b&i1!2O#
zSB&DEZ*}vZQU`wiYtfcat}#w^%~Wt5yhqF^oqX|&s9Pug2z86!(l%Nz*ccD{64~WM
z@2}IwV}~6}#YJk8Ez|YdfPrnSI2j2w(mtbDU$LvR0hX<{T}e8|c;!Ng`UP`{@A7;$
zIy8iJ?5?jf2QktR{{u+A*o+fwPe7&*AGcb?j6H=6tvYgRy^Zd>GD0b1_#x*nAl{6d
z*J#~U@O}<19a*Ca&z<(qJB+AZgfcfE6Y`TLOSP3MJva2|pm^WC1i}#@m<1Nqm6g<J
zGH5OeSJq=r4GoH^gX_RW{BI6Oy5gGPHQ%gQ*nv_B2tsxD-0o?9X`rYW>dp37x+t@^
zZ<QH~eSX=q$ZhVO{ZQUhap<6;ZOl3!dhI&uu|!AkhDP+R9Z)vk0nKU9%-h@BAaR7M
z0=G3|$N6&9QuTZ5+P(5GKZwg)5ALR)tZw)kzu&+Q-<7-~VSg1x{kCzP*x)GV{`K_f
z4Pu$_-;xR)MW-zB=%)v5?wqxH#Y|GOR>>6Qc$mdwO2Q(VcgJi-i2)7XJfY~*k$p}C
z_us39aj_cZJ?B-pB@u(XZA>z5>0&1wRn++SqCDwqCQ?`d*7F6TPA2i^9o+0}`VTP(
zR<e@^$F`1tpWKes$QXR3MwIZjE%=xFk!hM?qs)pl`zmMmA%3)kn3#YfCIQiSpq~1B
zH6q19@d+pOP2OHy$)e#A)l$tp;Ya&K%C9+X(vp*z6jQ%&Z}K*Zi_LHDG+*<lduKU5
zQ}MaE_4j(zcC6<iFE(MU2OT9QJBB$jAus*{O<&a6NoI>sezeR7lO}@z^iL$exsF!z
zUUsVUN=tn3a!1-<d$^k`GLOS?vNyr|0SZItv?dD?LYlRCGxLUbw%(SiR!3=mCgdO)
zmz>C5J2yA?cN0fUwj!hdnxML(rXyohIjfs~4!OPkV^w+$dkTffxZg&Es2;wa&!RTm
z;HkAZq;8__xe4$0Fg7;6lapV}(qa%Lo=x3?=CZq_)V7=?p6{zT9xydJCPT&EX!D;3
zk;loO7<0KUPM8h>u?_-gTZokLGr6qh!cE@b^URGZh`TLsInl{HE`7aUq~aC!t}&vx
z7z-Vxy!k+$qhSSkX)j1%@2F7ZI?7oeBvC6s5+MbC#noprLfxp@uj@C?MVwZ2Cmt%v
z1qn>Ivq0W~V2is&(_Wkb9oA3vSt1^+#Ba{}5So^jq351wXN;lKQ|Im8_)VQux1L0W
z(ghx+Vw8N$-uh^X8CHagyv1p6$(!cds^Zgt9%LU;;9IbzEzzVvB6WvLQSGbQxMS0u
z`Z-eCq|DtoxqSSd>lhY`g<ng;mz>AGO^qoZZSm2SJDr|m=X{lz(!^S%6;SUu)z@5X
z=hcvV-C8iVd(4z5FZZr+bBOZE*SfVEDxpyZ3<>VypcD>Vya}o!JH7ihvW}zoF|B8t
z1HHcNnx1^4|A#HG@{+~gnvtSb(85_Zy5?sl(Q5WF&YL&C!-DPyUpZzg$&bk$ElP|R
ze;{WO4D%xh);tPP<Gm++>&Kcl{rFAvqY~d9G!;4A6|{>u<wB?vi6~O%l8Vte-|Gd>
z@!}G`MadUV0L3hIm~p!D6|10jj>0$lbcU~qzX?GcX3fvaKnJaTkTD3~7*s5)e)^wz
zRB6ZeXRXYJI4CJA%lml~N(~JU8<U{J=0pJKqfzm$+SZR~@P#{iF#UY#A_c{lJP4)6
zm(})~0FWXVc>5DF=~d7Pw^iH~*auh<*aKqMi%>tpYBxhMU8)&~&ucXk%8+A!J(9)1
zk0xFz5NQg<Y5kpB=)F|I3OKb3O250<>uf-a1bhlv?++o)U)VqN-X|u0Q>*hPm9fn*
zyh1;{1PWEfj2dAeJQHk#ISV@jT<3jQ5;HWOuC*5;uYerP?0Kc?u-f$@+n_?G#UN??
z5mdu>4s{LU)JF$!z3zKuX|nWy1CyuEeiicbthBT^@_~1aSJ5*x5xMstYWw-o{HjSw
zAqD>oO_qMs`&lYOEuTMscFmN5ASFjf!ajS*YqaLS^vRfyzNPGA@{0_Oj><TZuqNIw
zhn0y;(%;PbWs^KfAn((RfQASD;)?S%s+wqnymBvz6K(Kt+_b%MzwI20KHs}SV#Ny0
z4OwdMIYwbDejXOs`E6LT&)#HJTY|i}iZx805baIsv%HmnpNCS?549eeti4yMQV%3S
zzoxu4$ihym=j5u2UGQo2N4Y%Z=Bf(wF6?p?lJImSGw<!%JpQUQA5o?oYOXW+>Q!h>
zvUQ{EZsBRpPL`w;8iV)?O-(7W7kop;{iCYE9NcOl<Y;Hc<D_p`gA_{@dBq8J#(dK)
ze@KOXwP)3owPFdB&CkZH3KNMzRq*WRx0#oGmX-9=I>zYEW&pfVaLVBQ;l8EAR!iNW
zF3zP-t?J+5tH{_jISIT-JWn<*(mSF4(z!JKEjdSwu@rswc_W%|-qOo7Ss2h{wvQpa
zMD5^X_q~^2IoMj4q1!UzG;()OeOn)1U;%V|;N>`n>ad^45#A7mrxaK~lJ=iJY!ed`
zaB<4sh<Dr!Y#i3;usiC%c65w~?bg-RRW<%G0f~uF^YxDH?Umw^a(CHnoEfLbu%V?g
zGd*n20w4keP&6;}?9t%&!2S(rVuFCUyDeWX)tgZ)xvqU^UVE@>w7b-D<Swi@k*8C3
z0CJ*agxHNKtP?_^1lQv=&Nh7hy*9YsdNDXW2&^N76bc+ut|mmT2^<_805ynwcJYRz
z?dly%!T|okp-bDP{?hq~3-{y4auj_C*D#>|e@4)IUl|*}2mOLYO8p_{+RysoB#>@R
zJ%3>jx|{0Pqm-V@h*t*PTmpg<+)&QCaaU=g0fd>ZAo>~*KK(Aywha2ScVIM#(Bevr
z_|rP?E)A0}?V%%-<`jL#kpJZgayu{8*xX{4`}&S}%NNu2g06cXJrjLSU3~ifE`B-x
zzVWONbX8h^o3=SRInAKmh=SVNe2`F~Cmx352Z;c`_^v=9Q~;MU8?oD<?o#K)@iyWe
z>qAHwil?@2)@9~AXjRJ1JBErRjx4qBw24Ua#A-?O;rP<4abjaHe3P5wFD~ZdSI>_g
z9NTo}!qSg?uT`S-Wg;+M{L2JG*c$4lI+67PZ6tw+#6pdoI#<dn*@Zq%#SdD`edFIK
zWKC~#$k7usq8A9%JyM}hyY`N1`Lb*7X~NJ{{`tJJ&~9d@EsuC)<u;|mBJ-*TK7LEW
zaI_YWqSNa=OW}D9K6mM?*>Kf~2W&jjE)vxRABxD5n<BGG2|v5}G5M18H~u~HaBR8N
z(oPc=7oaH9p@m(;l=tAVqse`Zv2P)Erkth;CV68KrmG^g_{NYqr^lfn=hx!{$76Yy
zhmWtPv$Hd{{25IYwexv+3~h{cw>_I&a*jlW9>?cb!(=*T#Bp(P|C6OpS>mC-XJf!T
zN00gXIFin(DxlE#%OXN(4|iqLB0xKdCRGje%=iQa-~Qt7mgzO(D@`DhFu{ambJODD
zx>Yex(n4MFjY%hANP2NGOcNK^&k(;}o&nwe){b-9yGAy?CcPOT1@piC{`JqFKh?<a
zNE8<zlo(!zy{#Vv>XDdn!-B_(AInd1?*$Xq-s#Jw(`U~v#-iR$s>$rX$MHn+MQFHi
zy!ZSeHxF@e@lkIrzD@c(@+6!Z1JV0+O=O<%!qbn?aX~P!UhFJPM=N5@SvqCC2ssO)
zx%;<IbrS;&y@T2cuMgW5ZJF>C{HNiI3pqegpOi(e9i?St&VS2@3>XJL=tW2hfzlz(
z80Uuj<izYlPg-NosQx!X%6G9-P6(Es!NqtIH)KX5WM>EhJUo_PliIG#o&SFChb1UW
z%sF(=`_N)MSg3RBd%DyCXu}yu52;$=BqO{t!5Q*%-oiP;A8YaLAt<p<(}6Ru=P=CI
z{9Edg35bGlQUwYM3IrX0e~X>DM<Lh_o;4#tcpgDK;|NwOq)YXgC+YW=7#y=iKG|Y|
zBu83;9>nu$Ro)ncV6}jFB+36a&cVOawk&w@?_hQlX(I2*dLFAB5rZBt7(-4HCc<=_
z9UzYCO=$dyrKvc!PSJYs|Iu{a;Z*-&yCm6rXK#{>Lo%~Rc1IzUy_K1nJ+oIF*?X@L
zLiWx|5>iHXW=6>SoZtJtUFVOwuEw|X{fy_ipL<xA8(hC$x`NAvJHwm*aQ)}|C~nHg
zt*p1*2N*u$_YZt^QWWy03a59LnixWD;s)~fe2Av3YqM8twh^Hw6Wa~^>c0JM8Uc*j
zJa2Xl!`yC)kOVVpQ`b0NC?a&t`ZWCl8^rFtR8MD673#)eS-r)D3Y}g5*=t0Js*Cs(
zf`!>-ZuO9GC&(({?+$%&;}(Zqq4E>%N3m7e2Bl;6jiIzpm~F%r%4T!E*k`+<T+gqP
z2rvui>-K302QuZ~^iF6f!I)QORqPv+HN5R*h^S%E4J_coMN7X8`R>%@lB~r$%p%g&
zBeZb*q!hFvpAagPiS>7I<4oqyZmT*s$MFLV8ZMMCl}5K-&dLz;%5-+ivy#5=5$^>8
z9URe62TauQCk|9Y!Y^FUT3oyDiX@{m#9xAk=B-EN&zzkTN=mprAr)4U#jO*=8h&kA
z$2}x2yE@q5z^%;Vo>qcX<r%oc%d1Sl%Uns@Cs2ZnY-+masor^Lz?Q<|X{cC}`5tqV
zs{Q2p72w__ZlA>d@}Ar|ZUV;14CvWGpjtL(jn-G7HMbwo+GvnU3ok7#ZOt~uf?)~l
zmTMYPr{?;l_cqJxUeK+Q43HQCZS(x^bbAWkGT1~I0E=SP2{@kraXy{#Ux8rv3eY!C
zuA<rWCe-_#RBdr)G511W!VLs*R(FUn(N-IVvhJ}<slYqgWxU*DY^RuiqNn%c_H$1>
zC~nY3r)b7BSLD|=Uvabs;6F8Z`{h}?3#bcnZD_6DI$;5`bi01qxf6$zSb+q1vtnrd
z2ju!L3T2(*?|4nORP(+*_#_^ynAb7&_~{6G6H<qww??^fE9ziWItz+2iQF@1x-jVE
z>(OO-e0q10)SnWx)q~G(m#ii9U*tsR0(vIr5#j0G%9pLL(bV7lmR<O!VnJ1NhbT^3
zqDxj-rY>=rhft>e{;VnSr3BL$YlJaPXiVigd9A+l48AtqB%{aSIR!OJC0j3_A|%Sm
z-*qrohtI{@7FSM_*W7-no8ML9j?MSTywrCNWK_7R>l`n?gc1Cs_<t_Iq#O<_nS%B>
zBH6I5KYL*{=85CaVk<_aLak?B8^q}%j94|D+uJQRb_WPNw`RO;c}rtmEtc!*)6{=g
zvQ;l+Ex1@x!e8||MUjT8Buu?@$~-3jD$%jUl>0Y2Ydi|!<w7vw9{J)V4Le%4lHWe{
zHICTTi!shf);TVwvM1EmU)SWjBfh6%*p`##B@p5Ns_zRc*e@pKWzJ(F#jQys9Y|S2
zJ>m6bQcDJsKfFIzG5Uxv*ho7QldIEiAa0)n9)O?<iMUJ0lWK=_U}URJF8}BV(n&>H
z>);4XKwjH`I`OZebytI2B{VVm3c8ajOm|%T^5nUNrDZ6fP4)JmDF<R&LL`)CbrTGI
z{P;c#oLQqFELSejfo!&V1GpMg!Wrr53#M$f`QFSlKF?BP5>`)KEHHh^Ex|~&^J=={
zNgq1wYh_>+hxZ~}K#!M_v7Wopgc{8@{&?)m*LtV}Y70`+!!W5sW;C6y4YZX9=!`IU
zX8i=TxNPMs<@BL)!iUFj`T``Sz~b0kiC&Y=K=^k7hG7jD<oGPC{@tuv1-0WUC=pWn
zuw*{M3EKq3!8bQL_~Cv5`(DG^v`tsB_5&f17y%x?m+<C8FdVFWD&61Df89MmJCLA}
zX1EAe(BB>$@nP*JuiJm5n#JflZ=J4YjiRN$1b63{v_CxuBLg&{&r0ebZNDkRq#t;&
z?x3tcSMqrZ8MTi;EK)p(n?bK^s5HN~dRz2u!L1bFF#F+r<<8UGs(N^4<C9LR{fiwx
zKhL=H0Gbq2!hl<U?rv;+iO2e6nPu(gU&>=$8d+(RgbR|td>Om_AI#sri=i^S_Fc4g
zF^J^qxK~$c?wln>DtKK;y&{flaQHS~NpAQ1&yDPlk$VE}chz0hyZ+EGiaK$XJn92M
zgh$)x7O&4~ZtRQ2gQwkV4I^`nn~pxyD<9XGy@U?B1-P0cORc#=V@mD1wdLsuZ%O3h
z>6vg{#JS!mMX6GAET4CCX<Jkm2ImzYO1n@_&LAu~<j6yp3Ey$W#^`*%*HxN+Bp@ZI
zO05v=ggsBGy|$<*<zCF$UZuqz#cHb1!~PNJYA7q1^NCJ{`>;k4eYCT`S2L}pdMd?S
z#ngNz-<QeNp{}OJyZ(`EF>?`*@7YV{intNWfz<=o79-l|Q28><K~~Z@+j<@<%srZb
z;?uTKl(5}$=htm_eK4zra(NDFZs<9oy@!fi(32c@FEDArUS!3saMC8NKO&)s%j!{?
zE|dy1jErv<i%h72nB^36?FQkPW1|I)wYj-GZUycDj?*inY3?^e)S=7|u!r{g2xRCe
zt-JGlu_AM;&6Fm7Uc6nEuu^fR=&NA^R^6L+%ml&nA>$)Ksms&R1GGT_`dkNhx@fop
zdLEGGEKrz|)uiV0_D{gXyzTq8JKACL4PCk`BYW^)$Ho>3Vt@2{@wwQ1>NmF8zNrtR
zaahl=mi>tD?trO7oC|oT4S)Is#B{sqN??HNJHB4hb6l3Us=5EIueVpZX!1d(z()|B
z7kC`7s{DIUFp=O*?B)A9B$J3v^gc&oyvy&!J!jBG-Qm>pJFyS4Wsk1Ql_=gIJ{9=h
zaYg;AuJ#!6`s7&2Bc3QE&pR*LKi_Q)pEAeDOuq|0tkm;p8}bo~ws=xre&j?)b@vFS
zY{SFTTDl==r?jVrEK~lLwF@*k3lG?&)FxRH2Lwb!WT->TYwz(S^RzqDJD#16(3zZJ
zvXWE^emuqqH$}2C_uWE8=Gt+gRLO(0d93C9Sz=f}GnW>ui*4|S?Rxp<a;dq#Q)49|
zdTH318AK{0u^W);dFxp=0d`hP+QfWalDJA!^^A>)t=J>-#crurbQ}z?BURO+NP|j?
zPuL>fhN^Omt!>8oT5ad?k>GY&=FFR;IN&~ybg-<`=N(EoLI6cNWgoBN(QCNcFK~K-
z??^E!#3Li9*t>d~AHkj0$&U}E_7&in8&`#yP=i?&F#h3qV}y0})^$Q<3+B!Kz%?1O
z7in5AO#j;IeRtmRvHN7Wa{*Hcp2k->w=Le^r6k{!;rG9rbx>)hvhFkGj(}6QME=h2
z)`gwPb=Q2I5BC>?hlB1n0XGzW-w@6hGGY)8K?nAqa40EE=*5K*a{^C|jNW+$=}jW1
zsLD;WMaN?3O~)S(v;LB!wZ~0hjP(9TAf5;4ct^oX>!jNHTUSY}^e%16($A~~)0dG-
z`z>UBe82gO-+m1j;E176mms2$P1m%y?|J*~<a}|b&_-$zb2DD*a>3!aR^k=IYtkG`
z!qdlWC_QW3?<0KiH_cL2Joz1Fmxc6kLnN+E=X?oY!$*e8=`^SpAi?}O!s{+2?K^Lt
zr74V_PXFIXqS0fV&Nv-r=_I|i#1YH2Bl?YK4sG&;f#rTnm#FBf?6^B)n{F@#zH35V
zZQ3Jdxk@xXwwjQTUYV;X7Fn4?h;tZOn4`}TpU5Qd-V~gy{*l~FuS6S3p*}s1Pe0F(
z?Zdd>w0DD0oA_D&LwT9-g`K58H!v$-=W-;Jx*X4i?Yq8?n2gEw>bWyuRr~n~WwZFl
z7}HLEmMF5Tj|h_WRNm`j(K2)p*F<uka!C5_rq)$62oLxQ1U&IcgQ#;aFWRA6^IDd3
zEcMB5HzPs@39gZZ>wDgPG4xPsLX|FCZ;<R|0B8d(hj*q@-HiMsf*Xf9SQ;QEbIz3N
zpAdIE=RWH&_o9mceL*}b>xrQke}DP!EG;9$S1Up%TQM-M{XKcziIf7%ujWemtzb;K
zs_DE(Wdp(7_;QMhifH5gE8xQm_Erz*?~MQax^oOi;Qn{a(ej0z#G#e7<1+v)BxaK4
zFY`&u(ONn{JD}Q9uXqGWU{cu8#1S?3MHzW8H7!;@WxzS)FyPdQZ@6pO?SZbahTw?u
z0j~fsFIWLp)6b{i|KkZq&nVQ!42j+lR6FOjS+OtG{ho5JV?kljIA(!f*7XRi=HhQ6
zNT5A*c7(|fi0@6PiCA90_EaYR(PKrfkTYYi`lPBy<rCOZ41N9j)vQr(O$=R6!!sag
z2_Au`NAMX>@2K;+%oWP`s%&fANW#6UYn@X99ix)u>M{!SxsS1y%<{NC8deN-H1Xo9
zggF-=O-$`gty$l!nb0(M9Bv`%AEgnLX?{HRi&{G5rgA|Jv6Cwl%<h=2iJ5o>|KXIR
zIOc53Q>2|SG#;d<3GL|aR`qJ-%dmVsC%`ISK&(3zaZWt-Zj6_VHIW(jS`-qOHN97t
zLa>SMbUc0m5xY){sx(XGC@svniG`|6{A96<MOdGv-yDtAwv;_f=`Ii(uL4UAr|}jR
zYesH^-CYd{Rj~}&*SCk8Y>)~H)5s`gIo;2gCA79d7oisg${G9_eO5*nt$wBcrBpX>
zzkd2*cb+W&*hi|pJJ-_P-93W)9aCPB4!2kaAyoI%2|C$p=R@X#-A-y%(@$!>?L^W^
zn*f%AZ%)>>8Zb#{LPK}Xw!YQ`F<z`Gp%>WJe9IfaH%VAJ9*`6VugQ`-6Xq^+otWif
z>Xjh#|JpYhri9Nv<BtoS0$Bd(<=<&F4UGwF#PTwy>f6qC;5moTUR)b00^q}WMb$HU
zXiq8gAoO3#8zXm`D|b8x9<WWkKZz~0`ENY~ic6G+?VZ^NUH{dOaMEy%>jdbb5uxip
zU~-qdB}!poYr^>_km`d!T0i05a^amrM%1Y91*C7ZyqqQR%HBMA%AKLTH8YSp0Lp;Y
zvD@(7N%(sCx9Sq(2hFGlR%xm8SCcTX`)Go=hy2Q9<?1%T->&TrTS-AN4)S*i7zVC_
znOxC~J)cU9-$@ta#Lh9}LnQ1Sv|JNx^w7>Kp(gzY+t5v+g>OSdX9D5bRAe`t7UNPM
z?Y0RpZO9i#%E~^~QJxWLvQG(llzy#_<5h8-x|NDPw`vb5aGELwLJR|H#s&79=tdp|
z1=acT$Fk`Vu|D(=n~avfJ0wWJ+x2YTNfS!BOAHO~)mvkx$?NNK!y}(+tH<tZC)7`g
z+-`PH=bsP5K6hl4w8$&6-*U#0v@oIAPx?~M@|Zg@i+Z)+N;2`!6QZJK)3<}&)GK=e
zgJyZizCIeGroiOQ>4aO9S`^8q^@EXrHn}v_p6Q9aT9CO4noYHt)HH;<K$tF{U~1Lm
z#mxL?19iiN0z<BxE865_Bza}DHeJ1+U0oGeG2Z3;Z8_oOP>jBmZ^ot?;=%qZ5n~30
zRFX3BKfeMU6&4BD&N_E&0j_x9VF>O8!omswb_ehALko68SC<B)v0h#Vq@4<+d0e;<
zVF0C<c<r}s5fJHAv4!s}v}Zx%s~<>$HJ)N|d;BwS13_S7^OZMzYf&nM``zvnI(#f%
zXp}4Xls#@&ZmOO}Fkz!HM=kOY=;Vf$)l<!_1ZmPlag*+^c|-(5!K$IA#xEpPCvokf
zM*ED}>BQoWH_d<79Qf+oThCAQm~Q#r11A-@zK4Wdhvr0#ukFMXK%$fdRB`f<1*CX?
zLOV;LD|PTWi6|{K^!+pCb2{|&SJFd`_>4&OfVpwb8L#NAGcF`5*M?k*FPsuW^ZlS(
z@pA0Udv^AF?{WL3FS->$3rxYp;dJ5h;_3Ixf5$+thHio}^;s*xK+ABBN`Uh1D%ze5
zAOLh80vhWA8~uO3p8r$)B!HH%Re6rsiNf&X6m*;NHG0cHqFdd5?k5h8htSO%=U&M=
zn_KnRc$hVa*6~5Xr~piF@T~uV7IQ)y7rP$Hi!b*N_6G_L^rDRyz<94N>%tyh16tQM
z{A(nA2JL)y?J`Kj1^3l0ju-x?SKO}$Mut=pMyH1uqe|=idUIEK^2;%!`YoI%$UOL|
z`K6^b$@|PFXhg~vl<TI<0XjA=&)(4gHbR(@8L}i+PCbw?gpzj=!bjoR`+PHrx*S??
zp5nX=YCWDG+7RBXaS#`<XQ=D9AIvND&m}&UQj`j_q$w^8C@mrH>r+{Ev2NC=F(}k3
z&R(ajT#Dl+S1y`}pf1I2Ho*|+aZWRCe9%7I!OtsrDv`O#L02i-iNJTA%$F1oOCCy~
zFBy|Uj2O<CEnTR8;1Z}5aSdwQA?*<`q0{M487C)rZbHdUx~7w0ekzY|MCn=n^R-g)
zGaNHB?WK_X77lC-n0_~Bu32CQD`}I%H<mTN`G<}6o!6fYcCP5%Ut|^uH8nLkwdDhA
zYh2poXd~68TVxSBK*L}stO;Lnz6V!)Crs7R-Gt|AMgkKb=7YK!`$aY>B=3;~5g1g|
z+a@EksUngcW^1=fHpL2UfF@lsy$SfHrIi&<pg)pF0l4)C5(BNIEiJxhM|)dV(%umO
zRE!392d#C(mZ<G+rnVqkDYB^SB1<$+D;xk%s!!ZSZ#&Z$&Mnc{r0U^ASMLnJf2R^F
z+?x6m77ZS})~H_Se<|RtqZN_wmlGcN=d1q%hdVwPlr<sk0bD17)6fSF#1uBm4y6y^
zuwn5gM|4_Fzh2_A-J*EQ`)<$02+n=9N)3)imoNGeMMu5SH_-+Op7Wc=oI5`UXakoo
zlzq+_PFnUQV$o7IU>Kmi+0g3}zbkuTr_N%C5<Q!R$sbVX&ZK|~a%)c_7Puc_XxP$K
zOucHFHGRZ+QvI&e;keS!=DqXXE<k<Ty1>lF8SM`SW_*DE`klV<`w3EKzqnshaz^c6
z%cb~f$aJqcNq6#Xpdlwu^fm-MOU82xKvKo3X)`Dzh7KkfsXs*pE<|fnnkLO)tLuge
zVyO&3Kz(y7LtVXCv^#_E+Fel{mBHtsnFx7d6oFwMdnBt)O0)$^uTX&_gE>mJlj-9a
z0;yoqR*+q+A8Dt^m9=YlhzzAJrW5GUr^@QGuxibMlHZ{SJ1EYx0&6Bt6d8VTr3@*r
zslV0`QfNoc**lP|n#!H7Qy?q#ge}U)U>iD%6w|gSRg8%fw)vw_+i`ak^v~p7!^`Hf
z1vsqbYADeteqOT4hFyB9qIr&ox6LS>gBuq#^X6I7N6CVjo!(N5pE>!H5|-U(T{?04
z*y+Qla7qxlLNHPjt=I98VV3-X9|A9Wv-$AM7d(p^97@L7#+c$o3v{?8Mibcu0|aq~
zXakHDv?s72I6kzrl(d92YXN!{Ac64Xzgbh#Pf_4|18#3<bQ?HGY(&5j0XKhXu$cmy
z5Ggk;G{YiNCe*d2J-D!YE?#9pp91*l5L6C6cVt8;{u-NN(Bor8#l#GPF@!e7^lf6^
zj+}4pe(gzq#QnN(WhB~`AR(u5u(n7q>h@3qN(qDgNem|+LVXGCLW%Y}LzkOvXTQ6S
zeZN#a9YuSt&1~aDEdD`Ho}fXk!kMqLt7=UFC)5md(h67P$x!?G5WB-E7#ITW2zD{r
z{<RzQluG0ZXkrpo>zcyE5ZN=ZZ<+%&{)5kTd0dMiC^d&4DhXgMk2W7|0x(h^;2zSJ
zEVKvd%#1z<Cua@sYnc1OPtss)%XZ0Whj#jd=afu%z27o+fyO=$y;^Uqn;QA-^}GcK
zel++BGZnNI1Q^uuUQed-euK96TRxoBTt$}~Fj0Y23^br5zkk-<dkbF<eGdV3>9MgV
zyA1IQ*mus$@k+ID8nBN7m0k*p7_@VNHBS2`T0ot-3Vtssh9F6^oUSx-)}T0W<a@Ie
zd8_irgQtH<g@lB_<CDQ3h5|;k3wQeR=KFiP+hi~A%hM_8#U)iG1<%ExlIkQDifYE#
zBJ1+;Z%!&FaL$j}G*E(LwP@skMcp#qb=p5-sFbhvD}wjbh5#z(mlO)xY+6IfIOND_
ziQ(AII*Gpgu8=g!@3%`#X7^?HvEMPzzYDpi+4D9DwGUm-MVe~ZwB@vwDGrGH)=~m8
zMokJH>cKW}3wF+tRZ}x_C+u@6CRgNMw;=!6!ybMOp}?-?6V4PRo6E#RI%3IX9jZYd
zY8@)WkySKlz1Yx_I)nd+PflB%B-jn1+Y=$HHh=#yS*&bU`1kCrD;(PUEm2|IWmn$^
z>sfQ{nZ=q&@?e7RG)&%C2n}3$;H=emBuiANc(eqZ0pP5x&|>C1OuY$oC}2SW)CIFF
zN=ix)*6G;0yn2;6Frd=g+xxZItqc)HnXjdw4HE(r_@?(SUbT;c2@&46?S}J9(;6-Z
z&tQs?uV1M_J`NQz{2A;nt?vK;=w$fTpjse0ZPRUl9i)2j_oq?gB(4E`u>Bm4eb`6N
zcX<$9upzUafQ^?nxr(&xeU99=ox}n26qVn87|JR4<+2XmuI|E!^l=QHKql3lB;Y(G
zy`DAQ=9zYzX?D8>HedNZPqXnBMNRIwfwSrXl+w>|BCAuw9Z?(*h-KWl9kG-O>Pj`^
z=|zpNz%5)L>+{a!Y{jA?TWtqhPCz3(fhBh+-g@-!!;pxI&iO~H)7jp>?(S~VU@ZyF
zX*c~=FX87tOE=$0Ja8JmcDDw!L*NbTJ#gYxY$iFHr^WRGSbV%R4;Jr5ZbmPqLM?M|
zG-5hf<>VTflKP~1SC1)QYB4WBnqd8V1zpL#he$kdvz}#~uiGEB54}p^<xO!Yn963?
zCYBT(*_Emn6tF7TK6b@|LYr=&o+ZX?5!+xpm`GBGhWSuVtd;N7Ff%DbR$CWea*vig
zTV}8pr^~-2Ypxgd_${n%^4As!U8~hQ&-0v@ML91=o;pn?PmdGOE>9=#8qk;PnCOwJ
z4QGxZ#nuU<C~8k7%x0WqWf2Hd1)009u0~z4*VK^s3VlemxvxDm_z6x7!9dxzFvskp
zc}C5VxkKV5)wcSR5bOPgRB1f%az!Q4%PmeFS5$^b=1z`{nEovV>0Q0>zw+IfulzT?
zt-0g^LUWSm<031=-!RBsInrj=g8~AkNDxV(KpmPe$(DfCF-rRP4)j_uIZfn>FF$F6
z6&dxpf;0jJAQCV_?*+l-0zKS(3)O4``fjx7OVJAkHbgftE(bb0F`y&`Y$+Yak17mr
zEcA8WB9c{2W*a`o!sN?@8h8YMFaD7-d|6ypy?>BmG49ytWc1kmBBEQ}gFZ5$sL%$;
zdaxX!aF2J!mn|AXYxhC0=R7RX$1f=f2ti_C-nRSA(F3-t{|1`JF1XqCiFj)ycP00Z
zd+;W&#d<$${3%C&@M6_7`)GV|nsH;UeftC$ga2nI{M9VB$IS@MMnH!LX2No=h9gZW
zzva*B`3Jw*I-la;t$3+}F!9wq$)lXi%dXa=mfs2xLFcn&#<;XmR)w|*Dg+re)Xe)4
znm!o1x}IqMUsJ`<o^{_7meC&~u7XaOwz(3OS4`;Q$CaoD92B+536Bf|@2iu>aj{tx
znRh0SIeH;b5w2P#da~LSbXFgkV%A*6Ch*(7mp|l2Dr+LEd9Ftp<8>iQqGDd`e!OaF
zMd-9~ddF&~vh{Nw&4it^X#I7qP*Uc|K;3>b3qkAV7@H@D&GfXE(Xu&F5h_huIa5zQ
zs#;ik;}xsQve%O1#V-H#on}oQi~|H~WgjuCp35KqYQyatzLzA{JVVZ^$yZ7*;`o+r
zCfkajbdAZEH{Sg<Vlmw>L2%D!|LzeDP+FiT4wf^zXNp9U-hO$V;=0fT_1aG}9+AJr
z7PP;9{eqc%J9+o(T;<3YxXv$I=2LLSOoJ@;GUK?>?lCb-;91d=Z!qn)?ka{UbHCoe
zVxi49;VHW*9AIc~H}q(YUfjJJO4*~Gg*)m+*H=}y27@!y>H0{8C*3mm@zK?(4);I>
zN#d~SVb1LK#I`#Sn1Jz#bI>WL#WjhSQD!gl0BQt4mYt(=l@myMtbjQDnVd|8{#=pK
zzDfQ)UF>6W-0Nue9okwLJ)Q$i{gVd!pI|J^R^>gvnfY#D%30#Q7uxsrKYoBFcx(dj
zvw!-5ZG{mYbI5E(Ky|~7U0pu#1&8@IwffXOuc-z8_yyTnhw8=ydQ<j%>9xA*YJcOY
zab&bB&q()I%oH2v*-uZk`ze@WY#L`j(+YE^mfO&xa!LL&!&2UeP@3Fu3&SuJKv`xN
z^8`7@p1qv!yV?F%qs^)3acZlBRRRtVZqANbzQeF9t*&NVx$ey`<2DV&jlTw{R1X6Q
zBXe}PW1jV5<@^#5@{XfOo^mC>F{CPhqLCY2D+VO_9oDqF4dPc~%t!uS7bC0+LU%r;
z4ZoKUn~p5s${lpuP7qk9Q(hY_kh1<6Dpgo&mItPCBkp*dhdANk;p6w+NlAkkq>rX@
zlx`znOblHh5kSic;`TB}6gZuWDa+lQTzfp^*17=P>6G;IH8WwSeBSA?x-P~+iJ=^x
zhQ^^A8~_{^ci&=)G0WUfDyp`d=5ycv#!m5-%FX~Mq2&kCqHczr_VH`+n^aj(-)-Ps
z+=9_XvMSI#XK1+JwFct{uX@65sS<ZA5o=Hp>M#rKcpQZ7F)~`gy$F+!7XUY_zWda$
z&CW4_;r2waI%R2(D`36(6Vx@v@V*1tU+3|Q7#s7yk;{f4K3@cw{WZQBqv+&MVi|;Y
zT7&b(|1F9=QoNfB(<y4Fejs(pLnf<b)i0uOgIG!4soT(jP<^u6L419V$Or4wm4$=E
z-{M^#H{Gd4uer)`CaPjeuiC1F`L-FFUp&o_Z<+6)w-pXGQT7T~Eh&x6rG=7V`3&mm
zwf7Ofb$LS4YPFCQleJ@}8ERsE9oNqIAKCE+JJ<^+Cv70J6k9cl<`rokf@~LqJ4#PM
zn~ANn%Zm7D?MVDkOfsD8`7>SZ8bNLZvh;GtNXsrSL_mX_RGx%;{^JCJ4oczf3!GeH
zijXlp>+uevgi5t6YF|B~&!OQ{OEPVqpRrT;D=8SfO@C7PFvnM8M})%uA;Zc!V2jIj
zY<vL38}#(Z9+%Z(uK;dNAlQIACQ}nnE<&FtIW;r0_do&)UrhjyS5&kWk>uBhYJn)N
zg9Wnv{`nN9GB6z-&?HvLVcNz^!qW>xfWWRKi`~Nv4-fBw7-mR8j0yGd=qQweay-c_
z|3%QIp->5hW4ZQRpSWFngFkfDKDls~_|jq#P%33O<8m*hFY<mZyi5Y=ee7Y97cIrA
zQKYv#az;Pyko9qp%?Q(NY{@)y10k3An&9JfxoKCmuI`41uwx-kGt$J%5;fBVct1v%
z^a?-1$jZ_&RHI6+eRk#j!9Ud2hW6qk@ja2b_AH+oWjm9rBmDgM8QGH$NN=mN60*wn
zbqrf2CZrdVgpjnYnv_t6W~A~QwMtz$QdJb2l~HnVq`9UThz4znFGulqGsou=ll&xG
zJn^N)bHkeFYun7dDP}v!BCb=odN1nLPsMlZ95JZ1Aa1UAH;j;OU!AJfivwQYykT8z
zU8+hP5}`}j&iq6{p6zHxpm)jt*@cX>K#bYOpJuh(;Py<ek=vEW$7jB@(OA)na?WMY
zeTK-&xUb;oW7OOvbr*wQ?7s!F6S&o(!^`?zu1#<?gA4n*E1SS$x{&|p0x%o#(&l;#
zM6$qU{4e2-Es%p)q{EkVdq}ohQx7vjT9!*Nk_ARaK^^yv->ePN+`P$(&daOey$+*C
zFqDUhIy4_Dg-`YLvd69a`P7}}13z%Rk{+s|Lx1*G8C`zUU26?y47wR=PwUBiksVD&
z$falD+H*gG8O9e}`I9?8PSH4%W1~8C$g+#4{gCOxKVOyWisPhgHt!$ZLEJ<(Ix=0H
z7dbuKkpg0i9-okiluX_u&cvbkPx&-7j7mZThe^>XswCI1%Cr+X$<u{+Xqe(}doxVd
ztGq4Ntkh!qlC51PwDxCAK)HwVR4qWZBp5Ye^A`^hc-_D$-TiZ@XdygA3m+glszv&e
zH7F|H<$H7Lbv-pD{}zeUOatB}vslcS^WK!kT{~>?2+$nlCwtl>+N)I`{Dy-ELjU&0
z-n>XPRM;(IMKp_cpt7>wWO%_07LS&>HM<n-8Gy0>&R@?yr0)Gtj}x7xr6n8AJVL#B
zzEVCM2x#FuXL>2@x(gvp#Srd%=Bxk9g)R&Va9<KCmyh!IwDi7V$}qKdcK<o@l8kVV
zI5Scsq+t64#Y^LCGCfVM57&G1*CYZmaic5FP)Bhi!8VGgg|ugs5+{Y2mO^FTnYR!9
zk+?_P<-V3%b{*NSD806b#n+r`R*aQRjBC!Jfv8T<kO_HI6jW*B>TR?UP5YDh{w_wW
z@7Zzuh<RE@R?num#LCv>8ZWOF{Y|rr=K8}Mgr$W#mB?MgyEYh;>WV{o<#bO6kY1ee
zi5)wxIDZKp>w?k-?u+O%63uxtuIxFLnwbp<Tj+dVUG1N}K4FPsI?``-mdtqcx^lAP
zU<2fPHaClA`&=Y_FbEl+4uNIPbHCU}bq|{6bBM#_r$`JcI6_UvwvYE@rAbKX2gRvT
zF{NC<2xs85K~gJ$xoKoz<ywSdN-o=c+Wq40ZYNoPMoZC;S`14O_p7fI78Su*jtHjI
zDOVLZGgm!H2S}SGVM+=jkFu%U$#vy%o%rmLPoD&-@j)d5wPr~f>>HSp!Z-+wD?nrf
zP04sc$1@lXLAk&k#*{$L9Lb}mtzgB>v^fR9Pe-*na<@Ug?A-K0cn@=ubIW`suRu&}
zEDQ<(?bIQ9Ft-52U0Yk+J}0(F_HMkonQ}Ejg#n$UJ5&dgDP-j<dNJ_f)C>+D4!JgX
zS`@GOoGo19!x$^*(_`qaseX_z8isr3;%ep&Y3S87B$oPAjoNE_j@PBdya=jPR2q=I
zb-up~jj)nPcIA<?DgO7gYI??d!Ljy~Zu71KjGF@aRln>+-z2(WBv+4*9O^E)L+mAW
zVGv|oA+vVIg;H!!4`&H8_P3c`YnEwQFzrq6J(fV;XqLHqm5oPNut%b8l`QnCP#B@J
zq1Iw*5J_vA=bXmf5=w4?`I~+r7O7tgg$>j5g7YY^%d>Im60^_y)8e8^bGdo6-!eU-
z#r;x!rk#XuW_m~*sTet~mI)u-Q`{+6KJMF6gOy*Uu{%g{`T*H*QJXY}2*HVc_BTTV
zwV9jEwND613BQ1~F8k955F5%Mp$tPsEiUCljz~yH#sV@m+9vhK$E%+I0Nnu?5ikp#
z5;lq%pbtS82>a-|oaAU?L|mZ_+%peaOF?z9>Iq~D0RaK#sZ}LKMQVJkD+$sZ#Qd`j
zFv?7x_Xi%F`xy@60oTW4FO_)%y-lDffP;v60{h_Gz?barBIQll2jf^|4AHL){nl_I
zh0k9~t_jk5;Ri~U=8hk<ln+;Mf`O;Fw;uKIk_TahV|u_OCx;Ws8i7KGu5Q7UjS8+>
z`y}TJiI{(z+s`(wrMgRH>*fn*tm8&jxqsp1SL6>0xvdf7)r%F>;BNj~<obQ~GgiRa
zX3h(*sLkJKm`sTzU^V_Wh03ZRo!E;jZW*D4#d67i{NGDwKcf`=oSDCKRPu=I_TbwL
z*28Pfc<OKeV)W?RkW;xzUS+{pD4Jc*zS3YJv^g@q^^UBBOPw)SljIe3o&t|5xfzcb
zmVxOZoqTakU=Ga*W^oRSMHTY;;T4T?M@xH!Wd6vpDfyLuT16dIc9UC<!)z+wc51LT
zV|%W^=mMRE0h1CK_QFI9kR|)sLXi4x7x#2}m6lEb7iFLvAZO1^q;ytVT8=>ljJOJ1
z1F5N5?!l4g*mx`PrirYil+-6cDj=;L>N_CmI??G5l9S>>h7cpv<?$J>1CQD1)<Vu9
zyfo3=C%_A=J766Ee<T^VA0R#u5K@Q2L_4x4msbY2&RFK0icUcL8+k<5$%prj9kZku
zenblL#=Ok`dqLIE`MuJ9xtex!_*rqGyiroypogWfLQ7{X(Yc?zw(`RI_+xT}uH0i`
zwUMOFC!^~*Ts)zbL5$iG&ezhc7y>3i8N3R(+e{iw<ci>_Pt3ave}W3jM2#|)96gpJ
zx!$|wX2-FPin_wchkYe?cR72j_wOz8v*F7Q0=4bQZA+^{Wve=VdxM+YF}2TaiBOf=
zT*+MdkzY$g!eSlQT+?sYmJV#V;vlk2OXRDj``qYJ6&y26*|Xbl7+SwWahU`=Qs?8p
zZj=5PEsKx9U}q^vO>GWf!~{WL$p&?E%;srvjch(XNx5DQa#%>vm?MgXT%Q)1+MmLu
zXERvBD_;1U^RMJsd0&*sPBY)#1)l@b+Q!}GC-#K5)l@@X{=60)^SUQjO#<3&F#i^f
z0jn@Pe{%LezFFv2o@^PAlr+qHF<oHTyu8;L-V9{#yLXk!rV=7d06v`tEghRJTmJ@-
zNCw%;4eC{Ci#62M$E^{-aGg-ncBKZhPoCri<LKOSPSV@jd=xA9O{!Q217wvRoa4uD
z0sjomdj=CcfQM&2dWAAO>N5YJS=(S&8LOQ1;$Dgw`tvU~M=w&A`o`SCd{5x)M*_+7
z{-T|i>mKyH!aXRT+Fv+XGNUia{y5*!cOlJflqF^}x~ODGl-geRXqAG;WKvF=xILSz
z#jDl7GDdNI`he!}46{v>_(XJ`L4h^ah==&P6v{@HT&Q>oCsdbb!D*N7hEQ^@?(?zA
zyMU)hC0n!!C{@i~HJxzBk`o>q+m(HD3L2SFg+{gJ3__V7c{Y$MraBX@*fQ|ZjRv32
z%aKt*7nn&8$N7`<H>H@mU(3RnM$Tw@ppdlU6Tima8(j+WFNIV;Z|kzJiy<i}bWKw2
zOQNiBkq?^|$_$I8A6>hztLxRu0L1d{1_v7r-sWGJUQ5k<2L}fyZFf{wKm$2avcA5~
z-sT5L-RLUN!up)(AiYlBAI2R96@#?~FnX}Hg&FL$3!2f6j?n>f@x(fWevnPOp>dXf
zet}E6RnrBy-r$!sZcW@L0K|J>@xpwm1c?iW*t2!OfRcRtyUWH7B%nluhTby-h}{_d
zwWd&-C@0;|z~p|qTV4xSouWbl2kJTp2S@(#o48D3lqp&}1!*>U%R#sQt4?_PvNRJ`
z6c}EL_<j0iee34b?P4zc^Wdy^IUb_7>g;U1+z<%&spp(gl1OAA*ZrG3QtPD})M5Ft
zhCxe&qVN0|du)?^a>6rI6Ngp@zKWo7iLc0zvlrYQEt>tET}-=Fz_UVNm-ZH=Q0E{!
z%}ysO`iG-r99NtCXA@A*>^a=89WAuq)SX^$Y{8$C3N?Z71ruVyyUOK1Y4KbRTO@Nu
zdp}u5EQWV{l$$01UV{mBt{v_87h%NC-=_$5Rvpthu8&`Iz&VA1<9mn+9eaJdXfF<U
zU!H>!ypC~T|7>vQpye!Dl10e5Y`C<%)ed|)UOt>p3QgzIM$vSHKE?O;k6$=nEb<ii
z(f*N`-(TiqXaAR=A2*_T&F;@!dqO&s&sD1+GGMGfUtguK-_+YSckw-ok;Si3s3lz|
zajWwC@2j}_Gs!e0o~&Z8I7g3tMswz^NeXa${Ulm4Z}2er^C32p2i5Auv@VH<1-~mq
znx%T6de?tph)*(q=+n;Gd)0pB_gqsVXC5PAE-*Aat#bs&xdXE??j@IenBH7z(&YL4
zFeD{qXvN3=i({k9%B`VV*7c1B2bc2`r-sHxm%w_kQB7p0*d7A`CU^vI`Xr4Tofa5W
zJjygHc%_)ERn?#J%@ny|NFAzH^bn4dr9IKdeMxs7I4G1>J-!)eo2L~LRjNMB6}hh*
zJ~cIUz7u8_0U(kG0LC=2eINFEZ)SbJ%Pc5%Ca#1VgXiO#y(tzTt>9;$fx<{*cNK3@
zaGz8<5svS^x?_hk$NAR0^}An$%;%N5?V>mVb8|NRD6{wF$S(rIi9Lg~(%;VsWj20(
z(=fi8SU?jxrhh(vhj<`q?0CleodW+KQSUvLp<B07QHq1ZdE~t$HVO1<&oJgh6{oMe
z*F{+BW6N<K=uB{NpuE1SWbmq|MfK$ysH#OqnY#a$&YmbH3uFFtPA7}aG0Ch=N9F|+
zsNr+u3MoYy{)~OuXu8JuAN_F8ibq9vgyt6fLVsMID)G*DYqFHX5>j`~F~N>TTIyN%
z2St@DR>YWzA2zoLv1<-b?6D`{N8D2TME_K7l7NL2akasOK3s`1kw4pf=JPgbj3aU3
z&UWjvFQ;DOKrqNgm(8k}u=l-ar?!i13>dDIM#}jvj4lzAuj`oPsU(2O>3%77a(`u;
zF}M8Srf|lm2C6S#zIcF%xQj*qv9J5fm-pPas8cKi7-C(V>zbO<ruQI5+!`^6S5B~0
zG4!szU9)PstG4`u%L4&@h2IHmCeNSmC{JgrT2(zNJ9W1pSMUrDTJ>CmE2#O3XL4Mm
zyH@0npT$^ilF8;UU8cHYybM_+9+}7M4{HPmrrUiY#{X@SV|-VldU~>D*!FB+bazgO
zGtJe{*zhElhs|f{mK8G}_O7?2mAc7({{mBObGi7c==U?;n#8#H1aq9qpSgBoqvbwn
z>GQYjH5q@T62`1B;me@dqr{EMEh)yv5i%&YLvP71k79F#>^fBl&yguMq+!hZD8URh
z>uUdP(X|Ez%Dz-~s<eEzheT<XOql;XWcmLTj|$V=ZlHN-X(sC@*l{y@Kybi)K<4hq
zjKe)j^*Z?nE~z&Qlc?<T*#X23)(+0uOd4eGWMUG381P4AYCugL6)Y7%nPU;xukTu%
zv#wXDf-1nqnLTR>){iQo$=46$@m9p;NqKFPs+LGCH_1KAmy%<)fu(}VUF0{MG_BED
z?WjJ>d_~VM70&6$PrVDY>a2yo6$^_H=)Tqr7TzU?ZWM1z)OI43OOjqHGzT(06Q!%I
zsgVKT#>C{OGGB7DX2GixhYriSfsG9wxWijpTi+((hsk~0cFzpYw4eb3;L>=WQdFNA
zSPXl6dpC6$d3H}AMiMtC5wLxs@O2K%?@#jLRqb##dINu2g6zTdw9-D<*)C-N{{9u4
zvQ3VTCJ1EeA4vSh^fo;m{Pb4RLpIu6jYdZfL0;e1_6d@T_DZ5$^{hv0#xiex55Zg3
zGfch1*V(sBm5<SmFEF(>_b0yn53X@*SJQ}}45}b$I10%@jfyDHe`{l8Y3IXu=MX!W
z_`zKgjA5qx8gBWR%m~4>pXwYEUp@Lu{<z^^HF?W4ndDd$G=`X9>lj1W=RD+o%s<E;
zc){1a8K*$9=c-B?+|R}wLFRQq_(vzP{vl_R<e%ctcJD3T@FvEo45ldxPt`bted%P;
zv}cbxW5u8BsHGQ7Gno-VRl@Z>SrLZZtk)m;<g%(#HLoDoMV!DDv7l3hpIEzXy2)X@
zCY1UK6K~t!!$Ln;K6@ao)AzLVo{rKpGo9KDiQ>;=1o*xpsKdM~!P%x_H}ZP=Bv>TW
zYJseX9<t_*hd+43w}RQ-u<b)vJiN2CmkNM^cO!DFh&FbnVZ=k?lTiJqOi8dvR3fKM
z?q<Ne3IU~<R{^kr#8t?b5Gwjy_!Lr*W6VOkOf*!Dy|;zI=&P=du5Un^=V9~Us7{_@
z7ytfvkj^0BJZ(EdX=<QRhmjv#WFwYBjGm?Cw~}KE$eRbuWsBhdUG6g^7-I%eg(rJU
z_kLd6$Jjo<Yz!jV*xwffXx@8zs}kcnE;-k(idOCY!i3+_8z)5~{C7UwZ+VI3y@Mmj
zq9er4+&e~)X>?&^yl^2f=@HC0Dt1x(__>bX5ev#Bdz042eko~Ae^WWZ-y#7Q`md&e
zd3XB>HNmZ8eqOvXV|E5jrXk{H8A?w1E&=U2LRm=BP-+k|Y?O7$i5|nmvnvy=l$)tO
z9d^hnMG&Jjl5v-Qx&F#ZAjW*f_dzs9j<iqtUixPwkM1H@=ifn#Tt*{>qu!6hnP#^?
zI|&D=ni3b)WeF>@@Q&o0ihF3^=U}IeW3fTlN}Q%sH91{Ml??L>(@*j{sl8@X7YNZv
zIKf8H`C4B0iap;5V%Cc^2M`$yi>(o~fH^PA%ajO4vRUXWwl8qvcYTmjQcVj>`BkxS
zLvgp`*~f{A2@+F?!|ITcDfssJ)|;VI_e>T$m4xNiF#Qw(+GuN$bZlnlYps#~qHo6>
zQI^KM4*HKA6(7rMpLQYx4$TEkMCpEU_>CT{xr#OYd|ILoM4r3nW$H@0t9gsmV2wUH
zI%zMm=PT`39{H@qBh0pb@P#P&Oh^&&+}6szAApgCOkigUTECs~xJugmbm)n_(R<Yp
zv0JlmaNM#s%mZr6nY)L#_=2CIPCn;_bK&#Tr9ay+(L37u!1IR4ee=FK>1ev(#<5Bw
zb(%i5`VUd|+E;;=)T!)m*zn`Eqs9n2JN!Mqi7~MxBp3xyPav41bugc$iQxRWdn?K7
zYG71w8awo~0t^H*Up@S7QtK2Nz33y@RLFXl1fmtmm=W*(K<v@L?&Yhi@wlq1UyA>A
zJYT$cD$SHg(_ll=&8$spW>a^&(}Z6keH$xxAnA_Jk&ylN4jH>ZbLGPZdes>3XBHDy
zBt-pb8%3LQjPBoEnp^f{+ONRI0_M>#pjr-5<m`F)bo&+bhGwop*6QRs`Lx@Cs|O(W
zym)<1`<ZS0_%1}q6bsL}K%I7SZm4{~zKyfC%=YG=3)`~MV_&rR!DVHTLgOt66ITM}
zK04}Ct~k>$%U1@LP&HZTH=%?XVoM&CLF&E4j;drR<lkr%-Mb2rrubo9KIX-}v85tP
zQ_Jw8RyR<zl01Q>g6}?G6z_2V@9QBHen1_VL8WsvB(8TLJ}2&Ztu6yt<9dFFh_3o4
z`LV9=oNhW)#<YnjQTEAahOMM1C#H@OS=QY6JKVc;qoZ7f9iIdPVU%Z321(xar=Iw;
zvbX`d6~xh;{c0>i%ICyRW1%`#BJVvf2o!@D`ZThz*qrEouJM&DU6In~7)M-GY<C73
zFfOEfq9mWI5@SNGTZ`zabzpm@PdGD3hcq<5()Vwuig=Tl_p?AF2NRR3>HTLk95SS2
zw}P7E-Z#jVXj>vOSP%hr-{7iu7)@AAL|ANi7nf$#T3qUsQTHx+?1yzRqlYV=P|2a4
z&aKjkZ<x!*vrL^9B^CaB5l$S)gzapa!iM8Sx2ig6LRm%5U$1+Yju57MUmG%tn#-zm
zyGIb}p<T>pXJ?+{bviLADJjl>Hn(LmOR2QoC`$LNnhZH2b3d5VW>2m!vww~pE3k>&
zY(~&VOPJa5Q~x-Vx^?T8r{na`v)S3%C3M7LVGgdE?nZLurn!Df`Vk&Emsa-2q8QtF
zjqCrAr1n=jvmY{6xBip-MkU;>A9rlOyQ+K`g-{ttARpns60oDb{?dD`+3AH8;a6aD
z)!OHZ=6^>i@JHQ{8~)bOnT*-xrgHAjs}c}}z#S)ht>SmTv1PP+q2QbS&qss#Dv^8$
zTg&JY*1g5>ZX&Ma!LLVZ3T4Q$C?!_F>*pPB7~KueOXVWmD}2av<00RyQ(TI|X8Vc5
zw+az`7xxs~zS*_ex|mMoLThZpp<X-*!8B&Je8Hv9>r04D;lNK)(K^?Ms;#Fsk{+<r
z&02`k688vbqx>cMS-qc(9fzq_lM4*gMxnDAA(=1;oe*Ny7gJMR{UZO47R`6vqsbSd
zki-7d)7bcF9_f7Yu`>JHvo@&ohVtl<uAPgEEX|dx@i>E`-`8DZxR3y$o*nOu@<o~~
zEB@_TR*i4&y8a~+yX0#y?+`QrrZzU*3sOlsq@gC6(FMK1VMEnZYnm5^gQM5M(plfY
zz|_Sh5uxtV{-^{wh%;r<W9PctiIjdCfWBW&T)Ffkt*`ducV;rJnQvW+cG=kJJ?$?(
zk)&&nDo4iYn5J@s+UVFJlqZvZ|C5T@99!67x;ipe-0YTn?M%JifE^dZJSB&QR&a<I
z^pkexDyr_)B&cqI*;odkNhB*q54=z=oA3Q>A56ZLCgmN)|6KA}d)+8-POVNKs>r4k
zsBxHKT}X9x#EHhlrJ2uFr?N3~EMq8L?30z#J+WB&C%Nb$vW}A*r}xRR?~7V)&<cOr
zNG^>lx1M!Fcw%sq0eu%gyohpMK9U}iuJLeu<+><D#veK??QZ_VYJ!8jyl33GrYH_@
z*zW>9_s8N#KVfaY6@M9XdV1>9!jCWK)oSdBe5|hx-}%j#Q?`G?JM*FzR8@5J4NHgh
zJqwX^fh-nKEr3ko%NNS@^mNbG=s|6rJP`|OX|-Bp|EjqlFydfUb^VO{`QpoQzQi+I
z^AghU^nE6Ne$sx1DhaSPDjwacBG&m=hEUfxGAbCi9@};YiP&jM!G_>ljUrEi`29!<
z*dF%-0$(uSx5gOp_h3oe%IbZ7X9D+#L{`q4w%H^*onyr|x**x27GWmH@biqJ(8fgE
zv2EERjbp|6dR2=O>4>r>YM7iseKNk8?ZHGE&iC3>{7^QTa3j8Wd6Gg4FoC^Rr%4xL
zug*A5Yd^Vh9H;NDWtSLlPNPlCpB6Qa$Y8>^{B>M9GOv&*Bp39^_L%|Yt-ePQgz{1F
zYDjb0SX3v|Ity*ORn(}Jl&_LTWOhbOnROd<8rsEJR&?wyvb4+-%8*YoM)OC_q*(9>
zVeUSwj?HO(c^18QlJz2<EDWw{h_hvV1$kf(`QD&o$3(%tl>rlq<dP{{r9j~TiT2X}
zKoF>q*;EqJ+lN*J^TxR@8MrNeFX>1Xc3mIJ(BQoLW^`m^rT47sCd_|+z2YozrU`Xo
zbC%)OjKt&#DScG7@~SR5nWCYgVIT35iSbP-ru(0-1xeJQ9l)uZTu0mqcN9s7s^W*J
zMq34SpHtgNG37NVwb0-Hwd5-{_mn3&gzRg%6xPIaEQ2$Zd|xrdQHQD&S}9}JD_-*x
zV^5qZpB)n%ce^(0raXz6v)+$b$+qz<bhr67O#EAnI!1v%ZSnYqnkwJ;=dM}@g9oo&
zkKOkqiKKC@H)zkE%v#TE4R?(ZVG@KVlFJP{N_hN6#J^p1`@5BFJM1ksbgR$0adzwX
zZ+Isp*mGtMAGkPDa8kxdbe4xKNO4M*7tZfqQuJw73NmQwrDpZH8LC!3s5kueJLpX9
z8so#e@I?9dsc5Tq9M5%_2YSO!eahlfR5dQc{7!Dv@|rm~M58S{FlAum;;9~uyJ%l;
z04?XIPlPH7uqww9(aue3;4HDQV@9p0VgKFBw^g|J;mr_Gc{Dacq9gC^+biP)4<0<Y
zXLB=J5gt>JA+`ykg5v4;>XSiB2?-mo+Ow2ox$dMyTE2H(Ck^UgiZ#~4MXh>zx`wdW
zZX7+Spo~@w0vocjsZKW|N-8SggFo&dyn(`H{yH}&k&ycoT~b|_>Y2kO)=6j*mUkeW
zn6ac21WX774DW8|2eYO1FjMyqhY4fLx~lixOHDP`SFD=~R#&0J6bekX$j0jzMD+E(
zphw7Att_hqhA23muW`y}##++a8?_4`;60(Vd8<>RdvEW#UwTHxjV(<H*Xi7&iznuj
z#0UbV$#u!r(3;PtuWSPTP2HuDnzNh@{PUt=XU+4wJ)WPKf;H`)aTXwWdH2=j%oEts
z9e7l^^`9Cob@tSqIQ36-VpJ#wiK>hd1dV&B5r+h8ha+l(${(cmGADAK6jZkCX?yG^
zf6G1B&?bk@qMG;CUB|NVw~;bHH#FDp1*gt<4Y|<+=bgmO=BBHLE9TsPBCZwsiV{7*
zmVtz61g$n;3h}ex*KJ&%ryN^)p>2K}JCHR`YV<%kAy_dlU`Z6F@E}=+@12pm)Twyz
zMxxFyVKJ2iNDD42D+Bf&5XIn1Sy=%aV9<2mfPd9h?7C=f=*9V}v~+^h|9$-#r4%=o
zRSgISHm+CO%hQ^E`&2H{SNo!j!qwNZgaqZk#U!VVjZoLBm>#IbZ(RP=UHU|-%rDu=
zrQ^QK)_I%jbk7Ti;L1^37rJz>ypE#@jPK9-=O$=nu|vJjcry&GoGj9&PTjm-y&0~z
zzTIGz*E)(NaqsQ*C><Bi3!UVYCw=2a2*iCsjcW`kL*-{u`YH6-gmqy$Qmu+VK7?Xh
zVjbfvE2|WAH}JOoyK#vz^0_8oWxy>WG5Aql$d31%_xp~|$7A_S>N>e>9iQ=E)q18g
zMMz8GeraJI5(FYBuw0bHPLiCAN|bD$XtC$-KoA{;#9&x{f2<A)zpFo)J^9t|X1)}Q
znwFL!fNZt~7zu3<v{#xs^A!Ud9p_C=Oqi?wX}1adK{H``dKA$8VDr3V<@=G5k%=D3
z(IJB0rC{{(&c{r<{O`l)|K|cYsKR0~HvR?ja2P_v;@8*H6V<_(Uv~{(5^Wa=HCaxU
z&O%~enZH`3wFcKS^1lBjIi7g?2h(Jk6y_nfiNq{+hYBWPdPK&=Ke=vaiZlz+DV~O<
zz$-;tpEg_(0XsK@ZfcOj<O=KAx=1u5YVP9f&D!rdi>4Bn%er1QY=PUsHreZ5zaKLk
z-S|4h7abxZJFSu!Jau4&<D12%p_eHgS}>zVw3zVd7Qy!q`#dX!UCWD}W6wx6aj+zZ
z2PI6%7tbCBTjZ|OhGqfLLAf^R<FS<!g=}01iJ&TRLeO{tn@nKRyL5G_%VjH?=)O7}
z0}h){Bench-#@^_MCd6^f~d-Xd9Nnp{Lkptu$_CC8>}kTN%1}w7VLN9$}THKZqWBJ
zi}wH0;z@|~hE3u5@fQ?CR}u%<7+XrJbUh!T)s38)9vEOi1T0%A5g!op-}C+Jnknq|
zd+r|Zkg1*BYxkBBwE4-}4fRRrAVEV57-x7cHuUGnBlNw``@X#$5(`d6HCJ5dMNY~H
zIsqJjka7qyC_y@R`I0MuBZ6Yycsw@rFCOa!Z^D;O62(}^wdoF|&yO#c?$q-<fm15>
zZsEf$HkH?Sdzs<MnCh(Q(i+acT+FwZvTec<GV)ikquq(~^eHv!RqLm^U2ks%Fg1zg
z9wsZ8r)ZZ=Q~$XkB#>)pDM~wW@muou`HLWr1=e_+l^m5JJK}eO=iPF2NnEcH3X`sX
zTfFre6jBP`PvJT?snPWy)76{BTkN-=PguWGp3qnIp)}ueaM*kD^sksrlWN>`d&fWV
zvVswP_L4r~cQbO=#hMErX5{173#cT-6>_W4nHh-K@XK^_cYfwa1aZBN7nFOm>%DNo
zGvyqz=MiX;CM2Ms`di}t(63J-uc(K8zF{mPRyO`|EVO#BJmx9BR3_%-wy5Nlyf~Z$
zGZlcOd>?;|*9+`Q_pz0$GOoJz-0i{!1*(f7vQJZ7>bif5x1}twpFVwRWn~q{gsle!
z>GS7zMl*Hf?ZC_ZZi9GC*FYjCltZ2$!CggR@|~fK4yG+X^)mNn7zILiv3%l(E@^90
zLO}sLID@|;AY6U5y|o2*mCHo&@8@P_LFhU=YkBX7oek<j^K$d@i1GDAEnFbJWB%KV
zg6CxI0hZkQE)mO}F6MZBGLt0tIWQxL5$AFBVYA3Qo~iPHCsm^oirxZ?oD5F7oc6<c
zlKMcs5pz;YDYOn%ybQU4f4Lz-O)A$o7IKxlyxp4QCQiEFj|8zmt=KpxS<%cN1WH>{
z9?Hw7N$CV{og+*d@75j4lcc}gnW%CbC%87*@c!i`N3$b80mu6lK@+;_mZpKu3Bj&^
z{>DY(-XEGZ3T-gWtur+f3W7U$ljh#3$x3&;yZz%?khD56_9tm46enLs-t-Quzd=2g
za>KEElluXd@efvuP|OM5mE?++^X=_vlHFUkV|<G_Yrg#k_~WgVFKXd5V_{%&F4;z0
z{UHp9XY79|j8$&_!YtD%S+?j`9YCe0FM-tUC&>QQiY7mdYDND_s$sC44e0twY%~pB
zp{MT%+E2(z<z7r8=O|QMKRp?M81Mt#??<3;d5sJWheuX<M1|EK9vvUM1QuW?|1DNd
zGW`kDilZ%k+Gs_FTRG$<VDtoWBT|as`p1tC9Mu1z>8itmTA!wY2&jmFbV!FtiwH<c
zmvl;pbayGB2na|wNDD{{NK1EjcXu~@i}&~adGF(+aL#%6-JO}8aovxd<|;frL2fAa
z;<*91-p!b`DJaI1xi3BboAa>)4T(2plDKf-_=B{7cYaR?`qIF&RjFWqbWT<@6;BN|
z+?xD=UNoF`3WDZZ7RfD*GO=Txs!!=ABXty^AreAGQUdRjGdR|-ac%4WUTt&N+g;K}
ze?O^by0Y{(S6zwlogKNcrx5b4mbT@sOGT(JFr?WlX>7WzIB+3My~Trs6Zl2KA^qe&
zKQjL_)`0uX4~Mp-dpyyi#Ry70F>#x=5IUKIWxghQlar>=nO2_^PwUNf)>R9KBD+kY
zY7_KsALF{nh1|VOE}@%HWAiz(bnB{5xAt+o7LAVOY99st`|>bh7HaFHI%XwBMN0xI
z21P9Xcl41qg-I>GVj>LTVy3315V*>~NPxO;1#n|)wqqDxs>E&^b)M|V@X*j6|A&6o
z(ed#V`IxF6d8*mT3I{6!Dk^A%t+%J%z{R$DeEty8tMZpBMgMbsS|ra2@pM{8`7rOS
zd#Wlb+HP)dwyEWLGDz*Ak5<aTbUF{^ISoyoDErrO1+<07o%+vpyAiI21&e+QudTEE
zTSxr^H+AL9y)Sq91u#4}X2T@(B-+m3=JBM`uZ_6Hy64^ZC>A^qqG#E9>}Yw86<<>!
z8Y3j%0VRh;vT23CxmTQpLi~PeLl0ep$JDI+C?7YcPuYrfxSE6mCr+;5%6%3q0VHqx
znb$;wVS+A*>Fct_14wTYTP+#)zaiIZ^-xKo_kEzGLwYZOgGWY2A@-19waEGyT^oHS
zek>*Dk{*=*RuO3J-qwwa$JaY1?17RNJyhU-Nwe!3)T-x6{8=yhEjHeG|7{PKbGxkr
z`%^pV{${CsOyrXko-?u=j6=8F(<tVoI!GG_dRrLML4D+e8isP~&99G{)is-&X71pR
zw{*xM)+rPF6&Cwo^%!biQ9kToBh`O5C~_j_5et|>^y1U-)m6a)2hyS<fhvve(nt5b
zZUdGQ$K@2_^O(4A{V-5|gIZrK;8yrr8g~}vIeQ?SvR`?5%dubD5@F*AJWtbLMW0;s
z%Iz^R_X2G5K~w?eHuFCe*hoIFFKQ<H8nIWyOG^xHII6QJ5el`e60GgqnEUjtsc-vV
z{@(Qd@JBd~KTBA?O+jd$q*thgnyBbZIMLA7&S%B&JZu6leQTy!w~v)OfBWm*_Vy0C
z1U+Ochp#Irj_ed2=vRj|7kEk@^G_b9yy;^n-MwSm&hJ2I+stP`gZ`}Z>zl`qea%&i
z7~f%dr-zD);!&mU(s78c;2pjR9U}{6$?KTa=Z@m2Wvz28ZXNiV`OEEd?cnv>ROGHi
zcdX#rqbYiAW_*`6I<j_hcY<I87LRGo%F{c;$E60ns5>SPu!6$1G6}Ns+%`?%H4sv}
zky1UN%>}jNLNU8&2xGK_GSr)ZIofft#)a5)(Y)dL8^bD?xvk4w-#(E=kwHU%^^BC>
z8K7*O(j;o8L4%%bq$;Jw)!qKw6mu4u;CtOM=~nwolEA=JZ_l%%LNx;#hlnOtk(=&t
ze{lUU&h~X6xGA&?e~U?HdiBc3Y`mzhomHPBeyPC}uB>UlZ}U-!*B8qdr-|22=1~m)
zfs5Sa?S^O%O!WR<cvZx$NIXfSEKetkN-MTuWbDvGWESvKQRPVB#rQr-S#dtjx+d9!
zFdKi{;cAb<`Fa^&IYPWKm&Nyso{C>OF1yg@!uKSt=q<b`zF{g-<!H~mn+|KvxtEDF
z>Y^?Bey_6|#yrF#%{8qON~QY1FCfB~F8WnA+1JOs5lMjOT{&|hdHkZ8B)_p}_`Q@)
z)x1VuAG`8!`p{W#I&sndC}o3|5vtB-5j2}=hDHoL0bv&C-o1SGdK~i@Mr|oFk<VRX
zD}Ou1vWwfWoS$2zb1Qm=lOt<_&CnmqS^nzNh;*8Cqot{gdUrBGd2h7SqxUJj@cbYE
z8YFyTX?0UmM0r(5c!2($Qg`ojB~#Z~#i^6XVVg2X?|^;SOgTMd*V6$HbK}cOj7?jY
z{BS?i>#3|`wqNtH8~8mkM=ub;_-{+TBX^oNa0y;J>cgGSEiiZel99leUqhqGZc;^S
zj4x0zLf<hVGU!Ek%ptbNF7{0r-|0)*|K5h|)9f<u9?7`N7i00<zKr<~{R`y;k>c5-
zGg;Al9{07rlx*%XkR$4spR5RBd`MlU--B>`z4Q`_da;^B$`etn6eGl!MO_6;G^YLf
z7AIe=n1ARD^NGnuhZwLt@HbL!Flz8JDa};2n<$iAYzixRuakEAx_nuul7*wb+~=_L
zE>RrO%&9JObaEqy5l)I4C6cJYdfW=#PI#<zkA^b;kKWRAoTN`CWe8S`?VTZ>pQiOU
z%;d5XwGJ0vefJf{6i!uPJEx{_2_?xhpI*!e+IsdF;n}6zRJDgGg1A^KL|vY)Db>WB
zM|q<G^jfJY?Gzx-QO2CtY}N6Bm4DdIe6qZ8ZcfKsqsDUfPWATwQmjSty0b>=udEN=
zXsJfd-hXO4Dp-kvPGf6M@j=LUBPJ_@aGzYqOs3?vb`A;jL7};g&qhnfN!F!6aMcK(
z)&M36l=QlDm}<HXmHIFR+$s<Axb%!&{}dI~!`@}njSEP3LZ#QEhW|J#BT8knLh8Sx
zW4wB}>1J{-iY<SkyR$d45u%a`Kj9~R)bo;1vfs)yQY=#_gz~MGNENq<7>%fpRgI5s
z*b|X2+L46{8OAS_bK|wjs+uNq)*pHiZSJA?cj|uNZ+?)GY9tZ%wg~lYQgA9uFUhyN
z7YRQ4BO`8+{Vrw;*?1}b3R%OdaVGdHQePfo<Z9x~M(U(eQU^RrC@-%Ybwtj4IxopZ
zF#bd1MQAEh0uP1XUDl*f{SekYH1sZwtJ05OFyCBI*98gWIuKc9SJa13KI)^ZXBVM8
zdDhj}ER=8ie)+}%g~)-CC}>}1W;Taou~&9MOkaRTkj9UcGPITNxQCA5VEg)N8@6Ns
z_FeQA#GS*&#<po}-LH(R%nUM?IKJ5LE~>1Q$duOsIOg=!iHO^^<;HL#Oe)(d-B0;l
zgGj)p{6a%tA3A7_MoJ_z`jdv&nc}7PqrSB2g<QXpl!KezhWV+W3=v-vb6)j+{Ps=L
zK!A+(A0Kt^TQ(16ZmH2B;>RZbOZLFkIUfu$sqjLz#zF&kDxQPg3uKB?{;IwogE7KL
zwR+5yq3FXenD8D)1^PtQC9l%;>=J3OqCPTAXk5dKFxqdbN@UvMWSwAZ9$RNMd0|`s
z%vTq2o?C`hZNu6WyPk1JfVlZLa`G*KS1=9^bckdybxE_N>RVkHha04P61mq*=sU>l
zTNu{p!ltP8-3Yyc{z(AG7gTo~mR;A2mb=fS1m@W57)_TVYd#K8{^@#Tj%seDZ|HTu
z-YvaX?a;|j-A${zect(YRC+5o>jtCpeOz2zjH~=9$tl^tM{eKG3J6kO!LA$ecnOGD
ze&ys?U7p#&^}N^d)4e80Uj_DAVA3vgJ!LMnm^J_iB~u=@C(1ZdTTs4d13AI;kE#Vh
zav(8}RMBW$S|6&<Hz6{ilqY;>3k%XT@4=^<KJrGXvA(|UM5$P&-sSRiuy}_n2#A&S
z@`3fk%CC4W`(<owY>qIDo^CnR#74?sH)P#(RVV~saP0GIj)`er@_6Y+GD-S8JG#x6
zCZD37)#dHaT3_;>HL07Kq}{@7BvS0D_ebAXeEqiN^@MbjeAsw0czxCpA0oz-(*-Qb
z>sj%N-o4i^bMhQ)d`Y{AhhZU+6^jHEJ$e&KgtnZudLLb50b!w+LHNdu&C`I7w}xNL
zo^nU%_y0+gqlpd^N7W#=5@WeX?@J=}ERez-jR~)tlDpgh4N;&;JBabo$OCg5o1Hh#
zS5bklYI_c%oi3RYl7a%XX!=qm=3~+e%FAPXoXc}#@5AjFZK#y_^X}nx)`&_!Ycgz3
z$8(a4-Vhh~`1np^=fGiSNX5^A-~rlMNY42O^9J>7AMexsA{T|B{`4qmPbQ#w-t~ei
z4T1!v61jm&=(NdB?+0M!-t@7JPR8D!Lb`T0B|fmKwzqk?Wr-M-yIv-Sg>akav{m{9
z9<g{ve{1Ja;*M+fe?=$z`*U8DpDRV&wSN6?Z^;nI&<v`WGRWl&<eU0v)QIDBkbafW
zJ$m30*MN{DjoF>G8p19_Nu#JoLU*4hT4#-tBD>9c9f7mT=b*}MN&I=7GFSa7hN00*
zRRnEboUWGed0w<Ld2xk4I<ipQFWRIOQRTk5o!*be?RP3^N?6)6ZR62%O)wBiI@A4g
zZ}St;`SYQR&*|Bl8;Yb)h57ObJwuCbk}kdG*$T)-OXt>}BWa-fFv!Lox4&QTgsyx4
zW@BBv`lOKoy_w0dD^A<VDH^stbJu~>gtjtp{#H?x_cNrZ#6&~z+W?&jROEU5n`#eK
zG;=iQ!M-Mc-F55$sfdY*2`H}%%F0>+17Va5g;d|d!b1M+!L|p{E>#){zKiAb0YYTk
zbn<T*M9~3!-?}aLw#AA7?5nM;p1dF4mwKD_E;CUe;8cDoB8>UfOr%&^<vlJx(gcqy
zCk>CwLoP3IGg7R7c=??gB=me+?EZY3*Dpg@lU0rK3oYaJ4H<M)u|92+6@Al)k<A&n
z#h0U+D0shffyDl}R9Qte)@~AyMp0*uf~Z6zU2pj7JQ4$5k+RzSUTel5$0LGVPVtW-
z&UbhY_6vS2ON)wpkdOVAL6$1fV|KTmb=QlA!T}52d>mDjW1kwoTq?#O^}eBg3_m5`
zFnP51vu{5YUlhvsYzB>NZ_S*`rO$dRIn=awbMo+Fu;dlzx}5}21|2?`sBj1ZiU6u8
z@MENs{jM(NR#r^iKRM#W73d@PgMY~S%hQ7{Q<sZKE_-mrz5<E^KuF3kLV0*33I`E^
zxR%}hf<2N!GhHz|X`Q2b8yF3nJf)a)CtBB-bok><^Hx&U142L0JjK#Nj<0FR&Uv2h
zWHx<G>HTn&N*u;&VQOl5xTXLSFx-cT(7K`_Y%>mG|MyU@fBR8HdB2yqM32x8B`5h}
zP*`dEusil0x`H?5PsJ<|WDJXMVYn2faTaK*>5Y2KP5W+A3G9-8So8<*NoSvl@rAT_
zGorGyIfTeH2kGMcFk;-jW!tuhas%-@AkBYN&HR#X^2gja^Oa!Rw_ZJwqA79|f%GfP
zacdkY#Ew^E-!gP*(@=_yqs_bPOKepoW2wJ%O`$ax<UgtnxUXiqZHDpr$?&npQsd@z
zWan<D*E7E$RHEm#==|V>c(Xs4tw0%yTw=ETDcOe;K_To@Td-_3AOW%GMc@$)N+uAv
zR=FIr18A4)ZDwI1M;q=4CI|4}tT10W-7dl_8!ZV~b}ADkY;lvL<>1d$rj_CsWG2Wh
zlc9|wRq?q_AV<ETsnJ$o{ym-Nxte!)neE}piSQrEeBZI;i*Ix1CNV|>nV8hwaeABi
zA$$=Uqv9p2cyucQl0`ZJYcAY^=}{7-HzmhJbAkfoxH0MbeZ|YYZl4714O4qzN~ACa
zYSyv-Qn?+}Bg%qAM*TXZmz`3aC4`;yp<7)`bF^ARkEF8~eYiDp_dZCPsK4QU^BJIp
zu?tAZ@8rI_bu`Ubk-H8}2hJF(`wAuHc*jJhNw^&4(Yt*)9^_JF_UXreMm>D3CWNtt
za@F08kt5cY)HF1@ijqNu)V|uE0j2=2U;EDEOdrGAzHzZwnk?+zT7wSMe7af^=L^`)
zgUsj^IBN^Lw^0qQ+Ja9jM3L=uuyhm1c=y>xjAU8o$cq-~s{oP-_Qt9?OWHsDxIjd%
z$6RX_HxX&ae#Cd0H4u?xHZDTrbg&|$5f5#ItB-}-x|&mf)IWBZddEHZml69yXD4wo
zvR9#8^ULQ*NsF^F{77g;E8dJpyLBW;B1PWh0aNt;Ne}LdJ487IYOhlatw*Zl#HH|T
z?d`4S$>xSJJ-RC?md?jZAR0oRKvCXI!|9?nu~F%{-5v3p@+wL;v%_oaQF<?$BzDmv
z{q4{XNF8(0A#_bWFYGCgMfE*%jr5qwLTO}{y5(Q8vL-F6Xb=nH1nsv<hj_0K=h@8G
z-K%3vb?;d)i=hNQ&>MCcXv4*%z6~rTwiU_ZJBb;|Hnxe$#7IyA0|*ax-$N;g&x~9u
zLk?PmK!o5S<KoJT*X_*z_U~?yyd^qNff5jd|McUSd1<gL@LT)l4FChMy?O=yJUKZp
zt`4NH!{A?D4odlV!DBksbgR7Y=we|Y=t3Icx$ne_c;G-qYlT&~IeBkpYNhR0oe<9N
zxW!WEPrb+?OqZ4xO&x(dth*10*Ifi-2uWXzH8=C`E9+yvA1<@s#H*;Nh~u#Hlgp3-
zRVcieCuy?C^^84sZ-U;yI|e6*rF(6#3w3#U$;B(57FRlMb|e2aN<)gwij_fF+cEgj
zmg<-M6(%xAuJry@IFb+~1wic$N_Q`NdV5*FJ>5u01q;?VKn6VGdwJ=$cE|y?1F+oz
zJ?5%?#YCA66|o1_!hijXt2%NV@AuX)x9eH8jg60DM#;);XFW(*9b8hd7X%(9pOrM-
z!f!sCYx321x~27rHS3}7I~m}S3!I}59RtJuoIis#4Bn8~397}E*TZ(;`okUwPjkG~
zV*kd_DsMyvVT+3A6gd>O<lSn<FNt&OWv<;%*5R|?6?^{QXh)r_FheGkKP8op8k(WG
zcs?!V4&}t-$k0P5LLZtiE*>$rM3fWx#Oxr8YmSq}tPfLCl_+Wd{T7(iNJCV9D&r-O
zI=*eRb-3dU8W%e*{i_!@FR}Jap;hsJ-~6B5b}2&?OoZ(J-Y)e{_e@7LOQYaP7(Wt;
zPkj<2@IijryEI?YjfSSCG;Uy2+&6~<a*cWF|F9bFnHzsQ`nrrL-P_mxyH(q6d4Dhy
zbwp?wq$oo@*d4YeB0p+d3I309*2E8#d5h$_mUNeb+TQ$sKhW_5RV!qaO0^QD|9+--
zKb|SybAN<64$;v6h-$ukl*(>u%wPXnpQ-i-u6K}zX0iEpJ^#Bv?v0)vl6cRI`Fk{9
zzVZ-@NNYSjBjLFw7DJ5yL%wP^{P)AaLN!^%1zkSi6N^43$)c|hdf@j&`{ti>b)8lk
zJnbKM=bZ^y<tq2eT_0GApnf%mlYUt6THtX#{-LixB6Ya@+6(pJOT(Jwl!FToY6KGt
zi^nt+0Do0fh=-5^>b0lqFd|OV%lg7&BD56&26$!VTPGuEIYE2=zi*x_cUO%qPj0;{
z#|8;&K;t3RhrEPzwo-|-(WR+yVmUMVe9Mao;jOBxo%tvH`K4B&tiGaiU#sp9m+fq9
z_fJ+g&#iWA$4~-@^FuaM3kMP8V)cyNg@z-#3i+70FbBZ2u(d@GK?KfKQs6)VhR(qG
zzoQM$r&}lElS0h~%m@>{cjWr-vY8Da$bkuqhX;a*CaPQ_Zw?;;>T8>&hamG7OM@+_
z_SO+$$(-jc3jdxE*R)bv7`e>fKc!xBxh$o>r7V`Y&mUW8kI`o&UYJ$NBnGTAu#{Fx
zWe)Kp?JA43R!j;LAR<gfvF})BKSFegl|Vda#Zv5*3>8a}%pB6&yjYGpI9B1d!_8QL
z?MP^s5pMObhTWvD6So)U|BB`}!~yqock+h2BnYJKh0XS3(v%oSM@K-IfD)6S8Fa<a
z%fO1%GbG^9L0eSt=g&#f_h0o{n7q)%aKwk`;{q@PXR+w_@3#caIA4OV0VxKixQ}%a
zGBnhyU2$(N8k~MsCkWAkridS@7PdKi?FeU5_{l9EkR<f7n(f@4>U5eBoFT~9th|K&
zZ$_Z0|L--b>z3D_UAp$GDzv%jH>h{^SBBPq#hudiYu_(br#hfgHvb%duJqb|srBL;
z@+D2hah{CXi#~Prb6(<5{^Q>Fby{t<J=*78vFc}=9)h~s;nX{f5)UTxdr-UT>v8N#
zZ}XX%)6`6A<TX!BOpNET7{x4kOalV0nv=W^vgktrGBw$@>zxrKfQ7i6557WVxy^jj
zqnm3KASu{@H#IjeEH6jtxKIJ{hteN_|6<as7_gTo%>|J7Fa@Faib+?FXaqCwoqrQM
zgbsm%TP)3b!SMA3KIkN((xcnM0cwQydTniuQ$9aFo)E;9@M+Ms-!v{rp0^;{h2ZQg
zN+EZg7mg+zQM6yj+aE8F|JI!U@c8PY<leFSKsp7V5`KG@8t0hSA6`bIA7c1oD8trs
z$fk>GoMN?@b=(gz&C_-hkXsneV%QIS`4j0REyWbra;qd3q(9K-rL5v{I^<dH`g{ES
zSs}S1y(4^AvHKcPW=406zC?}N+T}`5gy3XgDKc=v5a$cID1Vu{d$aEA3#XWLhz=rX
zhIsFPhO?}a?Sr|^yp>n=_qj>s{d!<M>9xldm#i4<i`*(PBlG?Nxyt`WV$f9pM&afb
z-}|AVsR<F5J{7YMO@dJct@~K@LES_p1|Jn1b@cHPI7aM&a9||3n+ELS$wiUs(Z)-H
zCeqW5z<wL<7f@FP!OlVfY&68Am&sD3y%lSPc~M-PMJRz>G&k@b-(PU}?@kks{4PsG
zgL3&pA-B#;;YJBU%pgL8yY)r~4oOeIsImmhL)OV3eWv}>?td@VjJ7Jz?n-%ViJzWV
zhJ_rh;r5YJqTb!AJi0s7#q;@)Eo9%nOXg#hPpN^SGG@<cW&PQrVecc1X#G^tj?8ZN
zCsvHecpK){vt<@qQCK-?7^l%#rx<z>B>GLo#Mdu!bg5_*)JDvsYqtcjq&2v#*+gkT
z;KCl*TDZ}Yy1JtqDqAg3CDtEC?T5ynUsMF8l0DE%U*)vyiZ9NK_$QEJ%^(^BwaiMn
zRD?c!h*(-W4*#zOC@d-Yg60X^fT8V?3^wOSW-ta{=1Q;B`#cs6@lH!iiwmoIPw2vV
zip~#vVX|UyRKal%&U;EPFH&$2&EPzcYO|Itb#?qw9|kU=1|(DKz`&BD2*g5BnM240
zzYrLn@Iau7BPc?HxK6hh8!HM5g3rt7HLE2mCKGU%9^P#)+)F%kbLGQC`FH=_Ti5-e
zwHO-Oie@p8Qe2skNQ^(Mv2^k|<JO<kz2n!%14;%S9;+^KE@vj&&YUU{1?Mvy4w~mA
zi)REarCljQ1)(?*Hao8+&dyEM-J+=yE;;<06M1=l$Di0UlY2k&B|GDM{2?#?@^ej(
zp^;x0F8>FvYmbZWk$tMuO^?IrWRsc`dk1@a>!dlQrx>2ko;?(BofaOpgW*f0A5*g5
zT6<wwA)lMkGfkukG)sUy8yCl5cuh`DP6XM)BkYQMEf7L(YG!72akA^Mpl1{lkuABM
zt^>C{896KqhmZr32FZpnrBqhg>@tdi;Hg&YJ4S{>DB+;TIMmMZ&ApXgGJyGx-M5e3
zNk*ajhIy}`y822y+YAM&GW_x9MrvwmboBI~D+JN&wtt<1<NpqQ(EZvwD_rINixHW{
zSo_yQ({5jlD&+C@Z%w$&3rcAuTzc{lg2;+6Lzc(6`eDUG&hZx0C2ds>Cnlx(slciL
z{IIMS>Gh~GD{@(EMla%9!m~<sIW|HlnTpHO$kw9K?xd!c7V*sl^?H9%Gh0&pp}Z!a
z1>q+@6tk<=M?u}ohXzQMM_SP~Avr@kWSLO=LCWH)qn|~PN_OIfo%grL25*0gw8LQr
ze?M4&xGsf0k}Rb}HWRE~ApJNE!sRa2Rb=(Z+5r_rzFAx#A&R*FIFy2h0&)#Muv#Cs
z*-bi}q@zx|myV~6npS&FQ0q(dvzkv;Mo$lgN^^^13jWH?y=nIVbAu0Y+i~t;{onmc
zSNRi-#Gm!#dz%ZEc6Q<9qJM0bQI&GlwSmFEF%|m|bM)Ne3+8<7O-pm~L7=L%S>~hm
zm~hL}zqSMVhZ`0Pi3~ZKTIW--bcvPTa=E>e)j?`*?G}UM)j>Rj1<psJCEM@Q+%{M*
zCncAH;yb#Z*`;^L|FnT1dDd~RKWFFxpB3!4G<D4v$<sH9UV9YMrIqe~EZ$G-CYFwA
zSozX=nCR*wW^%i~Q02AVX-*<TyQAZ!{#tXF_BdNidhe?3PxxJ;&~d-;Md<6P9Er#0
z)U0&Kg;I}~1j|}8RCg86c8`AB-k-Qo@>pEnxGQR{)U*wnQIH&}L@Zvj0iXL)Y_WNd
z>+wgJes5%6RAM<gv>1QD+N*g!?t*|)NgD?wyj#04ya0T&cq(rcLllgsT|zNp%4QQT
zW=4B+j};d;UR`m6JQ9dQo837<sRXXp9JNwM&{Bf@Yk#>L3iluI_~A@oCqam$e?JFp
z5AgfPpId&O?m-GNWM*>LLy>iaxk4M~s!gIo<oq9I9K;6nfN0%NPNJMeUgf-7#reuq
z@Ud;FY%}rU?Vd*%rRuupTljh^ZAbC1^2MJjc5{m|681Vd;@1~4Nq9}34+lz^m8%_7
z4=xAB#qT#u!6z$D#lXC^S!j2TB1+KJb+!AkYl(2E%OgR?Ar?OLjfn~|i*xSF$wvt%
zozw7wL+KJ6=UTcgmzS?wcKy2-5`pHcn52}H15ll_>l_lDKCn7#a<#2(YzV*Cz^`nn
zUww0R*yhpMI2NjoTltpOqgejp@i(o>N~h4WvNE#1A3A<M&7VOnv}zl3Jr1fj04HC&
zJmp_#pcO@jkr>=4qa^~>2?Xa*vo+I=&;_3W_SkF%4(Xsh&6b@se>IH?6?Qq%BSEdy
z96ThI)zoz1(`{^QbS2MlRkAqd1Lv%7{@4tAP59S5E^y{VV-k)pncH3Z6$amLc1lPG
zB@z}H*jrV&IJhACko~yZ`N{j1EXgNQ+;2ZpZ|U=+r+MI@i;6zEBg!Q5MTFh=Bien3
z`$YMF-c%UaS0=6+ZD*HPmPVhYES;xkRg85W5GnF=KOtn9d>?+QU|?S@&Aoljz5PZM
zYg{28ObQ|YxB1>4DYoZGQ;KlVj&s$+$B%Ujlp$?pPnlSc=ZmNQ-;cNNCYm2+js7fR
zb_)#-WpJpn@bL*Wt;iUk<i<P1E6<%rJ_-rU*^Ae_rdnCi!&v{-HKA>GL~p|4p#0~W
zse0#GfOCLq+1DXcMu+9Q{CDR^2(|*vIIC_)5NVhU8VL&f1}18mEuo&u$LVlOa+GxJ
zrK~T=T9MW_e+^6|XLa=E-L@Pa2^BYIJdk(GmZv4mD;VEoCqRWC-G?SKcOBu*ZRY3>
zo8}?=v@J0(!+gFj;Vd;@dZ5KJx~t+9x05h$uvTw2*Ws?kec=CNRL{uBx5%*Dp6+ex
z+1xHW_%IV<`+Pr_S7@BI)G8epO_O!xB2xK9kJ9%xxV+OKw4+#=PKP<#k|Un*oHbga
z(B`ix8wLtw3(7*W3@mQTS3w-4R7U8H`6pq+&-h0FYY@JO78TitWEkNAq^$$5Uzv_F
zezno%4bfxpQvyr$YP5N*2F73!OTbW0og^tSlHPrF<@WeFgoXGJhf!;$TcfItVt5rl
ze(mghQTP=O*HK@IDpdv*uf0vUL6uYv_&_YQe~bI`wS8EiSpG+tw625mpM%Bg?^?Fr
zhodgagy!hkHkd`Cq-3EBvP_t&M5=?bM4_X7g{nX0RT5o3*3YMeiQh%>KhD##`@`;h
z^<vd*uruE!+giRYJ#|{4SatGf6n?5&t#^3gr4MM}*E232+b!}U7}g={7Bl+9G)(@)
zCCnoC3M=oRvQ^ayIcPMfd2Ugs8=tFVic)7gqZw<;!Sp5Zk>Ga`a!4`{`Bv!m1a$C~
z7kll5<Rbm76Xvhr;1rG0`SPECCYANh$`Z@VgrSQqw$Z#)mn{~gUE@)CfMj=d!3DzZ
zv`0=EDmtVJOIC=8h#cPpOVR6c(5~LkIgFaxkFVcHL<Dk1y}~|##js1H_&Lr)_&DI?
zk>Le!mwG$Fy#fg=z{NyKJSb7)`!i_{@6D9&O-somc-Vjc&LX4wIBCXToxGW-McU%^
z4#vRdEZKe_tWOuZ_%iLzW9|z!)KcP%SASc4#Yf_ZbRGmLyIK46{KdJN((c0-FJybp
zVdXb)M#}Oym)Z8>PF5@Fo8NKOL=u!N$L<F{-iSec#JJ6yzf)e~pVZ&oW6A5B5HwP2
zmTKm`a6cP!h-N=-N@$kVtH}aAJ9g5fIaRa-QIytdxs*b-CD|bf&y&+TjY6L+eqG>U
zHMSkUs_Fc>k0j|vRAjtla7HE;Kw95rHMMcceA}eEy2N7z!NlIaL;-2pe%`u_167ZU
zXxs9`uw6V1q-nB7kJ1mEIq9$wL~S_YWpA4(W&8*qgX&+HE3TlRPLB;f^kb|y)Hv95
zkOqI{8db~=m!K?S@OY8bVOR8Z1rg{InD9V<`gi*n!}o2ET|8L3fbVHZrtW!JJ{+gd
z{GV)=yC9hfoNn*_o!=GF{0=5?06hiVowG5&)ZJEGvAC%yh^ViKh$!h$sf0Doke9aI
z8x97Kab6#yunD(YvLyOJNl7xnYg7+$r48)I_>tl9t>-U0Wl^|laYx<d^Iu%mjiaUR
zh#9Yt%D>ca3Jkw2^tFQ*fpGri^;H|^$#~pUv4IyqtrSz}%)&Ykq(;@7vK6OCnbP(C
zRN6&wQLB!abw^ocPk14CG8riSI+29g_4i^@@dJ{{wa<g&<IjuKu^`DNfe5nBJ1jTt
zHyYis#96}d1iS8O%t|fMKFaADM)dgf7-QRt=>&0DmS-+(`wo$R(1Y6H+kaxqy0&vJ
z5P1f#m5|2O5b}Cjl9K0GD@t=tF*0k>$#fgu8SZL()&3yogJ%V)#)HO!GME?cH5*6}
zase)$HA<MDjf?%j`+nZ8;VWLvbkO)_)26++eMMoR=cIYF9&_*XbWr{5GXhkMpysqR
zcXD)ShjcYh<rNfER8)fB8u6sI=oH*kmJ4^DR~A!Y#{DfVNH^Zgi=j>KErS;qT<}_x
zl1R3uYotLQ4k4Yhv)ZFc;CG1png5SEJ(NQ8(o$EEgrT@%&I$9XlW(LBccCt^yWU%8
zES*Od`l%o$QMHFUn?yEoS-V|S#??>DhkPX4Wf)a;Nk+$ZzluyQDu+$fRl<P9!2eNp
zXX|GM*ZAUSPS!OAHYEc`$Hkyg^@Q5Z$4u6H`UF$FJ~av?p<1i6b_7H`y<{1Q5sR3N
zv7I>_KdZIr><!zu_=c^{1v;iLExUEWT3~bm@v3E}da(NF!0;IwSZ2UcPaOfm@JKLs
zOD`Aw_m|e;B2&dz>}%Y$r!)Gu+B%sj8#|uAu?R8IJJGn9=9$0q^I~pf`kw6)O26y(
z6Vw{E>+X@CzwCpieVq87`U~<WMcnrpUT<8q20DypJp**{&<~ih`Hz}5E$_n|kj76T
z74<I1n+f1{!G2b#CBF>OB^2ZS0w%O(ye@!>imFqf44%LT$H(nw^t1W&1Nji}x4Kr1
z5B7GDf0n!dOqw3Pup5%Cpn$}`hl+{{&eVtk$7e;n15j6<xN@^f>RMR^V^B$nh$97$
zVbVtYe1#!o1ZhGcblP?#NJ~_wssFC#FP>=xWA?dj=dWJx^~FNr7d7sxx-43dW5U|@
z=r|bMe)k-keXq|j|MkRAN{no2bBu3+Ka3=~YBh-5PqszP=&YB|YaaXgu4Zq`w59DJ
zx)hAQUy){c-WslAD)l&AC$Nl@xaZd}&YMFNMVIGS`b8>6gyjCKghW|sT%FF@?rR;f
zJY@be4t^Gk{8-tOyZl%W%uDgJ2gShOUL$x?CvL*9sNt28$AehoWwE{caeJCt6K)SH
zR->PWOryWOjVX^vKo+<6eg;Yz>wDemlPKe&4k7(;&DiUwo<m63UU?XWhxYF?oTIS&
zp=Z6hN>(`|+dp?NlZCdncv|v(e_Q=qo>o+oPC)?%2g1*bG3naHs!Xw+gy}jl>8*1o
zptF~t#Bg(SgFaz&b2z~dWAf`K-jB^e#O<43FE<;0f397HCYMt-?7q`kzB1iIczbNI
z(cpqqQ$rlvDU_o)UNU_YHUr}iT++520qu)xKd;eEa);OVl(VzMz$AxU;bARiz#U%U
z`>9D<xHmUS262^cbK=-Dt_s809JVjg>dD^&bxX#q?$_-(K%cX6Li1bbu7Zq;v*lCG
zpNZDo0@Eu@>eWPb^}@)Hi56Erm#z}VCx0JJ)u8c8&fhyh4A)5r#)#ygBk}GptDha=
zzx9mEHq4b{_xB9p2K{5zqF%c;znvkJqX4c~Po9ojam2$`jpzl+XU~p_-!=AStlc>s
zPia-7xtpI^2)m<4j#dTC+{@9PQHf#F^S7LEiGWhJtFfWhfkni!KtN@>u3Pt}(b)+%
zuW|ZiAQ4I}I{wz?=Hw8)QPC*UXMUjS_>;cNEBqLI(PKXPE^oL%nQm_e37x{7{$b<W
z@)yA(X#S1xrs|A~?J-Z@msx+Fot-r^DIPU_`0ybb)U30j>gwtj`czkm-OHS7O$o;|
zLLEYWLMfGyS>#=>Z9eg*Am>m{K>;vx5pwhoF+$SPk1;SX-unhSCZ3dm*^OB?z)A*1
zzN4lfQiijp7q43SY52d^ssFxq5At_OQ`-TY=ngEY>Rq`LGB2+3Gs3;rk^blkySVCS
zF++N~vm`OOaxIjiIo-N8&bvP6(G`beeY^UK2@^U_XNaQ>cpe9NSGOPZALH_zL&3R~
z)jIKnOf0Rh#Qf48!TEk>#2)L~ORi5Eg|X}J?|ln6J-2S?*8b_8{3!?8MYh<|G?cuI
zsb!fXRD;G2O@gy^$4>W-tTK}exu@Y-iV=Rp&Q8(IWiG!WB2TCoF4fp*pI=O_Yc8i=
zbG^*rA7J{eOYjEMlj0XT9hN_xfdv(+K~XeFqFd<%OB3HYfWxi4DnbGJq!nBtkdU%`
z+c*ZIXCXNyhW8n-hewSq6ShfoV&XFB8sDc0zaO?LQ-`Jp9_Nm;TEvB#Noa#07YmEm
z_$D_dI;s3`fRZr$8^69K1+d%pr`>k-`Aw8>bu~BGh2@WLR)ka>kqiO63ICM7ToRI>
z{G0}j=*^A5xU<TiBl|YVZkryP`8{&ej;}aZ^&XWe9rs*QMU5bRS?xv+afo^>Rpi98
zs%jBw+c4iqZh<P*y46Q$xY7J3Y^~<GGI?Q{m=7xV{jZ;fGSn2Aos$ZEwT?9}cze-7
zj7y&AmvlEan`OMD|EEI1Yc6U3^+Pbl;Hx6Ad*8ggugx3{eRx>=i-cVU!z(7r#=7jJ
zQpd7o78gI1`dlk_yQ{=KzPUF;?d|e0Pdog4i>|~kh^4Qnn$(3)V};$LU)62zT%MHT
zT7KL3lOwQBis==UG&njeN$C5Q^ad6N)Be>CW9Dd%^8FHLX=H1%J|oal!HbBd-$>`U
z-Ph@s;m><w)~Le-{Uwki5l86!qUFHHfr-il)5{amR1+<&kFbOQVl8UJms2oh0!#2D
zKuF)OXdOdREpSwixiy9@wv$pJ3zEhaa&j_<>n{lY`;bGua#y{5sAMpo`VUqIe53Z`
zXREHm=Q%IURBDG#rwqjUOEw;xkxs6gZ+<&JviL(R&FIO%#p}BJWZmXCiv1W7DzZ;=
z(BM)7v60G)f$<HWvEyK|w4+z&(lST~sKHapI(v2Fzq%)>INW8MVh)<nKdrO`7Es|l
zl+97(6j~#E+gUii2?HTZ(TVr9!6MIG4fj<L^(+60?y0}Olvjkt@tos1qqA8LDOXiJ
z%QZ5>sOK18<Ju4}`uFBN$^KQKGO1O-dVgg?#O<$`QE=kMwK(TcFd=O%9sxnv>oN0h
zASo0dwh9ixKoDotO)>_*xYMe4hVlyGQwnt|ZtJ({F;NP=%=1D?kS0K)LKNf!&&aJL
zq?H`fM=P*q4i67oUwhsEoNjdUGcq#%y@|1fkDMJuS7qvf(05L6Pw%>c<0@9(xG5VK
z(PNk)&>8sBe}O4LURGB5W~J%nU1qI+6E3NB+Pb{BlYA(0MZ15i&Zzx#SA$YP#bp@E
z>X6&?R<g$h>x0MKl#$BwBWd09502$tV7F%IM7>b-D`S0d%uR|z*f`$rGr4LMOZL;Q
z;U$SBATXoKC2Eyubv`u?U(<YscRrtG4$~_CvWFfh-^DfR#SkG+Uh_OjWIJJ$mK|D`
zX+*-i$C~C>bginMWC`;KIV6=>*MZwlS?+He<l2mRd{SF|9`~T~4sUtHjMWyZFc(!+
zE05O5%i7bwCd%^bxulIrPkyP>`*bg_|12=R{(I_5ls61XG8^npU>yct`pz=cTKu12
zvWk+Rbmi6*u9B(JXa*bZMj_W0u!;peR_S&OjZZzSb7+~H2l6NV?Rob+vAn#zEt<v9
zKAgs(U2xm(36kr{tPccpSfB3c7mm`+S~^P4kyUcTTY2;54fCU)S;ox;Pn+!+lpMU&
z|GNtxCtuW?ngs&}FtS@=oXJaB?KwD0j!C#YR&k9QTQ5jiB4Im+saWB!LoY}Yi^lgk
zZIv^OIks{a`&w?I`d#xMoF@)%37%L9ge7de(m)D%>3TuB@4(I5&+3{)9iV!zdnTB0
zUdH8U&FO1+GT7FxMrb7uuy0nhG$y<{VcaD;b&lx2da$zcqdz=@4$BZ(RZR*hG$9ne
zV!b|_kYkK#63y;DlWi)U?Pay>8p$5K^lt8a6Vo@pNgbFH1i$AT-KCVr%^sw!xZn7h
zEN)CZ4IP?uodS`R-XHR`knQsc#v8;`wd1nziH?pAMkXe!XM#fZDBk?!=*W<*6h!o9
z4M)IcM4CJ+s}KZM(0_yJ&KFnamDSZ7AQK{N%G9R@p98cZ`j}3pB^L}5G@g2k|IXPG
zVh_NhyD79Mw#lkpx+TAn6rFW5$*w3A4S_vIfleZ*`qW|vp)RbDCa~(l36Zk9x1Yza
z9lrs4t$FrHrh{^okh^Zz;fljv$>F6i8qE`8LCFI?(V%B4WK^04r@b$S47Sj4W4I2!
z`P$~@$)|oNW0vV2$m~Y9duaXQkc`N{!olpIZIFd7atYV6q%TscDArGLNom>RUQL=0
z%{DH#nCs)j#_q2so-vBG$*K;QbUp_CpetJl^xlnp_Y>46&@;hM<5d<#Oud0kfO=ml
z10%AXLb$qNe%EZ?0>Rr(;M^Vlw|VLLUvEAcOjmAV7-n3zU8d`|kLNAfQR5wvtPiJL
zwr6cQa0v-FM|{Ap()cF$#o$wpodM~i@vGDJ#lpCvkq@k?{#7gGrb=fw2+H5TRQ~mT
z#HaEbD!=BU>*QW6S)DpZRUZ*4R)i*HKKK^r__?4ca&?uMs*DZt=2F=rTFi4{g0VUk
zAgDNaO*%`ocgl6@T20pUJ4gfL#uM>)?MLE!8O*{mu3+7~kIx6RwVb0C0C$CfK)+1l
z)$`=<Igt-O4Zoh{Y__1gu%YeTD~if`RTf*;gg-mPHD}UFygM#_O;v0VPPR24cl3Vd
z+ci2{XuR#qRzsH>Ple(qb`Q6Rwja+648hjN?Ev~naau|^!Zp*<%+X;~Qs2Fx_`Sf+
z8qAW<WtFU-wY*sk0`cg=?~NR^tcYZ47}<^V^RYS2BQd~W6%v7ASr>i>FE6k2sV+VS
zPskz%meXlHr_`Rpvpg3gK*)-nph1Ue@6@%~9te?Na<Bz<^OyAAdhr}(_9MeX_2ZPe
zn-l!;Q+=)fl#-w;(gAl7K8O2hyD2?Sk|Iq3n?sNen_yZL6^B&}zX67z0;Ad{d)4hP
zf$`18Hu=nM7?FARWBVKb1}l#G*U$ciyDr#DZ9a>-_X*{69fB(Zx=QvSj7cE1MfXMc
z%zrzh-?x926zxOjyc-cuYy6Q~-f*iPvt!j1YI{cfudfKv7bavL{-YNHxs)CvWh$g{
z;ySV&T&+Aas%ydz*@aJ5Kh=myX74F(OIvg9Gc6kUUOQgYJbN!~O>ye>n%l5+KaZ$M
z5YxX=Po^Yy{l?D0kxPmr9$Ltd?qz6^2<veKpOnC1PLSk}e-iKINpN!MJ~TYciTbP4
zK-b7fRP`;e6zHfv#|jg<S+o$KdXtT+17jLUtzg&V?%lfpOTri2-rin;_nk<-7#8Nh
zbIXk7!K6`u4w5}NdaX{u4lqS<#rY*Xi)V{{R5Q&`aj%j1s|W)a_z_;Sqn?Yl-@qde
z3A)@AWH7f}w;g;R61t_+M}wokU16-rFRt>|x~ia+RpQ1iTV1<9Mnfh8&nd8o)zdUy
z%s)D0azA5U<c~<W+C`n^4GV@r5*BWSGs;9B3>m+Co#$ZE8KMmME_rMxFo~2bqJRXv
zqxJ@b!Jr7_iI!#R@GCP%ODKhIdZA!}OE08Zn$$dIHiu|4sk82V*_oHQ<Tc&bM<>KH
zY%aJqgo`>mkCFpP03MO;#j_(^9LB!F<x~3_T_@l<#7Nor+fe;w&B{G;to2<-JuU3F
zydrk=k=ZHh(gT_b*$S9~wdSRul?0Y#d@ToBs4|ah?*AdCyw5(?&#th4hxPNmW8>!q
zPtY?~s^JR->)G`5^r*PF4XsU`Nr;|8^?^<|5_O~7#S|#x8=uWf_p`IJqt2;u(Wv^_
z8L-NN<scc_@RjZDNE^bWaPy~`3cCZdmh51)Rbl@a4<+K}dApur(!C$<jd^s|B)xaC
zmgt$wV!vc(9FE>kd`+WT)H-0I*`UYVk3qON>oeW-6sd^m`eX5^OXWqlX2asT_EOU^
z40i_s@gqXKZ*ir_(S!8?HN}8igGEdYKegcimcB&c!ceB0v#)^p%-UP+6*DEXdU-B<
zMgj%O8#oT7iMoSRi^}woFiwqZ9_40%3l8<q4<9Pbvqw7m4T%?WrV`)EU{>Q_lS!S^
z_w?Y?VHqHgpN&xG70SG)0Q|&ii(Ys_F78NkVxQo--<}SFxGgq<S1qc5xMaWM-U2dE
zJ9YEBEioi~KgR}Xv?*x^ejVwUz8@}7&3Q}uL{U)@2M6cJJ8D2|k$mdm%Yo{xk*B1A
zw3!|Q8u+GF+dfk^TpS!g^%xl$d$d1{buGbmq23K-X{e{?txWf2czsf6?I=Z+0v%a4
z6H_>lAlT-q!+PAvCi~4UsC5~@I*wHuq9XTBF&gqGg(U_GH;?OH{%gQ|GPJ5%aH2XF
zaq9$x^GD(xy{cfqGL?E&p1ynLu_GR3hN|=6aj#wcr&b%lg1#5@=UFXWs8MIt<!oOP
zRqyaBSyP~)`iN%q!<-LiFkDQ~erwtKwX?JH$Dy|vd|?2w{Z`s@<a&QE(*7$A{boVe
zqBZy;v=|iik!bM<#NY=YmZ|5ae0R%;;)t(lR=IBHB$_VXLtkIi*FdQh8ekGBNtm;#
z(oE}Vv0}M=;;Z^i{za|Tv`U~&1aZT46pYrbYm|V&UOE2t*WgX|LR0;`%mz!>;Vw|W
zn91k}-l8fiC_qJS)G08T4Ash>wN&0}-?X3KbzJyvu&dF_2N{XoH!UpK&_~6@%&qjM
zNJvV~+E=8Vekd)Gzu8|1<Pv=G{Nn$$0Q9`9coe}R0DYC%(c7<^VqSB*UqnIMwy>~}
zE44=2SUxTY@a^jQdP-@jQo1-8$^eirPfM|h(_z2@t5QVs_UjJ>YHvE=Ptv8H6h7$7
zEKxa%5(W>Z61d@M7ra(i-jqt$I41=zCo}xPK(@YY0Y&8-H;JkX)YTK_PN4u(*QdhW
z&$2S4hhA_07J6d%DT4*kj;Jdy={e<G!J}1_nL}q}pL$?QUeoC1bw6)&FX!92$2T>`
z<58eIft9^~$^~@17UQ41U$SW`O$LPyZ>pcC-773+#rDig`7u*`B07lv)6Z{st~r<E
zq%2r7dF<Ct^LmZ=ufisH_U>rcY*d#Fo97oik)9NoF=A<?zQF*XgG0G0oqwZex)I@v
zgw0if-RmHpD;&vwm<*y=k0#1R3fxG9QJVJ-5A6VKs69Dh2h7U9vF`8RThM)U>kYY`
zXw&$C7AAs7Ud=}`2*S9<(!hTn!Hch~Q-ZQz8!0P`wm|tC+-N5!CkeXd2=*~oIa#zo
z>=(An_Or%GQNgbT#u_j!vY@UXmJ_UO*m6MUEz3P=(OYv<Z*WeEDD=}@V>XsJdUr9T
zE6zFqrZA=BN>*&^_gHvnkR2dQTqAC+9|(+&bFpXrCPfLWKM&W}dSY!>E|JSs$1(^H
zdoW&_-dCx?aO$c-D~jq9$~j1~r}BCBW6)P6r=)JOIr-8*x~1b#Prw>qonFbuC&kgD
zXp|fPnumz`nZ5qhUaw?{8MhL2{_b{&sPok(E4L1^IKFt1*y^YsfKhEjaGao@QdoC5
zaDFg$e@ErcxrWzSr9Wn4|E3ZAGJ;_F@{YRYK(lBRc+r+<&tg?ctUqb}MleY<178gN
z?d({wNtjq#t|=!0s?FirU|?7RAa{5HbI0eao0}8_sO>tX@#V<&dKKYf!;12JJ9j?a
zf-4COl2P}DxKn6?MZR|$ytbNaNU|k0H#P0)@AsPC-frP>;UYr$%s-AH?BBx?FGYar
zL-%o;a`^vCIY-hu-l^WVot`JaPj)u-`xsaNA|+<+c6(6I0GLL$44>heqgSu5#v*ap
ztz&d3YgfQT^>gwEj;O@+U2o3v7W?1LN+hmdqv-r6p0Zd3s<bG~y7+S(raK08@tO2D
zVn)Qz1-9zdtsjPnkaIL`@MViLCU*Cak5`z8S@LB4c$Z5#A$O8fYqrRmaj|^bAHefH
z@5M`TYY(?0KB4!*`o@+uCXV!;gQjed-QA}qN+<8ifCO7asTz6n$LG6_Tz#rqy1I8m
zTi=4V0$`3}Re*<Q<{JH>KZCHXv>-XoR;SJ4;RQ>*f<RwiWH`=nPr^0Vis96Q`Wa>u
za`e#xr1z;*;G)lQMAkA0k^+E`T3cJ)Dq6OM9HH1FA$pC=Q(MIez8N2SP&(Or=|s<a
z^IbCT7z*b9h!TC*W%x5<Z-7=XJ<r{TYPd$6{;3N`<!hnqV3vXG7sMF`Rb5Ma4=;tg
zbP7w}-szpu1IRmE^(_gyuZUW5&he!sy97X^JosjV*)}Cgv$FEDm->G&`ZL)ukMPGD
zO1u~AHS*RsH$e^}OBbFW^PL`H^G^MKQ-Au@0w=!5*Y=HC!$<a=VYe^do;>&!fmLcY
zZ@$Fws48%L<4sR>Vg+LR@_umpBD9b2Okn6&*3^uAsZf_MVT;a&_h>Zj8OMxHK@=<A
zQ#*N3t9%Oz0$Pet^jAYI09t)-F}~kB08I|?9&l9<yB)|hBTqhdLNi~fV3I4qf~^<u
z(uRw<k9oHKF^`^(j*e!dvkN58LtX_g9^ROhq+haD(Wpp9|M*%9XamhlUAnvY{=br9
z%PUZ@vVKZ`2_uK)pGygiGXi0T%FdaE*FsWNo{a{EtG$n!q%|oqG)YP>qTnmFKIMYh
z8QNg*Calc!M<(f^rN=hzRm3NSQAyYE?=BrB`wI?>0NQV-PpePgk}0{R>uh3aj5iN$
z5P2kAP8-{3y^K8_G@+QF&xn(wbNss>m|l@^_ObiW?P(3gACr#tg3ql({e(Pf3=H!n
zjyMj4_4-GydkEsGR~z=6L_zlxyM;lS*ZU!i_t`VUMc+z`q<5g_!+3}o4KYH%gOrps
z?d;@p-4sU<0;^>)1HPsAAFa)aVD~-)t`}$dV$PDa^1$xpc}RxWhfSp7;+KGzR#*LM
zFgfOs*%Xt7Fye~oOZp7yL<4>O&#>Pv#t@vNd9ES=CV@mjEF2u=>w-Y=xYuYB;qUs;
z^nD2sn&%BX`hN`qs-sJCs;)Sz@nGOcK^b#<Z{!qt$qrgYza)KGV5@18#|7p`*O_jE
zqP;5yorq`B{3n1_L*awGfsV1?+3&|2@9m7(x*z>%6S~c<(eI0GhpB&jLX!T~_7%;&
zp4*@GM3*`DIbIFfx$h@MM|Kk=y)!vw2f_ueY*;^$pcEhthIt<vNoqiBD3{gdRnk<J
z?3Z*Y-%uS>Q2zAuO`mGew)RAYxkI+-@1JuV2b*D4o@*;=XzBdnW|9uniavBEjO`r4
zduPNJ+P(;1Vxi&(BO~O<_D0y`)PL~v-FGlzhqGH)ZKS2en<bk(Bz|IM);U=6{B=r7
z3eZVl;ZG%edqa?ZO!A$Bb3j%Wd8lZvdRbjV0|Ef#koPyce~Lpupl59z0(Tn`k9&v9
zc2#r5c6Iy41|5_L7g{mc$}1}q)zq90#9mN*x!Dp-CcE_%I~qi4s|o_q|9Sx)e6GB)
zJB-eIPv_dD7cRL=PH-YiUuLFQT#y&N^7u$K;zKn81kv<%hCFR+Q}z)anNNK$lf6_}
zCD0B41OWvDf(dql1?y2WqrXGPC_h+6Q!Gic7(>6TrqUyiD8goeDcd-k4jBM8yt!FB
z7Uo8DOEWo7aCYh<sW7F!*s`^~5Wb^LqFp+^dwzLva3I3c4>J}dzJ|8u10FRo5jUYE
zXHlrAv7|op^kQhHy1UbATO|RL2t*$z^M@9Uv)o-(8+Nk3tXPB2KWs8T#H81>G|FYl
zlRLr44($jX)&uAe6HNK&;q`(rfjz!UaTaPw^ZlGbW7q~|Vq+Vst$r#vJ5}WZl;|kQ
z$<D06&8jCy{MO2XiDqI}7DyA?_-m4ql5)$2;bUfIVk$7xNMx6RzYX$1AIu6Qp+P?Y
zFAY4pY3|EzH+=VRBn=|wX<9#o8K5+vq|9m_kvlO6sKK)M%O9hrAUO6LUJtJZse>=<
zRqVDs&m&lXXqi8$S;k8kmD2=wx$jVZlAJ#>fqsogR~_N%=}sS2fz|urYwQ^wkExTG
zfIO{2WvDtMQ63_;7v!ynPC#B(x1iad7de#ym=M$~kSQOM5TK?TyyA>)ReVXp6)z?3
zpJvO^Xg%00KjJ-@%0r<zu82fUr*9Wgh<5n*?GKw+&mOa<3_61Gg9OT*?=sibH#{i0
zFDylu1CkWuhl!=uUF~Fp&&L=*29u`a+v)nvWliBs{qiA)aIf8?chs!b1HZ@Q%12^k
zKFB5J^d_s8C05i{bx3k!oZeB$Y4cvj5gX!vw!qV3bb;M%#yV!6$?)*6cSltBaxd$n
z$o9cm1Q&2A0r6=E0A~RFj{DV`vElps`-f;1XUkGa#le%azP@t1c{z+Ysqd&aUYi@Z
zx~hky98Y^B!kURfu4tvyiSzH$FZ}H!?ofGvc7+&(G`|us<7UMJ_TAaVMMX_*e5qjY
z%4g1&!)Z0m(4)T2R#HI$02ZJ{ZtOVJ>4#++62;Oume9b<ndZ_srfFv88DP?|5(ufS
z?-ZVfw%YzUd*1r{QSrYC9#x^%Ik?W$V;e1q`%3OhMpx*ldZS(e+6rNlUU9!Gp2a;G
zV^%zCsT|-F%VAraCg4q}-)&$fK?Oq^4EhS2I+aC_s=1kM%_*bH7lXzl89m#*nXEIf
zwu~<>Rt+T$4xKMoRaFsaWctcVP&M$?{X82}xhAF7*)Kn+ET>f)3La>_&ayO<T-K?v
z{xP3JjbiT;pV@=_MLcD^qr^ltAbsk4n506-PwATeUZ=1cfgJf|vuSCo0nX0~W|@|+
zO}73nUw;-}M5=Hr9S>;LOIFR9+jUHpQ9c)B<l_34Zp0efnTwr}Zd7?Vq~^L)dmk18
zoYXioTlI8|jm7Xhb8~asPK721VvgJZ7SV7!Q~>9ce3&&;WZ;5~Y`0V9xycDMhG88B
z&y^ATDvfQ%odMQI6_ZjBT1T14nZrlGGP_m{Tkq1+L17$TwOY+c1Ty}`7;p`jxt+7+
zRi7C*|BUncsl}OT{vT;_YK2pukfCwK<-qM2Ils)~)i{-PZM2PsoeFHESpQ~xiGHza
z+DY#w$%IY;8p>xF|K2gA%udYj7hk>1M)^OU-a8)a|N9@WWERTKEF_y`g~%*r@4ffT
zx@^kI2qA>*z4zWEB*bOUkYumy^*h(={rUcGx7Q!9Tir6A*W-B}=iJw^V>v&!40K&_
z{&vo}sLaXqkyX3|qmn5(_*M&dy7OIHB9bQp7j|#7wV5O&udBX0`7rR>+d%0^oko{A
z780W?KIKtxlohD8H7xN)XvsM^?%c?B1YO|DvnviE<`+Lm01z^UZsG;sIQ%rtdgf+k
zQq25<x*y?O4{O*IRyJlZ-p0j^8eGxT(fPW4^n{<kZbjz41@TH`c(`J*TD|)|16%}*
zp&ihlfUo@Pl0yXLQ`+^kDUO>_jyigJaPXUqr`a@lWVy*_pRhgpfTVur@%8s_dXg~7
zv8-f4SG#%L$VL3$VfyE^VELU6F8Dvf`ihUZ=&){?LMW(rYL;8;Lm{Yw;L|xhJ9G7-
zdJC0agB7v4LAbr|(ah))xXoeVWhs`fG2K#=gICL&HZ>vF(9)Q;Z=k%wV$2}bxpR8|
z!&w)(mepA01`%WEfYx<A3YD57-IYznK$CkNa^B-P9FmGqZXXfh0I(}KoQC;S-7l-a
zEYLAKU>d1}c2ux@DM7^Qum}tT%vm8JAtj}E=^S9x0f48`Lo#HDAIfZF&K0z8_N^DZ
z+pSl^psb@3r#bQmWec^C+uIyV6cBWF_SExmBceDhG^4Dk=?Cm3uBo2hM671D#00k9
zhY2PEUhKzTv~)(!9;#o1ssa>sPs9&|s<n;H#&bO#jQamg$EkkVj5u=d{mm-ti={N~
z$M?aH4HL|)Q)|YLY>MkP+aFt__TDgORDKEM_~75)<eijyZ*Dhcf38U`W5eL*EUoU~
zA+jZT?6S1<+5YUBPNU!W*c$u!-ref;JC=$#^fx2yx_w%m@VaDvBYX!*AM|k)8DPW5
zE0+GVrlTGq<hJt`Y-6C@tzS2il>K!7j^JH%h};^r`tacc%-BpMVZecHCz)Vts%BA3
zsvFf-ha~1EGx%C3Lg&U=Sy?F~Bja4$siYd)=(^1Xt=6U!2+Z{LX?jhR5Ncae!*TKP
zFt9>YRJ><k?_W`NjQ`io(ulH<gu%TIj^v``DUW$OZgQ^8+VN%`lXnvNeR38SJK5Nb
zBL>1~&Fsyc6xOx;dkdih2QJ6UeIE-uV(2OHS2bj`Pu_Jns$rsMq-@Q!TPc0UYI9UG
z@2S;|EgZFiw0pp48d~N7j$tH1{9CkAFZqfkENxTEXegMi@9OWNLFzZSV(jD7Vr{36
z<o0X756}gIa?2$Hbtz!dc$FEq*L+6j=i(AeS1@?<Mj_&cg}Mh~%14O0d)aZwh8Lh+
zkaPl&+t{N2F|nA`?M%yAH-3&*8&e$s8u-N_39yP)X&&+58I@Q4c$eUlCrHtfYg=>8
zELZlGlXyD6TksuCZ`}CCHi&{|eZ0Uf1@whY7}H`)vRp23#Z+|bhoQyL=4r5w@yM+;
z$W>%5Rg0CrexNv(fWtHsBHomjDZO2=b$rn+P>RRcBrx#`-J=M{+=EzH<J~sL+FmlP
zva`r_(v0s&6+^xU=nS4j(C2S!&^=FA_{+Gqedq&GhQ%Kcq4}WDYoMvJ&uzYk$Nwg7
zLjhZjn><!eMw`PA^|`|0w1h(+G#~i07L92#M3S~AKP5xzAP^}~=m7bqQib;}xO)j&
zs*iokCRPKWWet~t(?f-kLd!#j_ukdC%x^)BITkOUCqd@>#U?L};>%BsOzJQU0q2Jx
zE2^=VKi_qxD5@W)nWZkfaYeWDzHR>cX}EUke<*6j9fy}1)Xtuqq-H6Y?w>(6vZ<w>
z#fTF^aC75L5vJUGVqvWzcPEJlY5~&}Y>Ok*Pn>mMAhiod;~9wla0xI-+LRug@x36@
zyZ6?0#US%%Xy^7o`-$+~cvmg^_4!uL{IbuZ5ATBI;B)^UD)cY}mZJH3qNHLI2z4v{
zk3z#R`~OjbUPafn<670TUtyxX;N_weMA*#V8sQFy%J6;Jar~2RxyC19r2~&<)z3Kv
zC|Qv7&;ETYpyELC0BP}*uV!2HfI5Y@0~)yE`j=g_?D5v!W8;Wn`X4g^8L&Oqk~aw&
zyKO5I=F5)~7izUGupl`Jf?K5bMts1=a+d9XzW;f8bLq>ccUQN`&@Y+Fa9D5n(gfc?
z;2EhHp$KP;!+^h7qz~z7g`pX~XRn+;@(!u4om4k<)v6evS1TrNO<dW?$|gR9yo+8+
zC9E^{^WVT!u{OmUt|Egc;x#wo-;db(`Sst_|DsUWn7@ySB$s0e_@ZF>yuPHb)Qr@6
zpzM~C>2v#gj+p-KB@8*Ui%yJ0v<WhHh0ox17EDYR>uYM9-QIWK>EOkM+6YZ`;4M;=
z&KBs(#ffrbm`2@)k*ohAy5<|vy$zE%xw%IaE3*oqeMX0=R64NVj4r+pkO?^cN+$ru
zC|xJ0Hu5`nrg%ixLIWcR$gDwgdjAgeNcQ`)<oE7P=8zm59#&LU(T8Zm?c~~CpJVx{
zZyE0!0B6t^|3QgWYBMd#;Z|pcy9wHa#ognS%2b3vq-UCR-^XqIw;X7lZCB_2tjX&z
z&hH#6YkkZ9->v`gr(O$f5jlt^UpO4?+sS8<J|T|6xY?-LZaegJw(yO2^TRlqn?$Fx
zINKcSO~NA{v(5Z6fkc#XW1dL@(+$@pSAK=NB$tUVcu1`{-z4EG`{H@%T!P%|tw>7k
zTGk;R+Y>-0@pM#^EzGsp23X{hF&v9}ddM@XgP>rFVw3o)MlRx{{0^(40V&@5<@ZFw
za+{e>0MwjJjkh4u26)RaU>h5lEo)uA@g&Jg+fB)eO^-um3h6TXcW?mXHG{TBdh9gG
z*WwEa(0SDBpax_BCqP<Q6$=e0)%x_r#9u~64w5if;s?Gaz=w?-9UZl|ylT~ua50={
zqSrUTII%LAI<J^Opi8vb6_)ze@8QpDK7W`J1ZEHSdvo=D4_1%X+gdR$AC?)-kfWr}
zzYHwY6`Xaumv#X}24lmX9><UDR7@C3BxoN<VoEBy*<Wwr9j#Ed@GeD;q>qhQTHuf<
zhn;(wNi5J@zFbmj?Jy4_F3M4B{xnvgxgE@eTWm_^!B?`6>?WdR57d54wqL9FgnHu4
z@M?-&f@0mxOB6U7Xk6f>0}^u+0ZMPz4KAW?C7F*<godwq)gF(h%zEfGy&*MZC0VK-
z*IPH}r0oeK&jNP>HTOqX9Tk*$x3p9~XL5M8N@g<T0z03BIuf9_upCQ^#2~mREG_f>
zbwaeV5r5c&UW9If+WtoIWJuJ@tis{ygWd<!k%U+`>!6VWW*TJq{Tfd7l#o8{B|P}+
zmq9m;<XIq8#^2?MDJg%l<MpOUAHynoIE&=r-&(?Eo+R9cc0{Y#?8MV|dz<R|-kJN(
z^-G+ct80%MlX<l)JFm7@$60rD4KA5P3(0HcP=MORRcGm6MCtf?w86ueUbLiF_JL&9
z=C#vIqb_0<n6e9=w*&~U*Ab4F=k2_ZclOz36_S33%s{UD9mTu&;gRi&|4dgXsSXlD
z6OvvVsaCVH|B6|PCmi;g#HIB8g?XkM?`zKALk{)+MMcF=cz7PpqV{VZc<7GlYO|79
zp3dr>vs0qMYz9Y_S*x*2HOd7pT{v|YlXqAjr)#s0tU3dT3Bt$E{bs_K$HsF$XPYY}
zxWg{VU{a}oo~frNn43rt?8QG5DeWpJ0ks&h;<4eSK&>@K2XP)fLHeZ4E0SpoZWBD%
zWDHa6D*%(;J%3KXZrFccV@uAO&~X!HuFs!8gXIi0)$aW6C7koxmp!SMY=$`xl`k%q
z{+v11$i_w{V)i<ad-Tb20CjXjNjfY(#a3cTd2{HGN+(^2_gRfki`_aqF5AM^Zb@}J
zeYm6hD>tP;=Rf4~>UC4zMR*=I37q9&EM*O%jocnj4q~7Ut6$j^yrIWR#MmBK_c9~f
z`bO+yrL^FnWX@CS<gL5Hk*YAsx~db*e7>Gk%-$xFGD^1Bu`S;@m>l5islCD>Q{z=6
z6G|I4ODvhLGVbo_Tsw#>fTOPF64Z>da;DX_*`EWlhEYT52NI-)`PX{N5peI+>6DPH
z7N}ZU;IcCGwe*atb=1<_`e3)R9f++)8)e{joUrs6xeBfMU96joBw_rN5hXZ3MMGfa
z0%Wp%U#@VLfyKdfvs)2QgF*BF`UaCcOLDWyd#b7r8%JGGyHwYcBd7$&9_uvdF5IJ}
zoRM;zYJLn7V<7(lR8Ub}UK=g^BI26@JI9>8Fw&Vr(vden#^-soC0(yP?gCSk`%1Da
z-X|sEthM{>Tl!b0<0l=YEzdKDG|rdxFPBR%x3(|0e1h04t3s!)cy|5>>s{Tf!)(Gl
z$1e6@`;WwMN*AUgQBlp-VFw<o$An9lA|=JOF|?#SQ5{=w0U@*Ugiax^<Cn6}N=4gH
zT1RW@(7m0%6F=VM>!%OSGQYZ74EWw$H=v|u{uI;qq6p{j-_y!TzvQh=LBiqsD!F{T
zM`FJ!Yd;mimSOASK*9?-y8~L@Fh!@t9K|p4nckne4(i(@#y1cf5>iFp&kschTRkR7
zV_^aX^c?t{{v^Jqj=y`-b@D(40q!?tRuUBV>T#^bB>y`bWOvnCxZpt%WDigQ{mAc)
z)<dzQVoU`h0=tLm1qDilqi#FDu0xL)u?Bl&0>mCXC4JT8^Jw|?xC`J%D54tFeb7`S
zXwUbf6;T9;uC}$cwahOe8t+k4qeijaaC<M~dm2;S>bH+uYd;niZ@34ky<gxj4ml8c
zU3m8MzB(M+W2a1mW~`g}7BA=tKnxZIi~{wDz6mHm!W<v;9k>KKQt5><^Z&p(;$1Mj
zaXGzhrKA7;z;B=o-{-3QTlfJ<`RVY@EEWa(nwR4u=S3YxBrOM7F*?<m!*3<>mCFfu
z^yd@CvG{+vk$EA|b$!+5ZEB}YSLS`BhI`$L87EjsZ1gY_OOT%jGF6Xwm(+*?Uu9R6
zUU(bfjE%i5u_Onr5(Zc<7Zy6-doA43pNJ=@9N%u%<3I(P8yWe533PlNKzxLob%>vc
zXqk%YCfPBZQ$RW+yZw=?Cf?pu03UQOO)^{$z56_)O!rP9ARI!}FUAid9Qgb`*3a5;
z%Pu9m&Mlxa1t2!^`J)xo@~zafnH9hVet{ewB-!6-zc#0cO~n?i)R7&E58Af1dcI1W
zA}*FRh=2!b|LmdRf1LoM#-OxU!|Z;b!yg~PE*{xwUdo19aV13}8?{?}MhtOI0^?%~
zeA|{P<ZrlcucIHYnH1Jt;ODq7_~Vq~7f#3sh-lfl5UV8>+qH#V$oJ{^ajvPmx#}5n
z<#Ddf8&&@Mq!RKpAT)zB^_}43<iXo4+LTM$IPk(oSB6ZpF&uWX{7b~BX83dcQNi>F
z$FH(Z<pEodjd3Oqi9YE4K5yl0U(5M<4Io>H^c}f-?uxqTU0hUs<}J7(=;E*2JcOSC
zTM8685Qxi;nh~XU&ww(I{-McAvWh!eB2^vP2!Eefz^t6OS+A)(zT8gyn>ksovb?+<
z7W48Z*-qCb0n+T$9{VbLNo)>ZO<f)3(#Ly`2HYnEQvfN6LMnN9s@EkS3XmOH^N-6W
z&Z3cM@!pP4@$tQNtg2<%O7@Z4E+6$Y-=v+bMYL-MqiZkVV=!ea4f{G;vl5rfA-OjY
zjWQ0l-eja_Oe1Zpj=Qz&ZK(sDzV7vsdCtGSzv^@Tch!))U)imCdBh?R87*LcyZTdM
zp|wE4F!_j=scLtYU)Rm~l+okBC(I`W1NO7yCw-4nHV5s3sLfeqHEgSNru4VorwZ42
zOJy+P6Wk+@?52~9ZSIMjkh`!W;jxh1)%scCqLPZf|M^z)a27()e>n?ZW){1=Z#nv*
zln4U7P_3|ch*O)C@99(NbcJ7+7iSfWvUidpHBs}HY2_pt71c82)lhAd24qkzwuOp`
zh?YRTEh;7W`sxGT9r_npKmjp>dmuHb-;6d23+kyya|9sdmDrNGV$FJwqF(;A;eG0J
z?)JV~{7f=waAmoze{@te=>2=(^E?qA!D1(<l$zL5ufY2Hr(nbRSzrl9Qq(p-P+DlR
zwjR*haFe6dLyIduTM&+oXH4^ZEfu-faXQ`BdY$<j+sJ<&DgAbWZTnK<VIEv&Vxf6&
z(*5JR|58%$8Z*o)Ck^?Je{<RWKP|xK{)xb!)exJA;I4uKBPHSzc9tSyVim&kj)Vao
zOv+-r_NDZ(!Bzom-<9a^(Ch9i2F`mH?LUzZ8M2aF*~%H9k?Ul+z-je)LAR!OJVSQA
zYtuK?ZfQweq^+XSzdob!zLZk6EXhSPsrsNUw$D@=^`e)>PpltXhA@XB11x*q8L8!q
z^rWArgx=vL?$Z=GXw+owp7~4SqHR0o^9|p75l8JT1G+yrgD>gD?CncU1{Sk`sso>y
zPSt%u$UAu6XZKv=I7x;fUrV%b2_hzYw*x<{z2OIr&k-+`(Smf=5Bi!@C;0YHy>)ax
zhfARWRG6(4g^`tUt3rXX40sLrx+-x9fnRX_0y6TFfHXuRYj?%&63RJBdwG0Zw?PRS
z5yz&8ezPSQbAIo;Z_PG#L&Q^K!?cL~(2{UUmC7%gYu4Xvk*JSA*j=S?)ZA?9L}==t
z_{f6dNIABU`GAXeq}w&pfWl{Cv{oV9vY9GY#G=bwOj~~<heotY*7*3ejP1Um{i<R9
z62W-hUL{qHElTKhk=4-|Fd##o*f0?P=vRBrK(lNeQDD3WaSQiQ=tEzw!hJ$<916+g
z-M1Ygr_Xce(Go&CdfyZVUlhi$4HhS_RZn%@OJ@ubC~7u}FIoMIOVvE`{ISTx*k@z5
zHzz!%-p3VWm{Q~^3Kq<c7%sqIB@nSLx+8WIG=0(}OgJ5|iW`_UW~tfxK@7FNSnm!a
zzCuf_$Vzeoao&W&XOn|Uj6^`%a6L1xoCLS45>ByNVZ}#XqPkmWARa`y{8#pe#>XFp
z=nEt>+`$6=^G)L{@M}P`dg(Xot$Yn3)Kx7kUQ4|HQmv3Ub#Od!{nN{hwe$y5wlt7c
zx{74m9#CNX3Y{RmHEAgTiC>;N1);*Ye)rs+v8!!5rj6=x5<K1V|EP-VXo{MVMNqTx
zPwsyz1h!`6$imXvm)$=uJpy4w|Dw&(uv$4;^v|PedK18m9X0aP+2V{nEka?f(fa!9
zxUwsq#QemT?nW13rBy*(`kMasgWM;lOtG#j#1R+^9I3E^7ZFdLo6RF&K2ey)C~(!8
z#C*i-;=`}~r~d&9ds2NknzQfDIJc}E?>n_``5LSQ02Z5YaUqe!F1WUpp(hiSx1;r}
zGWcAGlvnUZ({LXeWS?YRdq83JliApqA!aqWON7L4l${5T_)~60qIpP@pu!$g!&YH@
z4MZHI=z!`K(U@$(4GyA3`}6QtkP(}^yK6>cr4xV>e$w^<`xlU~gEkduiIQiQf>3ih
z6m#0T0w$+rN6-aHGsKuwYDy;}=pGoU{TlN_q-Rw(7Q1$W%oCMc3Q*COwS$8qd6k{%
zKT>v``w22!&{O~Xc?Z!5QM!kpWb~i^Px~VhS$6FTdxz|JAQEiUQdFC};%^!sH=mwX
z^1ECZ%ak^kuQWG#GnAIG?RhkRmdS}=9ooG!qcI**=Zw+JW{!AQ`$AhPow`u2Y2x%s
z>Us;2Y;>^%KTRNZ>x66Wg`Z}im3X(A-QCJpnb{E+#p;yu<KIf?{z<I9=DU@R*qdNW
z%4Ea|f8jn7j5|8}XU|S8rG^dR8|gmmA6KmG8{N(IJG$R~Fl>cEqo809XgH{+3^Dos
zO;zK<n?zp&k*7chQZ{%UZcIju4-5{1g5n`IIW%C9K?8t}w9YkY&b5OEAw%#C!2<>T
z35<Xt5)BbucdL~*Enl>2czSrCAieD)*Xl)k_0pO#O(-$$UjWn?HykpuH~{XOUPU`x
zx^_#s=-C~tJ6JbAr=&1Y;?;%{-wKp}_5U7X9<m}^6P>b+&q8zHJXlS(;81!~<K)tl
z(RX#l;Pty|F0MJAxM0tdy}b&Rys4$cWAA*G4#;BB$X|aWJjKk5FW7Fi_Tg!*5xKEX
zplGSzJl<;PK$5>GccKb6VXH&w^6}+V+OTMaU&H{Jld-scGX_;|pQLbV{Q3k_4Pc@P
zy9CpB|6~BJwSGn<nf_S2&FhD_Q}ys&KGjy!L9#rzehn|`HGHSKNDmYDA*bCe()w#4
zohf=}MI(;fDq|)oEno(w8>|4}=&b`OCR|x?U9s!bwSnpP;p&agAZN|ZHBV}67+-&w
zAcNq1y<6JT2jBVW=SjQTCn!>=I_J$Jl<K`eWgeOY*4spua;4)e%e`QxMF~+~EUJMV
zD56GFQW2i`O|7m=^-{gq_4$Hp-;5sD|3LtV9n))qn90@HE;sIeJN7cQnW%G?=syiS
z=TBlwtrQ=pwsDID2J^S}Xjjj6CUaknr;B+Xxcqw*Ib^5zVlXSUKD0LWORz?K-!ed+
zfR+RW1p$J>6;a)N`z|SZzU4&|G1?rpZFraapV0{q{hSQWKk7ZIhQh`Be}cub=d$uw
z`Ppzpe6^NerkN_7`e(;YGFCO<mloBWlg%X~`H3-9-K7v;ZMtC73y<)P#!w8k!fNaD
z;Us}j<ZTA@QTx^V@g6PLr?wu;7j-$#yuQ#QF+V>+4<;>WC*Y9<4Gvg-0j#8HXb!Un
z<**4?2nden6gLB{H&>s)WLFOo#p$9u&;Y`u2FiKqFLiGux+TD=v3&%@59J{U$xPY#
zQU_yX7_e{N*n8HYC0f^B`vf3f8QQ3isV;@P5b)Ktez5f!V`FpE)XAxmF7T084R|0y
z-q@)Ovx`{zoulpfZG&yW-gXedK<<SteLw25yG@D1`!?P(flV;gry0$wMYih{Z-=Fg
zuXZjo^>sVW3k6;Fe;&9O1GZH%IM3q6FBa8Nj5)+6a3z@3de&Z4xk6!jTBK(EV#dZd
zx$lO5iW=@%AzE=_fsACk{W#M&v6)iwKzs3H?ePa2N?46)`KzqUTA7Y=fugF!DOtb5
zAE3XNoK$n&L^KM1Dmpy)?O0!8$}wt1+&DXzn5x0tJIK)+my>z?!(yFSrZYLipS)s7
zW|Z@2(ltCkG(({^UNU@YNd48{#E1;)Cie}6FEhlFQ8>^c7y`ne!`@wOJr3A6i1_q;
zPD~;7hfZ3B?;959lPBW9FVd3;W_LT*>%0%iBXfiWdk&4WRG}kdVf}FH&Ci2TnvNvQ
z5R}uRM8Zni(khHZr@oV($ont$_M7(OF}EmG(TUbdN?HmASGS(JN4$)xkWR>AMCz3`
z;keU{-E|B=h)-Q?OzhZg+$|5i$iP3e>oL?+aeKIa|Dk3MW2m#T_`q#*j|n95NRzJs
z=lKE#!E@@f>jzhS3V{VBTkN59T&ckkTkO{oLVmEG(+EL`2R?kB9D7ntq<7gTU*FV+
zUkO}hSZ&Z_E{{lUmj`=wHJ?`;S{1F&x&ki(a|U^t@cIEnUj~MyiV4YvlwQkvM<FCv
zs&Fmbj^}S>TdJ+d;Y#bFb?f02Y=suB*P%W>@%}Yv9!YvI{MsV*^69Ikm=Wq3W9}Hh
z(E40lTbW$`s%hS=pR&G+Kuf^7jcHIhx2fFFu(gwaFZO7x@%O7Hwef?LVLkkJiUzzT
zJ1i^2N)G}Hk1CwkwDS;!r8pxj3{YMQM%PM~2)t(nec72+VS6kqiEFXxp9^v;h3=`D
zP5LZ1)x0Fnlt+|hd&PG6V+mwcHY1-&1?2Uyg{$%?9p31b#)Idva8wM*4O5kb*_qTG
z6N{hW_lb%F1JMB@RPjF)u$ZoayAlv>(A|?}s4yQKuwO^xcsqe$C80}}gQVovv&Pmw
z01H3`KrwN2^JA|`9e^Yy6dvyG^^x9TSYPx?*%n9{s5$;qn?Wa4u3_`*jF{(-=t|nG
zozu~lF1%BaBm(h)FGNliA;c*4<h1_HLPwYUU7}w%cS_&+I<<*hk<K!fiB4y{ck{li
z^;xggJDW_s&;$-R2(}x9m--wOk_GMl(6?ag6BBd~4IaeoDt<;chbSR7N|ncXN|N1?
zDIM4sHft;Y?bp2iI)>oU)@vlUrrpeIR_Q!dVRC(QkDka3XU}9lK0Z_b3%Mcjs=j}C
zMbB0*Lgvhvs%8o^Qa>25^X`11-Vm`Yra&8vLDoFBS5fdIPcgph&NHeg@>hO-!zT|X
z?)iVdcJTOC3Ahn-s;S0e;#1!Vkx1$3=`wDBrQ77-)tIg^-5wdZ2zfk^T>W>Bd>gWP
z-Iin3Uoymi>&da65)|V(I_%1fL?+=Ndj^)e<Mh-cRlEr+b*|n%oe^^Hr9Zi~g2Moi
z=3EBXu{88{a)~AYui%dW>wQ&U`yZra@*%R&;J+u?Qg`e8!R%$hB--WEiVJg*VTK1e
zt30wx7f16efE)~du{NLP6-~vDpzNB|ps6$4I3Ko2WJs2KaVSmmukw;5SLJd^HpyQ2
z&GE8RWcYLRuk?$0U_5}NTXvnMV92Qo<BEN?aNwb-ub->K4lBU2ZnXi+o~<O^qCR9-
z<y1`2{4HYXqYT63%*;m0tv)Gii&@`s*)MV7$O=R<p>3lOBRW3v+nklLX_DK(HwFL5
zGbCl7DDJxR9Lttx%h7M9_wKa<5#pe~O_p0a@aSFHcl59qS+)YGA26hRXR6-}cUlyn
zNde&;K>Qdj?38#&@I}#@eDwj(S&$(oAK$q7PC~B<cwS*WnGDhd{weI0hK>%-9>^de
z&|wf>e^m3AdODJek#Y<}>8b+9%Li@WzlYNMZz8k_c%qCM=iYzTyVFY6f$j@JjX3-+
zpTJK7f`VbDnr9KMcnx^$y^g4pd+&Q&sZi(ooJgA!iSKH>PyRVY>n_C#DAd5)?XjiP
zD3^;&P8oLz{#HgxoiRW;d2|B7nJ?)pS|6<7tw5(J2b)!l^1`W<-u>XU->BvLyXn0)
z8}V6*h?p{6;$G0Z`E9W9O567!8bQbgZ>K;q9xx9OmYY$|NF8Y|PZ6YN*tULPB*-gQ
z?w^SLvtD)eQ09bt!w55XOhWcdk#3epvH$pSJo#T`(Jv9<2YC^^R1&Y<5C_4N>ll>S
z>EWm4EL0<Z+s*Ic$+oXG=M}5H=Jo3~dK}u-{zLvFrD8OpyR@KZsz)swyAMPb!p2)v
zrr!D(8)bL;kj2aDT)r81@HN2U-)O*wB;RcFF%%ZgPTZok*nxiIIuJZhUAHI~-h76~
zjM9<Aae<4~caht#4d58a`NlASFA64N*qPDU**W})JHAi45&QY65$ucs><2`WaQO;X
zFdp8;ad{>Psp<gL{oip<zmFkwV_;L~^(t?>etMGIW6U332Zj0<eSpeGzsgPt)4>jY
z$=~gu@a>z!E6FcnKJv>2lQuA<oQ{nq$9UveeqRC22ROxxp39m^eJ0;zmI#+J9v^8b
zC}?D6g&vVZQu1S-0wevtzh`9o`T^u(hDi9SFhjCJrZGaJx#6u^je`|~9J0gG_R7+}
zd*+}>)2FqS;=b;hfpF16LSzPi+D+~%MMw(dlq){_th<_a%=5+TR^wQEle)MF_4U$*
zzJD_gl%Jj-HRK!nd{J7-A0HK0Et6Qe`ua7%nVXa$tC{46pAH)b2O-!t=m{D>vFv_9
zG|KWO!LHjfP@KY&5UE_`1M+ZGfVIsuL)6Q(^@cAhGf#)cr{l8=n1M#x<5^_5OLz-T
zKRDM>i6`~AqK+9ujG$x%!VSQNh7wfNGi-G&V}0H7Vu4|P{S`+?HbB10%F5BjVHE6S
zx8aeQRQjYVfSnZ7URk<dH#Rm>0!mbvqhuJseWugw5epgFlo|uh#H=ty)lHE4{PJx3
zaQqT;>-rUqZex&SgPvydPdwvpK~>__Rp+f61NfzBw{ef;TWwj?J8}%oMWB@c5*~id
z?EdHsM#p6C+v10JYqi+GBLim=?<2M(P6Hcn200X+owExd0O7tiF}dkxBT9dh^zL=O
zmkE1k7ba3q$n2kT7C!H8SEM4s+$iRin||ko_T;u)|JQw^8?UP!>^|)Tc)6K179<MO
zv=7uOy!Uo4A*#jACn_?+>@F*_W|)e)lcGHIOcEp4CZqM|!I26#ajvn*0rx^WL6KP{
zle|?a_TRBFHJWyii9p^7oRm*2NN&i50lMzL0PBQ=dmzvU2@aRV(>qv@8VG(I3vOmy
zDdl`3fE-yK+X;?!7}sk2r|pUVlYO)`YYY;nxcM|~ZKxKo*a3VL7~f8?y7DG%P;Md`
zn(Z^mJT<uZ-sM<=5`?n*0TE8GcMHTsIo=Q-NeXPw=f_i3-Sub&ar-VrLT@>y&10u~
z9JG+Cw?mZ*3|o;RtHa59JKeT@?TTYIzk}xK*RLnPf16x+WHC-OL?r8q_kZX;P`)h}
z!XwQc&PR@WBIyM`$1hTwbTjNBaiOjKlbsVUs6H}Cda~Q!E$a1_Um>e>{a%%#!{wEs
zR=CN`aVT8sROu(M@FLtz!7DP%MoXdFa{8tg$pqe9e2hk+te0fWsq3x0_-KtJj8nHe
zjbrKiqW9)m#NF7Rk!TdfGRUO|9?9Hh^}HIk?3OC0{IrST%C^*Psl}M?kIY#apxsan
zLY0p&j+VN<ab|)L@I2h3py+h?BD{=aU~7xpR{?#txs_EY)Kq!`0O61QX5uGu3JS=A
zQ2SktjKl<T!dBZ}()*ceR*aBFc0Z&8<_K_RCd1IVB0YWf$GlpFqlj5X!BL*5cNUs)
zUd^`C5VQgB1sQywmHl)}@;KASpdRUaCzj6zn_?tFT9_XzO<Pr?w((Bds9~6FyFB@K
zDVF@qGY-<k9oI(k(egD){FXa_qC;D_E{uPE)W&15Y;WBiWC*f46W^@?j?RdswB**e
z2Xlh_aUrp*axqYQnPWp$DnY`NEcgB;J=ah5^4eF%WEde5M!$Z><t={BVR}EwIvuTi
zF2-HPT3EXM`nC~vAk?+33f}1L-VNRyYh}?QT(suy?(g{t@U*Q){1Xoii+LOpN!IFU
z=4Hq(EN<HNOtX=)Puy$~dZ%k+%qO!^^e*vVMvx1Uu4uS>**e3R6`{9LnP@GN^TFjr
zO)uj4P;9OiM`}H0UyUtFO&1~$(KUpgg05Q!9{29uOV?nH?l%Lb9<Y`Y{nml;mO~z}
ztH#8{pn|fpl$hWe0J<lRwE4(reBA{EYoNq#Tx)Hn%jyW0Io2sID*7mwYpQz;t{7Od
zsAWf&AanTin>p7#h`)TQ@Qq0MZ9@qRb7$wd5(z5Y3t{%?879m=hBv24>1EHwRGEnr
z(%4^WL%fHJi%Yz(#&-4J%|3$1*qEvnI*$K){zKP&XpX~LWgQ<1kEaV><6*RtEmEBd
zB(>yVtEi}wfNBEE-=IsT9}ag1J+%aF)U$E?@?xd@szP(?(gPmw<Lr9W?a!$)(f(62
zK?t`VU3J`d{yDW)Ws|3A!Q`vYH1OCK?m#L`TcK--e$}H=Nw?D&n%vkKIPGH_v+|5H
zzX|=~cNS(`Qjz^v&uP^Wrpu$KtlNmgla(p;Y$7&d#+T+P&bqEK$L`ffxfTCCWm!Vu
zC+$=9F%Zx=GBVO;hIlLrtL&bElwAz^qu5std%vYy{$+FYvbJAH$w}tu!`d)2vp*ro
zpwNt8`zK($fMTJ!ML4hy%1O^Z=ONxL@L@n?$4d;e!#R3XqeCZZ0{6I9@pozLzj0e0
z_6!RI^Q+E(odssGv8Ho|EPA4Ik9|w~n%?+tKO=DvvTr#tOI<2H&iQE_vrCWO2j`ky
z$Lko9@E^XrS2bWu9A+a!V@MxM8}+h>xnH!9G-Bzc_|8QEy<mu%i+9ZDt?XGh9ZBZ+
zzQ*sqg7tnGjPQPe{b6fGRbbeA_B-#x`U>9LKZ!mC2Q9s=9+OZcF5*FJoWPhC6(MQ=
zUj62Ka>*3@aVnk*8^uPqSa`-W5~Yzm%v@8$741gMji$!cC|>HXcArH&(5+V_ac<r~
z<0#-1Xmni0fz#%-<&>JKRtOhXw%<NIv80W$si~=Y<ai8JXw#+S`~0I;s?qUt-oN)x
zT9nciDgtO9DzK8k*^U~7Mu0zpZpi}xvW+1Et3*c3&~YsP5xv*w4ns-(IX2IUL;$IU
zmL7hWl7Q{xtv$bwp~HEohf{Yv-DT)~h_kene6!7gay9q!hb|LASbR*fi8u?fHYqLc
zNPC&+9ht9}Th{db>#93*{TJ33%%K%ww{h;KyS9RM8g+NnnWnrMYb&n}@6&%}#Ku>~
zr}cM9)M7d>;o6uIl`ZRAf{~v&jobB3%^#E>w>w%~YLb}tgz-{KVCPv>KG@I+I~43{
ziOy7+kUHnbnaD<@el#j2XPm;L8_kQTba}I|SdSf7CU>Xbu9<X{TNb)FNQr{=IDPre
zK#{|^0T|a=5aTV#^8WTM$i|W8YYt{}J+vVJ8rax;P+Na2&ITzYx5pck#nQ{H8$l@o
z1e<?+7dps&y_KoLrNpBd7aUo>BbwV2{S*>%m?|lzH*1Ov;d+<;fbdEB$sLYTnW%R0
zO2&9^??Sx*!QoMCV5&2oE|8tD0K7DO@YYajV@y@G78`Rcj<~ouQJAEca4PqL3_Sr<
zcnF6u)vb!rv9+}YcL$=F@GQIJ+HaEh1n=D&Ee0PWhLQ=2My=53whje2n$a|A&*Vj@
zG0-=}D4eoISLDziF03*u1O+w^OC(_-v@K|<p>sgYN>J^-!$D;GuZZ3DoK_OuwKVW=
zwoL1Kra>X)N<OqELNoOdKiBq1`!ORLmh#JP<|;g|3;aL>Te!jHa?kLxMy-0^E`dU6
z&SRZTGY9FC4F3=9OqY?OK<2g!;Y_P<Dwz1*k<Qj^zO!2~GE%533KQ6YJXOfu64~Jh
zSmKAM4244Q8>p4y05pDjT6l8^0M|wabRO9lYC~ETt-6WR3m@qLn5;pPhH}}1qvhTg
zwgFvMl1v@;%mU_qGpApC57P~(ekXk@bo_5feE+&#iAeI0_2^Q@?X^^<qUP+~QMz&|
zDcZZx6$10*ST9s2qh72AQdjRA&$pKrT?Bnw-W-q*MLiJG<CuG!x~N6+Ji1AA{a9~X
zUh4Pfv;dc)YzDbe397vesb_y5W$>HE42%-EsZ{meKYT@Koh`i{PANAnm5P;-eZfx*
z5eN=zBx7P!G8ERxwg!FiBs>?-nqi+$0hsHexci39(~Pajc3Iiy6SA)lO_-#&pZmZP
z)AkYso-^v{0(dqV@JJ{d9AybIQXT^RgM;|=k3sMawFEGQY@D&zC2+q0Z8N)1TRm<2
zRC?m9K`i}#()KmK_OyS6<XLOMNiL=Kci~@8ok2DkC)p<-bt92}Gi5njSpEXQMUbHZ
znN;fH$ZMRyDl1JjedS*;Frr*u+bEk#FQlu(-ZJJ`lO{Rg)a@s4r`;U}!ORIF+Z_qw
zKGiZ5Kfl<SZLn!$$9G#CF9w^7ZcMYhHTj-zuL%LT$l&I@asB?8pDS<L_KOcte`{3f
zKLr!cZW&^{j4)*^vfrp@D{t9EY1_W{1=>wj;hctG!G*l9S3Dh{XN9cSno6OyNe|%J
zFuKLF)$XCu*`(a1WAkf*_gJ;6ie~VHzh6WXy|D2a<qjpT@p$8f^64(tY1Ey+%!six
zr%cq<bYQJ(jBKcHomcJlp?xPzgl-K+HK<oTVTX{sAVeLhQUyPk1}hhI`(;gl8GsH9
zN&%z;0cQxxrUZFLoj3TWw$C~3k=&@rk-}}OM22Eo#2h~irKxvhBoCD!(E0uy7#M(w
z;g+?``m^g6++|eZ4&1=Y_b$5Qk8g{|VM(Tb>QSk()z%XEhoRKH<N$96h-iV^pA@Y5
zZ?hI0X+`qHE$3C@hds@HmAe1CbA-R(sO2J~n(KhRA;{6w4%B<HA-W!y7ZE&B9UhLG
zSR7YHD8PtK)X~*3YJFQ?_E0ru(fFR7{w2k|afCv}r1sW^4NrL(mC&P1Mm^$(X@+bg
zibm8hr=^5B2<<C=d4*}(^WIp{jh%j{<	xtp#VB4V_!ov(KeHI7{QO&O5O`1xHlR
z{4Fr*6fVlf7=uGzefB;Amp4z5%DixvO?IwW>D2#)b7+72HhgtcZOk<XPrJ+^qw?VK
zbwHbAV9nUfjLtlsN7CjY#YKQ!D-x_{py!;abMOahL0TBkkuErledf2PeT!UfCm7$J
zy6<IGy_|>@tnv^3Hzz^c#1+%~Z!{cTT1mA!3DF|k1i%2~!*(i6*JM|WHnz^)2MNOC
z+x~7sEi5jS4W=ULRi<3`^39g_-J!G8D<xG`J?qKnUV<V1Hs!v{|HiiCH_RHFh+BOX
zptRz0&nm_C6r8hd^$t0ZugZq&ZSM3m)oDSw6gigYhEH|_%rf|;=O06qAJI_w)l7Zj
z<IVe1TemwNpTO_#ry(He#XVc=v!#BIS^fK8s#pN=Rqc%91sSQ)8%Ly)aE+c7UB{Vn
zyJMXl!h0|6LW=srf~>1>xLy@?*&jtc*l<DXDm1jMQh4G+q$^2vDAt{GkW8ufu>QjA
z?c5(1Sy77Bloyx^gw$Uw%bJ+Sv<MqVM3Zj&ZKfTo5VekZM_6Y5J=4|~B5fxyjDVmv
z6E3DlO?IAD|4$2W?jc`P_R@_&daa+~$Fd_x^gwL}@g-9tLQG{!s;XNv^_?3VMfCJQ
z_JD}Y_zkGx>jMQlv*SSu@4ja;?+FXwK#<2uVhr#HC?8BUVE~ji@IXuQoLv%3j!?<*
z@aK>;Zry2v<?7x+nc-6YKr-`F&JjNzVcEqYJqO5xwJ&<Q->ed1B;d_L*<1D&?i;2;
z*8l~iJJPE9EYNqQff*VU=&Uhu9NFXhxT0;h{xK0r8tz`;?;KzJXnM01(jBd@XJ}%U
z*rO1i^_wJ)k+qc5<??pK=0^dpPUtAUQzXqC54b)(T<dc~^U<VQOLiu`(66y29PKDG
zG?<({7;4g%6d}QoZP(kDnyqSyq1%ZKu`_N!n{vMMFz+2&XwI;^`K=2HA}{hkj+uoo
zyP5&4%Sa718f9xLrFYOm#4|p{v-<jEM`w~nwo2xj>d~DKFB;CRS<8|G{~FY?7OfMH
ze3Z(yT}%$y@SIC2lmOuqh(&gHcU!zN^v0UNxdcXRby_0XU6e}&(1^=(cbEg9PQnCK
z2^;qMo*3WJ{;{$T24%nmVHt}K`wJN9z`qGrTx_s{0yI<6(lR$rF{+&zw3U^{j{$p-
zy6O6y#dHb7*B8HFRpeSG-mWD2`uY<gZYS(M?$N5_bX(`SM@(N-`-5p9_Dvsg3lOK@
z^~wMeAqWYudFw^+&aoF&5_P=le%hIzezy}8$%Upx=MhG6;o@sw_V~{A1A@qIn7X;Y
z6E|HK>DKd3c*E5mZm*xCn7_f6RW;2W`jA#!TA-N3+I4H#vYP*7H}cL0DNDX2onz0`
zY&FJcN}8pAgI<24+OBRpTb>is&0Wpw>L9x(+&wfR^ddin=fBl{rbLiy_YU_oiW4!`
z=EOB@Yol-F94?YX5~`1;D@ui*mKy4Ut&l$YQ2UAK$U?~pBQ<gE%vN;HU7CpU*=zV0
z>sEao9wUDVIK7`|_vfZdI9Gip-44T^iV-+0TL~H8smC7EyS(SO6U@}|PH_Ivufg+(
z9hNXO?R{ra+5|csR@*G?pMWm4ZDk*FykM55&~U1;rRd~@02I^~%3?K;TY>N4`>=ko
zTFi1Y%uoW!acX!umeF##u$X>WQFq)0a%-}&^L82u7;-{0dY0}M#NK}nu=6*J&p9Ec
zno9iy!fJr5LCOwzI~N9j9^is&#~XWN(JasAiWUfv`sQYO!v%7V4xS@r{QvcY@6jTb
zIE9LwHnqHek^X|F<;~`1aTAc4fcaWjysSy*kLWCi6&v4NTW+=OjL`3w`chVo{<*Nf
z<WcM1m`>?lS(?)KP6|Qp3pMSDkJ-)BJry392r-Tv*XyJ8-oP3a6D@gINjmuCv2S#?
z^?2_ytt=IGn8TrcD~UM?r$k`=o4gZo;)Xl?4UIomuJ*yRJbzq`*}X5ZwnpP@Uutlv
zC_HD|w2EAN2W_zC2hA0XsEOoV!>cP~h4vp?mZw?Zyc(WnleANxEb_W#of*h`HbGIT
zSGw(P3oS^4dN<9Xh{L8AyWc}EKAgagHtA|O^#aM%Zg+AlQHwOdC&=&dFV9AJTUA@a
zI7Mc1zyvB%b#=AyA(P{s?r+~v(cX|9Asv?7-H3eKzkSsAX&ad<zYpR7>5iL|T(VBT
zYHuRBO|N3_yBk#~Ah}oDSnL?n)k2uQ&RqlxJ=*mNFtpsRC%n{!fE()0`+M}4t#x%h
zfB*h%vw!-3s0NK3*z7inQYrWl9RD&3%J*;8y}i!`mCT!3z4=0c0-yBvI`|m5gud1f
zdWCx0w|2sX?5K@D2NO?aZvD|8e16B(8_1QxU$@X*i}@P5{U80>KC9;=GdNtlbyMs?
z%he<OT&=?1YEB}Yp>kOvVSKgxR!5U!?s9FCl-Jll0tMDgu3-W@^Xlb`zGFzrTbno`
zNdb&0a*a_K*IOvq-E}%VZ7J=zN?0zPp=siT7g3@uT~Q&c@b}4mV(Y#8i&_v>JmL7U
zdcPhkjxizR=B6IBH}%IyL6Xd&9aB@l;=GffZHMi+ZA|TRZe+Vs|ADNWTwEPvV{u<d
zbYROR5e^81KrdAhFg~rr!N3*TwfNleK?T9mE<a2{5TC_Kw!zXIA|pD~#PO+SjzUjE
zD)<j#mt0j95M9T$H|A|-rQ;zJabO^9>OYsn@vwjDct@1rOiNdH)B2z9u!}Qbc(+Jt
z*ooIu&tr&SF}Jq&!<yvDN+VngZt~$fXVM{$Rz%J$*Rmbw$@3X;9*G>1=hHnAI&N<y
z8n3V^E*j2Ld3qV8et0gHqOT=&<l!J={p)h8<p7Q0iSfiTPPvAR**~qgQ5Wqa=UAr7
zi43Re;QBcU2O<Fx*2Q|*ki+`Jv)XP@^|?630kU3ktE7@nn5gZ{h`_qDZOSG+$9qEd
z*1jRGi<@g&)r!~ceQ4r+ieFCzc{P&nxkjr?DxL~)`2N-mXwG^>9{ydArhExf+BEgH
zJFZ}Zn|$1gg#Vl#ADCeNDu46n=Y9YFJ=7D3&jn2(;79Z^y}~tfXAQti|M+n`>ZJs=
z{I;xivD%P}E{0lIx1_YZ%&pL9eS7`b`w|&61<2K&mEgUs`%ylvG$5@tB&)!l+dkUH
zsUh*bReR<pB&;Aw_)NEX<+$A+(=+<<|KX#p6wGl;{9pL*rs?<y)z8`Rn&R2nn?!sj
z*wA%7qP0s{dQpEhCWuCZFG2gG`J7hd7~dBwuK)L{`zP-vqVQM5Z=r;rj<JIkWM{yW
z{+ra(gVRTxf>c)8k?43)nt4IaAEb$C$xrjirpoZQS){8z(L9>R=#&y67ZagwpR+Nv
zCHt%VK!A^htM<<G*Jp|`)El^H`)1kJCuxit*}TQuHp3o`Z>vYz^lXL&GL$KJ6*E(G
ztqlO~adfOG&0z3i>w0`uY<d@G>`~6b*DF3=UV6=Z6=s0@cYd{_3B#JAXc-3lAJ+mY
zPVQSIGBoFc2Z%;=68mu%DfrrmaQ66RksD6>M<-sxP0B)Z6gwLbM7ETEpEJ(1x24d4
zld-cfuszX2ifB0VGjGxOx{%lLN3}x9!cjne;pl_x`#)$JwT8g1c9|F9p&Rn;93W!u
z1Db!7jW}`t-J<-`V-uevTThZI&Wpev>DX2P`qLiDg+D$&%l<@7uD*+<>vqs+_I^)#
zIcNS^z%^+5e0n}xYl6H^4RX?B%;nKOrU=s3&RcP#F3TUy&QXH%z7&<SX3lmI)s07O
z>7|R{IotIRauXBYsAl+%$O>QL*WFhr4P8-D|1e}O4xN5?MfrH5!Nf#1@0byqQ~6wY
zgS6Yk+Na5exGv@@1#a~D!K|HTt^i(Qi6lG%A%aZ3(mww5<|&(-^;Wgh^aLTfdMu-z
z4R@h|24k)6lZrreh@rX$xN|P7JbCh@Gm7rT@1&<#Xhbn?p0o5|lyx+M2=iWz1W`V}
zo_n`%-uPY9$>fQa7)ZO!j~De4+Qzx@%7?)p?yJX{N{<dK?HVNJ{7d(<%LMIE7o?Cm
zL2h*RLw|68n3zb!T~D8`OOauKv7Ae2dTAF@Kw!9!=NOY)eidLy^N+OuKRE8(;*EyB
z4%+qD{T1<<)_D`_Lcvk?mf#n(cOe6M{tZ%8QE&e1+(KCiBTjvwWCk6b4LfV}Q{L{^
z`seO;x<3A%_RW>nZ9g`gUq(c_pyQXmA3nv{G>e(GZ|a>1Bdy;v_iZn{+9CaUA<P{q
zciO{xYg@nbI^Yjrd4m}K!FBPz<9t?|h2ITPyr&X(x{z$xhr50Oq@G>RsSb?!WJl^q
zU1)#yGg9s?-n&I0@I>NVCQ&Dsk?o#GWFvA7m=BXP+3hm|BX}kNk^`T9dQKfWl=s$l
z8vN~`d&pO5s2(d+FXa*xoTlpD2nlj0KeQtYVwDVyNh@H6-Ak*hmXLIV>Q@{nA(~l3
zODmdMsrlkX^O-C|%;y|w(5OJ7E<V%qu^Mk-njpg-1GAwoa}=I*KLo++P-FYrQtczh
zf-O(*MQE5Q83sTxxmv}b`h;%B`=NNc7IuayS&)|wcV}S%^ZtEF!UbfCISS}|<50JT
zG`w4#&m=N`<%1oW>p!R-(<vwD8J=pElwY5qOJAnNXizhC+-XR#s&W)|^(LB*L)T>W
zVtvN*OgCpBrK5Is#U_a%*><;`2B*Sf9WS(i>u6CjPPk5VnzLuF`9-FR*eEw`r?^GK
zz_$)s@g-x6#4)~V8qWu_)kfC@YFQY4RO5EPC}B2QE>zk4L*D%o6iz;8XEcs5A<~}3
zK7)CK$pE76TIP7V9pJYCyDxAXCKM5Ol&3apHX?mU+7}fv96k6L$ZJaVTjlsE|G>IU
zw?F>t55*zmv$VEADgWg@S37IqA~8-N=%>JMomT4mn)?F#20i6pStR5ozN&utPeC(I
z5<*B)2ainAacEc}yfOEhb#>y#<`gUe15C=29IE)K;-*UrY>u#Llcy&Dm>PZ(#KFf9
zPK%)*6YC=_V2`3FaA$CI`8;2Rnua>M7JN!Vs~3Bsu0>KOXsEr1SYutnUHQiTMJo~y
z4`1zYjOB33rTs?}UZBGp^jh1MI32Zy=_o>b-rK)%-ZyFNkbK0l+Cus@T415?NwM8%
z@4Hy-TImlFTmr5EtNZ5n5|SB&K4qzvR@T>i1M%U94ahK}B?(La*36Gf*l1_OHZ~L~
z$w|7V(HCB_Jz7?(%`l=;Mn&n#Sx!_e8L<S}S)ibUMYv;XzWGmp_bVKQ!Nf7GM12Nc
z*ufwA5OonvG2$Y8SRPqkR@Sz*1`1f?Xxp<3A0o*pO&_t7slh;A;BhyncvepyXqYm`
zuB}H_*d|G3iqaE6suGy^!DtKGKa_M6x;N#xUjaA&LN1;}S`QTwPOR-TJjRX%PB_tK
zt`5jsrgEQlK-PyRWP1Pb!5%iTq@jgHFru*^R>%X*4zsFe3lU*b#v%J@H@v;=20Lw0
zV*l6G`SKC5mKR%%{uyIS9@m$@6Xw$TAuQi5%DvE1YkvN0;V%npoozkG&DFc&zh+&#
z{n+JD<!BzP)xy4%g-FAbQ(3R{`@1rU^5=UZ3*S>O&8eHtsB`tNZW=#|G>-jVB;po4
zy1Elp>Xk70nxpK*;@xh{BTj8wHM>~`7ZVI$#sJeq38^6IA2Lxg*%r0a&<lXJ*CuQz
zntULZO$ctv@jeS=`@4blGtuBj4JCd1AYcd10ubSj1*~w77=<jnPx&g_kA=KG_0OO=
zv(Zzbr>u}u7x;~Oir))r2hpV4#lbkNx*>yj!MH(G`U9wG+`#4qmdBib7Hidx6ZOq{
z2h6F1D81$qKPar3R!7$lgdu)`8M<3YIISovd(Ka}wVvr$QrK`@%ZWj3Q?0@bCN->^
zH{e5p<ww9$J5GZ>5$K!QMpsz)nL610_irA(HRdx|iuHXji!snch}yCo<uR{UR8_r4
zU%V~=C@w8DNdy~}bGQHh9vl62np)-Yad##w2Sv0BWXD;`(mTG9?1dD4R6r>Un}3Au
z$v3uZb0?>%PrNU0Pa&|`xVT4DOXjZLmJ@bM+ne7W$C2dLT{6F%);Bviq+-7s*jCh<
zzyDa^`*ANJYW%@rDg62Z3IFle$ePqOvQ1ZLv|hbk=*2b?7YE)K4&RW4%ApeGjq571
z;$iNM*=>Y(M}k-K2eLVLOGd34>jFMk*;M1qc>Wu-<RvZNz`qvgcp$K=@mXmZAKrsO
zZ(OSGr#cWnm;u6Th;{niE}5wa9UkaHV9%5zs1o!%Rv$7vdQ|UoAuxaZjjzsSJ<iz*
z@Y6%R>k!$bn6ILtt7{5&vB(eDgDel0v02cE-Uy?l+K2taum1#u2PkBC;7$TrvFk=D
zi-h`Cu_`Mu4>q81nL*fHn2THun{<!U+LU`r*1(YGCxcyxwa~bH;8E|kYoxEX@O~lb
zzdogo|9o)tXs*!Vk4&(=kCUs7PN}{zc+RLJV_-=W1w|%A5P_2j7F|Fzl2q|zMCW<8
z$oJEW_g^l5cPn~geD^D7bZxofb>h(NAjMJA(H0-&z7d6mTp7r)BF_~*armHxtIydM
zD{^(q&TFOdcU)Y#459t~Yj`UA=+R5Y{5Acj7nmPp<kQzcmvMCWzVu5K;!VR_+g`V~
zZ#fi=G9bz?WLBEBinpeG#-}Wjb9T|baiwO?eph5}4bS*HQm|dwazYa#CAVj>d*!xe
zeLQ10=+Dc@`UidJZg!g}jw)1j*tra@0*tqijY#ngWF8Qw<0Z_I)EfylNj_msnCEZa
zfWFcd2XsZy&}F;$xrLVr-ueCNd7dWF0~Xv<;|lYI&LOG9IOvmCAZakMGXzy__$p>S
zzn{=SeZSo{lJmA^4$<YKj4GznSr8Pu2M;Q<Af*4V6O84QRQ#N<Du?vG+?Q5U*ZvUh
ze#Dc!rE4DO8>6vKB>5n<ZwT+SZa1~1Jwg8Jc3|x2)>1e|ZR-&UsmPI2(M+AJOJQ4a
z-#pzH8Mr5T8JWjh4s1VIg{xl;#ov48akIN7BpZn;{#Gr|t<1>#Gxs-r&$dr!uKpfD
zgBd@v!JU%g;tW1oIkC_q3tKXanD*|J_$TRh@?%z&*31o^2^Z2v^4llMM1dV+eAwvn
zkGanAHRg}ef`Wv_%l9zSGFBD+-rZB-kudw$ny%v><F-LJSS?J=++M*N<|Y)e#DDX~
zJ7K>A`g@~5`6ZbqGIU2!0SFE`Q}{H(zv*aBuyQO9lery`Shdaz&(M1e%-twMnZDbD
zD=Coy+8(T%zrzy-S3r;1I0HLDwB}VZ{`Rg_i$h6)ycXc+U~rUQmIiT%DfVd1W;Y@6
zvuqJjB1mV2ffLcl_BzAQ;u&ta)BmBmjp+fijpGK0n%8nhkCR>Ji@Aii1b0ZyhiYiY
z;(^BFN&q%F-KXmNM6a>M^u?=Kos^ly=pPhKhsE1F<@%@HHZ#6Gdk5NI&W~TX9k8@M
zuT;uRWZ8LtwSP5`AnzNU$)Qi1BA31@px(Nx4$nOWS?c?txSDm9?n%Wt<6MmGwD*k<
zsy6;5&i-FK$FtuC3EcgvRg{T$rr<aAZXhXm3Jo)}S}NNyaHNNny3o1(Deu01cex<Z
zYv}(yQy0=(BL}C$kCYYi`L^!hq`3XDcFj``fHa#H*6!5?cJHLaLSq~P;JRP|I7-9}
zTC|UYE1*C$@$pGPZJcv_HL$V|wBU}LkkahkwQxv<^K1ZX)_NF$w|qu<6F9SzFdpyL
z?z=&P2)x#Z&qJEplj$nJ!zdu|s&I5I$!i8bF-LkydFVAm4)xdFFd4C3y}-ygeznJj
z$3@;6cKH80*>13tT$SIEzU;!i^kbh3h7V$U!;r+55A=}@QTVWbL0Fi*KM(oBoT<cg
zd7Zl1HCDalT(ZRXj4SC@*bm)?Hh<UBH%FT^JBy^kgGs(SFKmj<pyR#@S;)uMyE1S6
z`pD;`^-oV2X}%pXll`BU{R2CPG?&;Eg_>gJWTz3KiltF}Z*{HwmwP4>bt$j^%oYFn
z|J8Kf@l^NkAD4#Ps3ggX!i~!AWRC{fo3bUFY#}2N=@2I)GEW^v2+7{ryOWHAW0XBI
zj?IblyH5A_*PnVkJe>3SyvKE2&)4hqrHF}@N+0EYp=k)CR~wpjLNz)t<iUa%TcyG~
zi~4Zua$|U}V>tFK9S|#G;`{2Tj@JMW*sc+Ia}0uayK&1p=RU>8o~fL7PT_@IE&~g<
zWSuvU3O;@MTv4GSojwclFulEnTKWe8pvAP?hVd-R2e6_0j?f%TRC@IVG}#s)iRwU8
zW4I3C(g8~#W@_`e1})5ZRGXXM|Flqx6}b|9CzBiw4xFDkK|;nY16-1=m|+$NnrXOu
z|0SrP3RO*iF|}|qfGHS^O;dViuWi{>{;#(5^Y}e9C4RomV|P6%VsE=#J#bN<VD20x
zQo8M53t0zs%P%4|SM7uOIqV4sOE3`nAx_7RU)eTQ;tqtZLG#Hh`as`;*qH=Txr8M;
zP0h7EOVai6!>PiJb>YAe(z;B2Tkux>p*s0dHR;Eb_j;qs$Ek3*x<5?CW=P)2F|*5j
zMO9@t0%?E+RzPUy8&D@A`WL>|Isf}KZtX2!ZOORMTWpw4BdZMh3SF~Ix^z?*DbDC&
zoX(o0g<FcsG$n_e5Ss0BeM9Rk=j29ouc-P^3^G(LbMK!fqh^G2LZ=)3n)Vcd2mQpQ
z)P<-sP1_^_>JS@n1;X-N1O01<&C`OsYKMPNcQWB32(oqGGld7Eoh>DA9aDbXG^gi7
zwG~GhQR{ITh!HAuLKIo_e_ummyBu%o`qPHr8m;z4f%!I9EVJal%;0;APE4RGKdDO|
zc7P<-qOiR1u_e7iZy3LTV6#Jx-^2rqF1_4MljFLKohI^Vdmo?eoO74kJXXL`J1#f(
z-<`~Eaxt3088h{Vca05OogN)6ewvfDX4|A>b#CI1UF<b`g9k-z2j8+B)y_;hdTc2K
zpEU>{KL1VM$XIE|&*5&K-_avboCYqF{%Fc_2p1k^TF#4S26%U}Nx?D6Tq-?qD@2|2
z)5}17!+XTMmpg-H86Tn3cTBg0J!&X7Vok@}T~1S9$C_i_{0F1E2Gx{R^6aNjOpd4A
z?{n~8z0(UqIZXc$61p__vt@>qy3k>nZSdsi@$?u&%C<|)(6bX!u3SIGfu(Qw8f!20
zfwcdf5fV{a+S}WOwxDG~z}Fw(0UiU;_r3PO+L}9hFFS#f>KN;en6Ej`7X4bs3yLb6
zW*AV$F~ZCjrj;D}p`*wj58PBW)4YELl!A{Th96`Wu47d#d4+YI{rw4a!?G`ErX^2-
z%(G^D#M!L=Njf-cLLtEIfC+ZaZY)cZa;zRp405eifW+^h=|Qj*_&X|6`+REIYj?Er
zdjYz3>uz$?`GXR?d?m+7Sx;kQV@Meb0Iz91nfhmIwgx8V=DUPFr+*wDdET&2C6Qda
zOX)@;;P$9UeWb6g2N7_(_N;9ltHQN?pFS3?7VbRpN1Su7t-f92$<oBnE#e#4-usKB
zz$gmrePD0}Gj&8iG&#tH$p*3LlYe<~5U5ewq)ME5eTFY?szv@RGqE+ktBoD7R^(1H
z>N=EZ6|4V&`z=%$JCrrX#UMYBt;f<zQ^gT**Yi=uolG!`!fy4Kxxh{VV>`<~EtWj&
zr_^m*zX$;M<oR})mjNlS2N0+YDo!W6imyE}X8xVdKTRS~DCi*>;-6!tPXHLp%Qn-Z
zlHL9YEOlbrh#!Q5R9h!xa|+dJNZ=_49;>izLFXDG&46T|1L-rkh=D|Nv~HfW3(SUJ
zJra)l*8uUjcvu`7Jz=-r*+**lCC*htAsQaVc`%ndCCM1YTU9_$^C-2NQ_g8!B3S>|
zHMFW2g7N)b=gtgsa9L0gd}$T}8-!8K{*gm>W+g*@J6z?)0i?5W+WR@h^>-ug9c~K0
zk-UxtQwK|O<4Z-XBOFU&kM;!Z2~jH1ZNh4DQ}#$b+6Zw4mk-Bu{#~FD0YEHTlD#k;
zXzZ-{{haJIE7f+mMkm^WEf^r0L1*J(D3Z8$js`Lp`1gtf;>Kf)Tc=bz+ho!OHS%sr
zdM_p;*WV2zfM?>{3!~0{kpc>BNeT#Kfcfr%WWd6Y5BAa!8uSQc+C(uL!7ebv2fS;h
zV_A^&xYmL-`niB0jS~32Y4}?st4cBaL9H&K&Fq`9FU)Q{1?S|t%Q4>4f`S*w<5Pep
zYzy8>W|#@_At$xY6+&<JP?%5*>^l5%->u4niaJL!kP;r%v}d-gl?>}??8E+&rLg;o
z*{cP|dI6Rb&u$K*|9m2L3+LbOXRA59N=kU0TCUooH6{6Ti;7w1@k`>WSuDPGGaTzN
zB{QLPgdJ}ccz7&Rd5$7thqP{5v)h7Ba<%}em&nTklYy!0mt*=Ijj5zj6tf4?`w!_$
zp4{T~F9{DVg&RHS7tQe1l&u1#`&41K?R{M@^nIG_wf0P1_K7h$VC`tmMn~~;=JtMk
z#R~VVl&oJn#1G4%Q<7kvqnBe~vo={g`Vkf`I1_e^|9hrYV0)S1$AR+^jDnhbYYAst
zTmhr4AQLV-J8SFPcL-_Hk0K#jm_?QQqDo)9R*9`Km<9*j_p(}be%5f&{#%+Acte2?
z!L}`J%*&V1P1sj4ha^Wpik@8CQ!iRRGq~hUn%d9$3pWa#B6*d9b+D=PMg!F&i(t7s
z{O?8OSC&oDvK&|;>z-!?`4^flmRGcQ)7pEmM-bxMjH|s)TZ%kgS>CofsjcB1{=zH+
zdf5N)#UGm#{dVvhtJ7I9MIahg_(yvw?6A-^T|{H`{|@@#G&S#)i(?<?=3}@@$5#D4
zaoS0ERhZyQp7}9yX$dpG$+Wducw_6&R`uZ5`OEjl6Ky)n@Q9Qh6<+y+(Y~z+hS%Kf
zK#zjW4Sa>vC!R6Vgqz<)CTpAka-nepK?M{pfFhoqI&Z=wQ`Hb(ZkNYuC5!<;N)GN+
zy9as+P_ux0J;+etf}AE7aMy0{LdhxFc>-}~6%28dWjQ9wQEF-BU^ONRvNd@$S^5Gc
z{9|zc`Ly|Y^(jK;+)2J;2h$W3r_EIrZyM}DV+H6TM@>Qv?)%>+DtkfbBj(+~^2qtB
zM&#KeaBFr8u5Rwl45>SuzAvHz$WWp@RZpWKWB3ePx#Zzu7b+mZ;5CdAYe>=cm^bEg
zzp530>G9fFXEVxD^><B#4%3Tly|4%d#1+Hw-2;WjYiZ4%RpWQLYK#r;ZW~MPnCmZB
zA%V=%^wag;y0IkUZ5joddFZd_7*H^??hE7(vc_0<SnFNq%rYL9=VY4c#5ZFKIcRPI
zyrce61`L;g6Xcr<%kKa747A^#FW29Gc1itfavs~P)E)|ZWFG<!F8srKh+K-zP24Rn
zSOQ@>$l=;sqRuC$q%_WLY-#?MTc$oI0XF5+@SOnyxawJ;@DDV($;KGu-UJ#8N-&@)
zZHkBMcB;UpY-~2Z<Nc=S-siwRGHOnHK4Pab1?%g#T*|*tavr7pi#SoZ_PX#T-6)7E
zpm<5n@_aRA3cKB}h07(&izp+UH!4EEFgUH+tu9)E{%3>Cs^?-nCwDB%>*w#Fh_+eY
zmb2Jv5q?l+N{6zwA@+RxRg`Uh??PLHqp_r^%O>ILH7OrR_9J_xkh_Rw`ROq_L;KDT
z@k6srg0~u6Z)w0ej@)fKGV{HF<aOrLc(<%BiP1ufi`#wf)P=T7@f~1O0h%@$`Ve@*
zth~LwFIx2>qJbZa?Xe?+z~!lnHqpMO%ZuL9a7sXn!V$p==Eoif`&)I=%lgIohBIK`
z?~u|8{s`!fdHvsb4|XlId^C({hhU?F3;+--<PAVCv8bdZ48#9LdFeCj<^RpF)s9Z(
zm*<w4jQPYq=;etgU>rL$G9s%A)Z<FgaWEuEMKP@@b16nd_-KBRi77VcrPDJq{?)&P
z9@?s?ask@sis$}W&fAW@Jq4!uOfZ_hvo7nRQ;gvs>e5Wh3KwSyEa3GL9ZFQ;ac7TE
ztGp?@UEI%xqxk&$goRtA=ZqjGRy|@I^~da(`p^kj0>f-P>qmC3#txVDKpgJhFu9`D
zZ5)mqF%+a0uTu(9T>t^|Lfd9z4qIGN!Jlpw0C9QHWv8qPGJzjM)G>Vcp4V6Jy?5H=
zi7y^+)99Oq0~Xe)&%Mi=k6(aq?|py&6tEeQ-~D+M;!{A(rS>5dGTxFk7^3f-9`#$j
zVnptz6MSEqNS&gk6%?|dvtk~@?ym+FHcr3aL86CKvY^qnZD!hHVmRJ@Mmw80SWsrA
z(=k&?Snldqx0NRfAIez#X`C|-Bpo%8!=J^XV(5CE5HCL5`8d4+l?OCDfi?i40KgWP
zc>?P0L0G^8!RtDFlphIv^|)sRE*Li%rl?+ren*xmgIt`bN!K*(1P@2RNu}P?L4tuR
zmv4zSP^nZu<(oMjDFL4!RMZb$Ttua$q`oJqA=b87qa%6U`w)=tZ0+aAtfb4#$!QOS
z9e8?(CI7z4ck1Lxut<mX>Ap*NF$XCSpzcUe`*&p7)oBOZH+Vd1&#NGM?Td{8W$^dk
zSH7jEXg3M+a#+7)4`E;&#F7CVl<PjI)DQdnUHuBd@F#Tf+GC$f2FrI#)EiH4;jHJY
z{NLgt7aqr^;!S&X8N<W#Afi@0BFh)I=v4Q0HuE@FQ`5&3HO~iW&=Seb`gQUcGu|~2
z=bcVQy6o4UHwl;>P`G_X)$=JSzQ+ueX~N~x*%`W?*zVE`{RX(UE}?Ru=t__jk-6;q
zRXkVj_c}0k(db`+l-ay)tssIjFG4mcbU^(+<WE>qY+M|$|GuQGtPBatuz<E|Z|Xi-
zx;|S2B6`p>LgWh44DPxCJ^lTR&#gAMgqLfEWKaMWcLZiRAUbEdlK$OD>~Y5{YTGbq
zX=%yFs>lm34I-kwrF+QtXDsOtwrhP3<rMrsek#w5`k_m*QDW1GNz}Vb(JG2EI!CMh
z%rvlb8+-EaF77zQ0;6Wbj~rh3yw<d1T>B`ZGD#>lCB60IAaio^BCY*ktwk0#{gQmj
z>^C;I?R*wx%O|nXL=-ee-=dn4kNV~|AfysL3q1LbEU(J4{p;N`Ct{B7Yd9pGY(i1y
z=7Jq07MWQef=k`+Ju_pE?0o#Nbs&}xV<Z6cr1#qTzz}^AP&M1;cKmBy2I=f$-fo7r
z@S;Fm1#~hYk<id^225oc08TZz^qVf$!0=p?2+SK-^^@VJP-f;Bz^gUk`{1))4s2tP
zwXC<oC@KdVG?x9-A6c3{l$xVwrsv-(;k&{;wA7Wk9*L_0!sTz$x4fqAyw(mLX=_)(
z_0;w{+{ceO%ALXT4u)G4=2*bn=(t_$Gbc2D<JFy<FkLs1jYi{|5Q;L}B7fqkg-UL}
z9TkBi4hdJMlM+>J?jY%WH8t<Kj$`7ky02#zzFQ>Ho(A>G-Mj(06%4X)>ZT@9=k~s~
z$@Lja1urHmsD|{=1xrV%mG8|808$GG5Db%U<SIb>3fy7OCxN7s+`oa;ZGc|HQ)j^T
zGKeL3#*?K<1pGya;%n4WYhXzOpm;L21OT#&^!NSzc9CIs$TmhGFeXg35odbPX{RZZ
z{u9PtYAjoxFkqa?p&m4KGn9NH`*ko8iMdhhx}!8h=cEFjkwLfbleym$u<1c~jd|47
zbGcWYKTT9?&+dvEFP_C4L%k0-+V8dFLDxcTV<w0=YDMn%cG(Nef-ALKM`7Xtr<MO?
z6>zX{o%oXN_q~dE_H%)}Inz7IBegd>8~eS2f@%wMv<wXavxXi1gQJrZ%?<H{3BbT7
zh_w^YAcTV!(&81l0B7fo?||^S4|PB1;WAFw1Y4%R<?$Sx+yO*&!3(FV)V;a`<IoS<
ztrztM9S|VW_2dkt-yVOdpz5yv-<~i<G&{GXEsp1kC&X#^tMJN=*5aq5B|N*kz3GD{
z0Tz`B8^4~jLxKs4+BX<cRBMi>e49;&xUL!D%`?~WUyLg~;=2jf^Wz6>T>$~_v$5!R
zaJA_TxR)#`hvNI09Vh&|g9T)p0CKHAhcIY!=T+^IEgX9LI2V($4G|g7HEH5cow(NA
zX;%!|fu@MlFy>%@nzpI%aItVZ-OL;ZhL%B^i<lx=`hVyIXjgn|fJ;QRL3?ww=5ajO
zLM?9jPlPk|Y(khM=tsdr-X=r8u#}{IWt7+Vue^c~Q~;pPx~IgUi~qhyph=F6=d?<K
z140Pz2>C4{n(*6h^Y<4%L{ixI`i@pV9}8Nds&0n}%!jgmUw1(h^E$7#;s)0>wLnDA
zkJ2URyc3e@w@B{4R)fn>2uSjV(Z@R-74b?E&)K~3B|^V`7}pJBn|*2}U`^J($*hWi
zC^*XzCPL7<UGPo;=N?M;{mlh)JU&MeberHdIUq4O>{JS<4nT79*IDLmN6D^uV7#K0
z&3&5>OtfHi&;Ci5^0NEsQThM#PShg#M!&m{qYWfP-PUWV3MlWmTDV)`HOIlteH$jf
z=M!phjq)E)=Z!h6><_m-Q=Ye2J^MP3ueKiUn&g+27EKc9l5mSp%ShO3PS(g;9eQFi
zm~KxK06nodRkr7sGED(w5UyFH1W*AU4=C&SwdZekKBmNQ;7!#njEr9KEE}6yg(vfS
z7axM#$<WZyPw+2-OAr+OJW;i|$wbZnBtU;$0t+(aFHG(PE3|QvJTk(MKQrl>MUysk
z?d&r@Iw{W1&2Bs%zAfv(`E1+KN18`0GpNvb^lo}+(cFfI0b_VMY2Nu5*O@qL>tmQR
z@P#+rcTZaWp^dNF*N`)6%s219dw@Sn1&`s%4ai+C*mPmIlcrtug4`-tD;{$N4dNR#
zIv9gN%u_Pa90vn*02skM$f=YSs9WtP>+$PSGUdcR6~zC<U<%_{LueM>BRScjO7l}|
zm$v)?RD9^_y4~`1S$Uqlt-;B6y556*y(<s?pwOcEy0#JAn{n4ZSJt6KrA&jYfrsP)
zJ+hl$IB0?PV@efY<sV1$duabi_wP!hd&L%9u=ByAIt41wFT6G9jm3Hj2I_$65TPSu
zdp^I1GR`|>Gmbcwf~DOn+g7k80XJ%)0vW?A=htPc{yWn>D$se@hkfp^w#vs`fji7{
zgcv`<O5bh6?t)Aw)}sD@mUcK9IMnFJW55Q;#BXmR==H8cin|z6f|*B3^!c#sS@mey
zShhb2Q^PK-8HpW$euFRq-cdy^(N~W^<vGk=UsQA{pvXtes_Tml4;B_^NI_H$ZfS<>
zDO>$>MM;5v<*3kq8q(9E4I{L0!93-Se3x%Ax;JjBktVI$=3wPBLMH6<^wUq3IX)<p
zd6Un?hGH_R@_E)>B&Ah*=#iu%$KCLQB)<IsFAlmLNYiwrFyUt;R}?Ve($GkN`5*{a
zJ;34WC*c@bhzEc?1hfqez1bXRWn<%Jq!Cq65Z>A96VjB&mUI1uEeO7}NJYx0V0pRW
zH$|ujtmzZ`n4}gq=`oqy7QegtWEqP@MTZ`(bANGzXPs2;OOmF(t;^mPoIj&xbpPTh
zhFjrpLt5Ub_)$}#jWyGpKYL*G_}E_xFgWlCC5ZGpJg7v@5FdvH#BK2HY#X3eF)QwG
zSKMD2bACKmZ~viT#G>KV6Hr`=pL_n!3}=`(pjT-I28TbqrC+ZSJ=v9S(zi2%4^98-
zj)bPcU2env3x?uTM=5see^?Gcz+8*RC;!?4wGWtVe82^CdZ!G6#%+aBK{i*0eSUp^
zvCS%Ve=TCfX)01gegBD*c&cNOk%K(i8Nmv&EXPVSGC0MY=~P-ZT-buoNpA_2Frs+n
zW>M;+w42VOy2)BUFPJj_F$ja&@!dN=*RLz{+w6>;kI#~lEnY!HOQP0ZE+spD+(m$b
zqGj@N25Wj5Yx;Io!9IJWlF)HZrE$H}6m-~rLss~IW;5NVyM6aVI!;iu=e0{6B8=|;
zprly8;u)iGlIcwe#YIw$TF}_)t_(*T1yeNTjk?w;f}-BV%fArY-$LHc3FipAF|-8E
Ra8kfORV6LOf?M}P{|`xh17-jK
literal 0
HcmV?d00001
diff --git a/milena/doc/img/src/genericity.svg b/milena/doc/img/src/genericity.svg
new file mode 100644
index 0000000..ef8b914
--- /dev/null
+++ b/milena/doc/img/src/genericity.svg
@@ -0,0 +1,354 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!-- Created with Inkscape (http://www.inkscape.org/) -->
+
+<svg
+ xmlns:dc="http://purl.org/dc/elements/1.1/"
+ xmlns:cc="http://creativecommons.org/ns#"
+ xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+ xmlns:svg="http://www.w3.org/2000/svg"
+ xmlns="http://www.w3.org/2000/svg"
+ xmlns:xlink="http://www.w3.org/1999/xlink"
+ xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
+ xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"
+ width="2099.5906"
+ height="1353.4995"
+ id="svg2"
+ version="1.1"
+ inkscape:version="0.47 r22583"
+ sodipodi:docname="genericity.svg">
+ <defs
+ id="defs4">
+ <marker
+ inkscape:stockid="Arrow1Lend"
+ orient="auto"
+ refY="0"
+ refX="0"
+ id="Arrow1Lend"
+ style="overflow:visible">
+ <path
+ id="path4297"
+ d="M 0,0 5,-5 -12.5,0 5,5 0,0 z"
+ style="fill-rule:evenodd;stroke:#000000;stroke-width:1pt;marker-start:none"
+ transform="matrix(-0.8,0,0,-0.8,-10,0)" />
+ </marker>
+ <marker
+ inkscape:stockid="Arrow1Lstart"
+ orient="auto"
+ refY="0"
+ refX="0"
+ id="Arrow1Lstart"
+ style="overflow:visible">
+ <path
+ id="path4294"
+ d="M 0,0 5,-5 -12.5,0 5,5 0,0 z"
+ style="fill-rule:evenodd;stroke:#000000;stroke-width:1pt;marker-start:none"
+ transform="matrix(0.8,0,0,0.8,10,0)" />
+ </marker>
+ <inkscape:perspective
+ sodipodi:type="inkscape:persp3d"
+ inkscape:vp_x="0 : 526.18109 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_z="744.09448 : 526.18109 : 1"
+ inkscape:persp3d-origin="372.04724 : 350.78739 : 1"
+ id="perspective10" />
+ <inkscape:perspective
+ id="perspective2956"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective3030"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective3104"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective3253"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective3363"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective3437"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ </defs>
+ <sodipodi:namedview
+ id="base"
+ pagecolor="#ffffff"
+ bordercolor="#666666"
+ borderopacity="1.0"
+ inkscape:pageopacity="0.0"
+ inkscape:pageshadow="2"
+ inkscape:zoom="0.32014852"
+ inkscape:cx="1160.7884"
+ inkscape:cy="650.81386"
+ inkscape:document-units="px"
+ inkscape:current-layer="layer1"
+ showgrid="false"
+ inkscape:window-width="1592"
+ inkscape:window-height="1149"
+ inkscape:window-x="0"
+ inkscape:window-y="0"
+ inkscape:window-maximized="1" />
+ <metadata
+ id="metadata7">
+ <rdf:RDF>
+ <cc:Work
+ rdf:about="">
+ <dc:format>image/svg+xml</dc:format>
+ <dc:type
+ rdf:resource="http://purl.org/dc/dcmitype/StillImage" />
+ <dc:title></dc:title>
+ </cc:Work>
+ </rdf:RDF>
+ </metadata>
+ <g
+ inkscape:label="Calque 1"
+ inkscape:groupmode="layer"
+ id="layer1"
+ transform="translate(596.74498,224.88271)">
+ <rect
+ style="fill:#fff6d5;stroke:#000000;stroke-width:4;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none"
+ id="rect2885"
+ width="893.33527"
+ height="409.18506"
+ x="602.84521"
+ y="419.55798"
+ transform="translate(-596.74498,-224.88271)"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <g
+ id="g3451"
+ transform="translate(-318.60213,374.82603)"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75">
+ <image
+ xlink:href="file:///home/lazzara/svn/lrde-publis/trunk/lrde-a2si/levillain.10.icip/figs/bunny-holefilled.png"
+ width="450"
+ height="450"
+ id="image2958"
+ x="-278.14285"
+ y="-138.3521" />
+ <image
+ xlink:href="file:///home/lazzara/svn/lrde-publis/trunk/lrde-a2si/levillain.10.icip/figs/classgradi.png"
+ width="454.28574"
+ height="448.57144"
+ id="image3106"
+ x="-278.14285"
+ y="305.21933" />
+ <image
+ xlink:href="file:///home/lazzara/svn/lrde-publis/trunk/lrde-a2si/levillain.10.icip/figs/lena.png"
+ width="452"
+ height="460.57141"
+ id="image3365"
+ x="-278.14285"
+ y="-599.70874" />
+ </g>
+ <g
+ id="g3456"
+ transform="translate(371.7026,371.07634)"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75">
+ <image
+ xlink:href="file:///home/lazzara/svn/lrde-publis/trunk/lrde-a2si/levillain.10.icip/figs/bunny-holefilled-complex-pinv-curv-area-500-segm.png"
+ width="450"
+ height="450"
+ id="image3032"
+ x="681.14288"
+ y="-144.06639" />
+ <image
+ xlink:href="file:///home/lazzara/svn/lrde-publis/trunk/lrde-a2si/levillain.10.icip/figs/classwst2b.png"
+ width="448.57144"
+ height="451.42856"
+ id="image3255"
+ x="681.14288"
+ y="305.21933" />
+ <image
+ xlink:href="file:///home/lazzara/svn/lrde-publis/trunk/lrde-a2si/levillain.10.icip/figs/lena-s.png"
+ width="449.14285"
+ height="449.14285"
+ id="image3439"
+ x="681.14288"
+ y="-595.06647" />
+ </g>
+ <g
+ id="g3471"
+ transform="translate(-74.310365,74.19173)"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75">
+ <rect
+ y="271.47461"
+ x="109.32426"
+ height="212.40141"
+ width="368.57892"
+ id="rect3463"
+ style="fill:#f9f9f9;stroke:#000000;stroke-width:0.96799999;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none" />
+ <text
+ id="text3465"
+ y="355.81046"
+ x="299.86081"
+ style="font-size:40px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ xml:space="preserve"><tspan
+ style="font-size:64px;text-align:center;text-anchor:middle"
+ y="355.81046"
+ x="299.86081"
+ id="tspan3467"
+ sodipodi:role="line">Area</tspan><tspan
+ style="font-size:64px;text-align:center;text-anchor:middle"
+ id="tspan3469"
+ y="435.81046"
+ x="299.86081"
+ sodipodi:role="line">Closing</tspan></text>
+ </g>
+ <g
+ transform="translate(390.90885,74.19173)"
+ id="g3477"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75">
+ <rect
+ style="fill:#f9f9f9;stroke:#000000;stroke-width:0.96799999;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none"
+ id="rect3479"
+ width="368.57892"
+ height="212.40141"
+ x="109.32426"
+ y="271.47461" />
+ <text
+ xml:space="preserve"
+ style="font-size:40px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ x="299.86081"
+ y="355.81046"
+ id="text3481"><tspan
+ sodipodi:role="line"
+ id="tspan3483"
+ x="299.86081"
+ y="355.81046"
+ style="font-size:64px;text-align:center;text-anchor:middle">Watershed</tspan><tspan
+ sodipodi:role="line"
+ x="299.86081"
+ y="435.81046"
+ id="tspan3485"
+ style="font-size:64px;text-align:center;text-anchor:middle">Flooding</tspan></text>
+ </g>
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;stroke-miterlimit:4;stroke-dasharray:none;marker-end:url(#Arrow1Lend)"
+ d="m -150.28848,868.07272 97.529943,0"
+ id="path3489"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;display:inline;stroke-miterlimit:4;stroke-dasharray:none"
+ d="m -53.008405,868.20786 0,-417.43053"
+ id="path3497"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;marker-end:url(#Arrow1Lend);display:inline"
+ d="m -146.72366,447.18288 176.387083,0"
+ id="path3499"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;stroke-miterlimit:4;stroke-dasharray:none;marker-end:url(#Arrow1Lend)"
+ d="m -145.2169,33.058748 94.965307,1.096259"
+ id="path3503"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;stroke-miterlimit:4;stroke-dasharray:none"
+ d="M -50.802426,34.161751 -53.005703,453.80623"
+ id="path3505"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;stroke-miterlimit:4;stroke-dasharray:none;marker-end:url(#Arrow1Lend)"
+ d="m 399.54619,447.18288 99.84184,0"
+ id="path3507"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;marker-end:url(#Arrow1Lend);stroke-miterlimit:4;stroke-dasharray:none"
+ d="m 870.39821,451.60025 182.96699,0"
+ id="path3511"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;stroke-miterlimit:4;stroke-dasharray:none"
+ d="m 971.82078,915.97919 0,-934.831771"
+ id="path3513"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;marker-end:url(#Arrow1Lend);stroke-miterlimit:4;stroke-dasharray:none"
+ d="m 971.46519,915.4238 93.19981,0"
+ id="path3515"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <path
+ style="fill:none;stroke:#000000;stroke-width:4;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;stroke-miterlimit:4;stroke-dasharray:none;marker-end:url(#Arrow1Lend)"
+ d="m 971.65505,-18.849357 80.69205,0"
+ id="path3517"
+ inkscape:connector-type="polyline"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75" />
+ <text
+ xml:space="preserve"
+ style="font-size:40px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ x="181.01906"
+ y="285.25821"
+ id="text2887"
+ inkscape:export-filename="/home/lazzara/git/oln/milena/doc/img/genericity.png"
+ inkscape:export-xdpi="75"
+ inkscape:export-ydpi="75"><tspan
+ sodipodi:role="line"
+ id="tspan2889"
+ x="181.01906"
+ y="285.25821"
+ style="font-size:64px">Generic workflow</tspan></text>
+ </g>
+</svg>
diff --git a/milena/doc/mainpage.dox b/milena/doc/mainpage.dox
index df9eac0..cf9bec8 100644
--- a/milena/doc/mainpage.dox
+++ b/milena/doc/mainpage.dox
@@ -7,6 +7,11 @@ image processing as well as algorithms, in particular in the field of
mathematical morphology. It is used as the base image processing
library in all Olena modules.
+<center><h2>Write once, run on many kinds of images.</h2></center>
+\image html genericity.png
+\image latex genericity.png
+
+
\subsection mlnimagestruct Images and Structures
Milena can be used on:
@@ -57,49 +62,51 @@ image.
\section mln_sec Overview of Milena.
- <UL>
- <LI> \ref mln
- <LI> \ref mln::accu
- <LI> \ref mln::algebra
- <LI> \ref mln::arith
- <LI> \ref mln::binarization
- <LI> \ref mln::border
- <LI> \ref mln::canvas
- <LI> \ref mln::convert
- <LI> \ref mln::data
- <LI> \ref mln::debug
- <LI> \ref mln::display
- <LI> \ref mln::draw
- <LI> \ref mln::estim
- <LI> \ref mln::extension
- <LI> \ref mln::fun
- <LI> \ref mln::geom
- <LI> \ref mln::graph
- <LI> \ref mln::histo
- <LI> \ref mln::io
- <LI> \ref mln::labeling
- <LI> \ref mln::data
- <LI> \ref mln::linear
- <LI> \ref mln::literal
- <LI> \ref mln::logical
- <LI> \ref mln::make
- <LI> \ref mln::math
- <LI> \ref mln::metal
- <LI> \ref mln::morpho
- <LI> \ref mln::norm
- <LI> \ref mln::opt
- <LI> \ref mln::pw
- <LI> \ref mln::registration
- <LI> \ref mln::set
- <LI> \ref mln::tag
- <LI> \ref mln::test
- <LI> \ref mln::topo
- <LI> \ref mln::trace
- <LI> \ref mln::trait
- <LI> \ref mln::transform
- <LI> \ref mln::util
- <LI> \ref mln::value
- <LI> \ref mln::win
+ Here is a list of the different namespaces/categories of algorithms
+ provided in Milena.
+
+ \li \ref mln
+ \li \ref mln::accu
+ \li \ref mln::algebra
+ \li \ref mln::arith
+ \li \ref mln::binarization
+ \li \ref mln::border
+ \li \ref mln::canvas
+ \li \ref mln::convert
+ \li \ref mln::data
+ \li \ref mln::debug
+ \li \ref mln::display
+ \li \ref mln::draw
+ \li \ref mln::estim
+ \li \ref mln::extension
+ \li \ref mln::fun
+ \li \ref mln::geom
+ \li \ref mln::graph
+ \li \ref mln::histo
+ \li \ref mln::io
+ \li \ref mln::labeling
+ \li \ref mln::data
+ \li \ref mln::linear
+ \li \ref mln::literal
+ \li \ref mln::logical
+ \li \ref mln::make
+ \li \ref mln::math
+ \li \ref mln::metal
+ \li \ref mln::morpho
+ \li \ref mln::norm
+ \li \ref mln::opt
+ \li \ref mln::pw
+ \li \ref mln::registration
+ \li \ref mln::set
+ \li \ref mln::tag
+ \li \ref mln::test
+ \li \ref mln::topo
+ \li \ref mln::trace
+ \li \ref mln::trait
+ \li \ref mln::transform
+ \li \ref mln::util
+ \li \ref mln::value
+ \li \ref mln::win
\section copyright Copyright and License.
Copyright (C) 2007, 2008, 2009, 2010, 2011 EPITA Research and
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-74-g83e82f6 Fix titles.
by Guillaume Lazzara
* doc/contributors.dox,
* doc/license.dox,
* doc/license_commercial.dox,
* doc/mainpage.dox: Here.
---
ChangeLog | 9 +++++++++
doc/contributors.dox | 2 +-
doc/license.dox | 2 +-
doc/license_commercial.dox | 2 +-
doc/mainpage.dox | 2 +-
5 files changed, 13 insertions(+), 4 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index b94441e..2c8eb1c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,14 @@
2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+ Fix titles.
+
+ * doc/contributors.dox,
+ * doc/license.dox,
+ * doc/license_commercial.dox,
+ * doc/mainpage.dox: Here.
+
+2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+
Add a description about Olena.
* doc/Doxyfile.in: Add IMAGE_PATH entry.
diff --git a/doc/contributors.dox b/doc/contributors.dox
index 82b35c6..93d761e 100644
--- a/doc/contributors.dox
+++ b/doc/contributors.dox
@@ -1,4 +1,4 @@
-/** \page Contributors List of contributors
+/** \page contributors Contributors
\htmlinclude contributors.html
diff --git a/doc/license.dox b/doc/license.dox
index eb9a8e4..55bb41d 100644
--- a/doc/license.dox
+++ b/doc/license.dox
@@ -1,4 +1,4 @@
-/*! \page gplv2 "GNU General Public License (GPL)"
+/*! \page gplv2 GNU General Public License (GPL)
Olena and its modules are Copyright (C) 2011 EPITA Research and
Development Laboratory (LRDE).
diff --git a/doc/license_commercial.dox b/doc/license_commercial.dox
index af2f094..05603fa 100644
--- a/doc/license_commercial.dox
+++ b/doc/license_commercial.dox
@@ -1,4 +1,4 @@
-/*! \page commercial "Commercial Use"
+/*! \page commercial Commercial Use
Even though Olena is released under GPL v2, using Olena in a close
source commercial products may be possible.
diff --git a/doc/mainpage.dox b/doc/mainpage.dox
index 4d72915..17b72cc 100644
--- a/doc/mainpage.dox
+++ b/doc/mainpage.dox
@@ -14,7 +14,7 @@
<td>
\li \ref about
\li <a class="el" href="http://www.lrde.epita.fr/cgi-bin/twiki/view/Olena/Publications">Bibliography</a>
- \li <a class="el" href="a00028.html">Contributors</a>
+ \li \ref contributors
</td>
<td>
\li \ref moduleslist
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-72-gcf3ec15 Add new documentation groups.
by Guillaume Lazzara
* doc/Doxyfile.in: Use .dox file.
* doc/Makefile.am: New dependencies.
* mln/core/concept/object.hh: Move doc content...
* doc/mainpage.dox: ... here.
* doc/groups/accu.hh,
* doc/groups/graph.hh,
* doc/groups/images.hh,
* doc/groups/main.hh,
* doc/groups/neighb.hh,
* doc/groups/site_set.hh,
* doc/groups/util.hh,
* doc/groups/win.hh: Rename ...
* doc/groups/accu.dox,
* doc/groups/graph.dox,
* doc/groups/images.dox,
* doc/groups/main.dox,
* doc/groups/neighb.dox,
* doc/groups/site_set.dox,
* doc/groups/util.dox,
* doc/groups/win.dox: ... as this.
---
milena/ChangeLog | 29 ++++++
milena/doc/Doxyfile.in | 5 +-
milena/doc/Makefile.am | 19 ++--
milena/doc/groups/{accu.hh => accu.dox} | 0
milena/doc/groups/{graph.hh => graph.dox} | 0
milena/doc/groups/{images.hh => images.dox} | 0
milena/doc/groups/{main.hh => main.dox} | 0
milena/doc/groups/{neighb.hh => neighb.dox} | 0
milena/doc/groups/{site_set.hh => site_set.dox} | 0
milena/doc/groups/{util.hh => util.dox} | 0
milena/doc/groups/{win.hh => win.dox} | 0
milena/doc/mainpage.dox | 121 +++++++++++++++++++++++
milena/mln/core/concept/object.hh | 73 --------------
13 files changed, 163 insertions(+), 84 deletions(-)
rename milena/doc/groups/{accu.hh => accu.dox} (100%)
rename milena/doc/groups/{graph.hh => graph.dox} (100%)
rename milena/doc/groups/{images.hh => images.dox} (100%)
rename milena/doc/groups/{main.hh => main.dox} (100%)
rename milena/doc/groups/{neighb.hh => neighb.dox} (100%)
rename milena/doc/groups/{site_set.hh => site_set.dox} (100%)
rename milena/doc/groups/{util.hh => util.dox} (100%)
rename milena/doc/groups/{win.hh => win.dox} (100%)
create mode 100644 milena/doc/mainpage.dox
diff --git a/milena/ChangeLog b/milena/ChangeLog
index 6a780fc..6aec8b7 100644
--- a/milena/ChangeLog
+++ b/milena/ChangeLog
@@ -1,3 +1,32 @@
+2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+
+ Add new documentation groups.
+
+ * doc/Doxyfile.in: Use .dox file.
+
+ * doc/Makefile.am: New dependencies.
+
+ * mln/core/concept/object.hh: Move doc content...
+ * doc/mainpage.dox: ... here.
+
+ * doc/groups/accu.hh,
+ * doc/groups/graph.hh,
+ * doc/groups/images.hh,
+ * doc/groups/main.hh,
+ * doc/groups/neighb.hh,
+ * doc/groups/site_set.hh,
+ * doc/groups/util.hh,
+ * doc/groups/win.hh: Rename ...
+
+ * doc/groups/accu.dox,
+ * doc/groups/graph.dox,
+ * doc/groups/images.dox,
+ * doc/groups/main.dox,
+ * doc/groups/neighb.dox,
+ * doc/groups/site_set.dox,
+ * doc/groups/util.dox,
+ * doc/groups/win.dox: ... as this.
+
2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
* mln/registration/icp.hh: Cite article.
diff --git a/milena/doc/Doxyfile.in b/milena/doc/Doxyfile.in
index 7790ce6..5cbaaef 100644
--- a/milena/doc/Doxyfile.in
+++ b/milena/doc/Doxyfile.in
@@ -84,7 +84,7 @@ FILE_VERSION_FILTER =
#---------------------------------------------------------------------------
QUIET = YES
WARNINGS = YES
-WARN_IF_UNDOCUMENTED = YES
+WARN_IF_UNDOCUMENTED = NO
WARN_IF_DOC_ERROR = YES
WARN_NO_PARAMDOC = NO
WARN_FORMAT = "$file:$line: $text"
@@ -96,7 +96,8 @@ INPUT = @top_srcdir@/milena
FILE_PATTERNS = *.cc \
*.hh \
*.hxx \
- *.hcc
+ *.hcc \
+ *.dox
RECURSIVE = YES
EXCLUDE = @top_srcdir@/milena/sandbox \
@top_srcdir@/milena/trash \
diff --git a/milena/doc/Makefile.am b/milena/doc/Makefile.am
index 99dd4c6..3437f58 100644
--- a/milena/doc/Makefile.am
+++ b/milena/doc/Makefile.am
@@ -693,14 +693,14 @@ regen-am:
EXTRA_DIST += \
$(DATA_html_dirs) \
$(DOXYFILE).in \
- groups/accu.hh \
- groups/graph.hh \
- groups/images.hh \
- groups/main.hh \
- groups/neighb.hh \
- groups/site_set.hh \
- groups/util.hh \
- groups/win.hh \
+ groups/accu.dox \
+ groups/graph.dox \
+ groups/images.dox \
+ groups/main.dox \
+ groups/neighb.dox \
+ groups/site_set.dox \
+ groups/util.dox \
+ groups/win.dox \
img/small-enlarged.png \
img/small.png \
doxyfuns.sty \
@@ -708,7 +708,8 @@ EXTRA_DIST += \
tools/clearbanner.sh \
tools/sample_utils.hh \
tools/split_sample.sh \
- tools/todoxygen.sh
+ tools/todoxygen.sh \
+ mainpage.dox
## ---------- ##
diff --git a/milena/doc/groups/accu.hh b/milena/doc/groups/accu.dox
similarity index 100%
rename from milena/doc/groups/accu.hh
rename to milena/doc/groups/accu.dox
diff --git a/milena/doc/groups/graph.hh b/milena/doc/groups/graph.dox
similarity index 100%
rename from milena/doc/groups/graph.hh
rename to milena/doc/groups/graph.dox
diff --git a/milena/doc/groups/images.hh b/milena/doc/groups/images.dox
similarity index 100%
rename from milena/doc/groups/images.hh
rename to milena/doc/groups/images.dox
diff --git a/milena/doc/groups/main.hh b/milena/doc/groups/main.dox
similarity index 100%
rename from milena/doc/groups/main.hh
rename to milena/doc/groups/main.dox
diff --git a/milena/doc/groups/neighb.hh b/milena/doc/groups/neighb.dox
similarity index 100%
rename from milena/doc/groups/neighb.hh
rename to milena/doc/groups/neighb.dox
diff --git a/milena/doc/groups/site_set.hh b/milena/doc/groups/site_set.dox
similarity index 100%
rename from milena/doc/groups/site_set.hh
rename to milena/doc/groups/site_set.dox
diff --git a/milena/doc/groups/util.hh b/milena/doc/groups/util.dox
similarity index 100%
rename from milena/doc/groups/util.hh
rename to milena/doc/groups/util.dox
diff --git a/milena/doc/groups/win.hh b/milena/doc/groups/win.dox
similarity index 100%
rename from milena/doc/groups/win.hh
rename to milena/doc/groups/win.dox
diff --git a/milena/doc/mainpage.dox b/milena/doc/mainpage.dox
new file mode 100644
index 0000000..df9eac0
--- /dev/null
+++ b/milena/doc/mainpage.dox
@@ -0,0 +1,121 @@
+/** \mainpage Documentation of milena
+
+ \section intro_sec Introduction
+
+Milena is a generic C++ library, providing many data structures for
+image processing as well as algorithms, in particular in the field of
+mathematical morphology. It is used as the base image processing
+library in all Olena modules.
+
+\subsection mlnimagestruct Images and Structures
+
+Milena can be used on:
+
+\li images based on regular grids (1D, 2D, 3D, ...),
+\li images based on (undirected) graphs,
+\li images based on complexes (http://en.wikipedia.org/wiki/Simplicial_complex),
+\li any subset of the previous image types,
+\li any of the previous images modified by a geometrical/topological transformation.
+
+Core concepts of morphology and topology like adjacency and (resp.)
+structuring elements are represented by neighborhood and window
+(resp.) objects. Milena provides classical neighborhoods and windows
+(e.g., for 4- and 8-connexity on regular 2D grids, adjacent vertices
+on graphs, etc.), but users can define their own objects, and use them
+seamlessly with existing or new algorithms.
+
+\subsection mlnvalues Values
+
+Milena supports many different value types and may therefore be used
+with virtually any real-world image inputs and outputs, as well as new
+and original image types.
+
+Currently handled values types are:
+
+\li Boolean (binary) values,
+\li n-bit integers (with fixed n),
+\li n-bit floating-point values (with fixed n),
+\li n-bit gray-level values (with fixed n),
+\li RGB, HSI, HSL color values,
+\li labels (with no arithmetic),
+\li fixed-size tuples and vectors of the previous value types.
+
+These value types can be freely used with any image type and any
+algorithm, provided the combination is valid. Users can provide
+missing definitions when they want to handle non-covered cases. For
+instance, they can write a definition of the supremum on the set of
+RGB values to compute the morphological dilation of an RGB color
+image.
+
+
+ \section quickstart Quick Start
+
+ If you are new at using Milena, take a look at the following pages :
+
+ \li \ref quickref
+ \li \ref tutorial
+
+ \section mln_sec Overview of Milena.
+
+ <UL>
+ <LI> \ref mln
+ <LI> \ref mln::accu
+ <LI> \ref mln::algebra
+ <LI> \ref mln::arith
+ <LI> \ref mln::binarization
+ <LI> \ref mln::border
+ <LI> \ref mln::canvas
+ <LI> \ref mln::convert
+ <LI> \ref mln::data
+ <LI> \ref mln::debug
+ <LI> \ref mln::display
+ <LI> \ref mln::draw
+ <LI> \ref mln::estim
+ <LI> \ref mln::extension
+ <LI> \ref mln::fun
+ <LI> \ref mln::geom
+ <LI> \ref mln::graph
+ <LI> \ref mln::histo
+ <LI> \ref mln::io
+ <LI> \ref mln::labeling
+ <LI> \ref mln::data
+ <LI> \ref mln::linear
+ <LI> \ref mln::literal
+ <LI> \ref mln::logical
+ <LI> \ref mln::make
+ <LI> \ref mln::math
+ <LI> \ref mln::metal
+ <LI> \ref mln::morpho
+ <LI> \ref mln::norm
+ <LI> \ref mln::opt
+ <LI> \ref mln::pw
+ <LI> \ref mln::registration
+ <LI> \ref mln::set
+ <LI> \ref mln::tag
+ <LI> \ref mln::test
+ <LI> \ref mln::topo
+ <LI> \ref mln::trace
+ <LI> \ref mln::trait
+ <LI> \ref mln::transform
+ <LI> \ref mln::util
+ <LI> \ref mln::value
+ <LI> \ref mln::win
+
+ \section copyright Copyright and License.
+ Copyright (C) 2007, 2008, 2009, 2010, 2011 EPITA Research and
+ Development (LRDE)
+
+ This documentation is part of Olena.
+
+ Olena is free software: you can redistribute it and/or modify it under
+ the terms of the GNU General Public License as published by the Free
+ Software Foundation, version 2 of the License.
+
+ Olena is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with Olena. If not, see <http://www.gnu.org/licenses/>.
+ */
diff --git a/milena/mln/core/concept/object.hh b/milena/mln/core/concept/object.hh
index 4d58fbf..6fefeba 100644
--- a/milena/mln/core/concept/object.hh
+++ b/milena/mln/core/concept/object.hh
@@ -54,79 +54,6 @@
# include <mln/metal/math/all.hh>
-/** \mainpage Documentation of milena
-
- \section intro_sec Introduction
-
- This is the documentation of Milena.
-
- \section mln_sec Overview of Milena.
-
- <UL>
- <LI> \ref mln
- <LI> \ref mln::accu
- <LI> \ref mln::algebra
- <LI> \ref mln::arith
- <LI> \ref mln::binarization
- <LI> \ref mln::border
- <LI> \ref mln::canvas
- <LI> \ref mln::convert
- <LI> \ref mln::data
- <LI> \ref mln::debug
- <LI> \ref mln::display
- <LI> \ref mln::draw
- <LI> \ref mln::estim
- <LI> \ref mln::extension
- <LI> \ref mln::fun
- <LI> \ref mln::geom
- <LI> \ref mln::graph
- <LI> \ref mln::histo
- <LI> \ref mln::io
- <LI> \ref mln::labeling
- <LI> \ref mln::data
- <LI> \ref mln::linear
- <LI> \ref mln::literal
- <LI> \ref mln::logical
- <LI> \ref mln::make
- <LI> \ref mln::math
- <LI> \ref mln::metal
- <LI> \ref mln::morpho
- <LI> \ref mln::norm
- <LI> \ref mln::opt
- <LI> \ref mln::pw
- <LI> \ref mln::registration
- <LI> \ref mln::set
- <LI> \ref mln::tag
- <LI> \ref mln::test
- <LI> \ref mln::topo
- <LI> \ref mln::trace
- <LI> \ref mln::trait
- <LI> \ref mln::transform
- <LI> \ref mln::util
- <LI> \ref mln::value
- <LI> \ref mln::win
-
- \section copyright Copyright and License.
- Copyright (C) 2007, 2008, 2009, 2010, 2011 EPITA Research and
- Development (LRDE)
-
- This documentation is part of Olena.
-
- Olena is free software: you can redistribute it and/or modify it under
- the terms of the GNU General Public License as published by the Free
- Software Foundation, version 2 of the License.
-
- Olena is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- General Public License for more details.
-
- You should have received a copy of the GNU General Public License
- along with Olena. If not, see <http://www.gnu.org/licenses/>.
- */
-
-
-
/** \namespace mln
\brief The namespace mln corresponds to the Milena (mini-Olena) project.
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-71-g93614be Add a description about Olena.
by Guillaume Lazzara
* doc/Doxyfile.in: Add IMAGE_PATH entry.
* doc/Makefile.am: Add image dependency.
* doc/about.dox: New.
* doc/doc.bib: New reference.
* doc/images/olena_layers.png,
* doc/images/olena_layers.svg: New.
* doc/mainpage.dox: Rename section.
---
ChangeLog | 17 ++
doc/Doxyfile.in | 2 +-
doc/Makefile.am | 11 +-
doc/about.dox | 13 ++
doc/doc.bib | 28 +++-
doc/images/olena_layers.png | Bin 0 -> 25711 bytes
doc/images/olena_layers.svg | 368 +++++++++++++++++++++++++++++++++++++++++++
doc/mainpage.dox | 2 +-
8 files changed, 436 insertions(+), 5 deletions(-)
create mode 100644 doc/about.dox
create mode 100644 doc/images/olena_layers.png
create mode 100644 doc/images/olena_layers.svg
diff --git a/ChangeLog b/ChangeLog
index 5379982..b94441e 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,20 @@
+2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+
+ Add a description about Olena.
+
+ * doc/Doxyfile.in: Add IMAGE_PATH entry.
+
+ * doc/Makefile.am: Add image dependency.
+
+ * doc/about.dox: New.
+
+ * doc/doc.bib: New reference.
+
+ * doc/images/olena_layers.png,
+ * doc/images/olena_layers.svg: New.
+
+ * doc/mainpage.dox: Rename section.
+
2011-10-28 Guillaume Lazzara <z(a)lrde.epita.fr>
Add more content in general documentation.
diff --git a/doc/Doxyfile.in b/doc/Doxyfile.in
index b53bbe8..777a6ac 100644
--- a/doc/Doxyfile.in
+++ b/doc/Doxyfile.in
@@ -106,7 +106,7 @@ EXAMPLE_PATTERNS = *.cc \
*.cc.raw \
*.txt
EXAMPLE_RECURSIVE = NO
-IMAGE_PATH =
+IMAGE_PATH = @top_srcdir@/doc/images
INPUT_FILTER =
FILTER_PATTERNS =
FILTER_SOURCE_FILES = NO
diff --git a/doc/Makefile.am b/doc/Makefile.am
index 81c34b1..099b497 100644
--- a/doc/Makefile.am
+++ b/doc/Makefile.am
@@ -53,6 +53,12 @@ regen-doc:
DOXYFILE_USER = Doxyfile_user
+
+# Images
+REFMAN_IMAGES = $(srcdir)/images/olena_layers.png
+EXTRA_DIST = $(REFMAN_IMAGES)
+
+# Documentation content
REFMAN_deps = $(srcdir)/footer.html \
$(srcdir)/header.html \
$(srcdir)/doxygen.css \
@@ -67,7 +73,8 @@ REFMAN_deps = $(srcdir)/footer.html \
$(srcdir)/online_resources.dox \
$(srcdir)/get_sources.dox \
contributors.html \
- lrde_olena.html
+ lrde_olena.html \
+ $(REFMAN_IMAGES)
# Sed is used to generate Doxyfile from Doxyfile.in instead of
# configure, because the former is way faster than the latter.
@@ -93,7 +100,7 @@ USER_REFMAN = user-refman
# HTML.
-EXTRA_DIST = $(srcdir)/user-refman.stamp
+EXTRA_DIST += $(srcdir)/user-refman.stamp
MAINTAINERCLEANFILES += $(srcdir)/user-refman.stamp
$(srcdir)/user-refman.stamp: $(srcdir)/$(DOXYFILE).in $(REFMAN_deps)
@rm -f $@.tmp
diff --git a/doc/about.dox b/doc/about.dox
new file mode 100644
index 0000000..d2e10bb
--- /dev/null
+++ b/doc/about.dox
@@ -0,0 +1,13 @@
+/*! \page about What is Olena ?
+
+ Olena is a Free Software platform dedicated to efficient and generic image processing.
+
+ It tends to provide :
+ \li an image processing library (<a href="/milena/doc/user-refman/html/index.html">Milena</a>),
+ \li various modules for dedicated types of image processing,
+ \li a set of applications,
+ \li language bindings (e.g. Python),
+
+ \image html olena_layers.png "The Olena Platform Architecture."
+ \image latex olena_layers.png "The Olena platform" width=5cm
+*/
\ No newline at end of file
diff --git a/doc/doc.bib b/doc/doc.bib
index b116a98..46208e5 100644
--- a/doc/doc.bib
+++ b/doc/doc.bib
@@ -1,4 +1,13 @@
-(a)Article{besl.92.pami,
+@article{sauvola00pr,
+ author = {J. Sauvola and M. Pietikäinen},
+ title = {Adaptive document image binarization},
+ journal = {PATTERN RECOGNITION},
+ year = {2000},
+ volume = {33},
+ pages = {225--236}
+}
+
+(a)Article{besl.1992.pami,
author = "P.J. Besl and N.D. McKay",
title = "A Method for Registration of 3-D Shapes",
journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",
@@ -9,4 +18,21 @@
doi = "http://doi.ieeecomputersociety.org/10.1109/34.121791",
publisher = "IEEE Computer Society",
address = "Los Alamitos, CA, USA"
+}
+
+@article{badekas2005ppriaa,
+ booktitle = {Progress in Pattern Recognition, Image Analysis and Applications},
+ title = {Automatic Evaluation of Document Binarization Results},
+ pages = {1005--1014},
+ year = {2005},
+ author = {E. Badekas and N. Papamarkos}
+}
+
+@article{shafait2008drr,
+ author = {Faisal Shafait and Daniel Keysers and Thomas M. Breuel},
+ title = {Efficient Implementation of Local Adaptive Thresholding Techniques Using Integral Images},
+ journal = {Document Recognition and Retrieval XV},
+ year = {2008},
+ address = {San Jose, CA},
+ month = {Jan}
}
\ No newline at end of file
diff --git a/doc/images/olena_layers.png b/doc/images/olena_layers.png
new file mode 100644
index 0000000000000000000000000000000000000000..56edabd7f501be12197869358eb7f678af5f4334
GIT binary patch
literal 25711
zcmZsC1yoe=yDcCHf)Y|9CEY09AV{Zl4Inj0cb7`HbTf4Kh;(-hBOnY&Nq5K4yyO3^
z_wKs)Jz$u1CQkg$mwWFo;-jkE3+&g}C@3f|6y&8ffae?Faf|g7`2V3cA`N&#b=8oQ
zLa837*a0@4nkmUiqdflm&22A91YUXOB(Lv^f`UW%?}3Vvk@X6A5z|dUMFw*on*tA&
zVwEWC8U^JIih{JHme-&CY)?NTna9CpUtiDWb96NNufIMVX}{n3_C{XLj+qrLl4$;k
zhs=9v8T3Ldej;Mj_$5i7nn~O3;P*ESx*l7|K_h;l1e8-ziK4OFix%B_o>pip^ep{)
z+|$a6h$i@zxSybIHK-d7#~WT|VquBE6TkY>6N0l9nJMls?2A0?x8om;Zo%L*ZX1ev
z+abT$;_ll2f}H<($y5Zs6j)0fi$-2WrT_U~m*!W(Zp<Or#0f&KICk?tIhK1ix3&i7
z9XnM%e=Z#>HT-nBKi9N)D7wbY&HakYEO@Mt9LxWAsd}2rQ+fJ?RvaS)YJCW2y!@eY
z;V*`lEN&6fwEs5$ZA1{YNYlnkr7nb5>gG4DRZO&1D|GbSBfs}HRBL`DmTBHO4n%95
zhpj&&e3uCmq`-aJc8ZM>ng7oE&WlJ!5=(N+VUZbBIMkUD7@1#dImPJVupDc~tzPZt
zozZB=f|EjYa}bixT&edn`ozS~^tp+<aQ0Y+0A9uP%@Nk(#%jC3GX0x3UkVH9KWw8=
zcieJ<wPmK<cLhEl$6?(c=Y5!+nwkj7wHQtyiVi=HFoD_-n?zH^*ziwP$>FiMv7FCO
z?cZGwQqOfdLe!#KELNnbc7E?Hm-`<>=b}wi(Vjfv%7lH+2rVyv9NKv9eWC4fw4eZ%
zeQM66T6k+7cim@LrGpmq?IR3)O&J25>gXkhA<FZUH0|X6Yb>$OZxr~i<ei+3IXqDv
zJUZPC8(%+?GV2{v?~J6uz++(74yk|+M-wXFQw5Z}`+LW0(Dh>HoY^q4iF{+ea}Fa>
zsxxNn(%W^*<ziR+?(VKs9m?p7+&{sxfxUJ!8nocPCb}{`>WG4YKcG+wyZ4yY9nXBH
z*<Xf+^~A@w@mpXmQu~exs;Uz`MIOZkzDMB<3tEBr^&CO1&gM$hvm&Aj3}6}26k-_$
z?z;=W`2tUNO4785^Y+>VKcR5-N0EnIPxeRM9UnB_iQAGmGpVgwLDef%DRX>RFsOX4
zOo?Q=H(X8@f0zBcJDrz@8@tmLSMzpSWjZldx92M>mj}S&F-P_+Ew)~}4?I^L>_`1n
zr+2kc0-0^`=E*3YDAecq^TT#<z<p1z!3={@yGZBYkK5ivZ`0E6B%%O_SSEGHb!aap
z9@@t|4F;dfU75jZBY)Pxq;(5%zb#5`N?iPJ9rw3T|2-WNSPtxm^IEA!yD**jb?_he
zz1PV7p)re1xCGCA>ye<#Hok)q6Dw<^u>0QWuU@>VI;B+Kab@D8c4SVI+fKrF3(ZCY
zK6PPXgQUpFmxla~Bx1hj_?D3*#Qn6ieQ=3~ZRk;sn6I>*ax^nBF@vIvOoWa~M!1H$
zdj99`Z~0nUNz9r*N}0*XSe~BmZMlL_0|V7rSrwC*HKavnilUfF2xuZ@q{AmCCI*_V
zlpzlLkob6-d>kTt%P3+(0!c}6YvQ+BLy0;oJ!XhlO4+OW{wSI0Sv_@bZjGeaSdxAm
zQWD@qq=$#(CtCD*^7dNrrNV<pQ)S}6;}ej*X2r(9jFOU+q-}YR`k!lvjwVv`IN}V(
zmqNI61s;OdIz~v?7=Z^E<fxn{WwsthjZG7LWzmPV2CBh*^@)Pob2q8fq(JtPkC>G{
zl9bgVt3nc%u{yRhC51%;hGj5pK0LIR8dPV+cjECouqzU=(2fvsu>N}}%E?95_eVzH
zkaA+t1Uof@RwEo1r6Xu%q<6MItJBj<6JcY&l#!B>p7@3#Q#$PfPcm<u=a7x1U|HX5
z5LIpg42OY6oGNixT3#NjNHIA-41s=ecD8n=Uf@D$#hw09n^;?0$Zs3}K-lfmA3TZU
z_lb<M6_0~iLU!;Ig^J<U&VDn~(PEY7uSC-#G)puZ1N?*WOUp-^xXgi*5%A+nPEJwK
zev$Zl$LZK*xx>Q3;!SX_RFUHI+beTMl}uT<aeH`l^m(14&&44OV)zpuk|5=0!y2Vi
z`6mA(%@~+xG=+TyK3$<s6ha#x|HE$a&(*ok6fdiGwesjpm3OVX?MMpSRx_iFBwzZw
zbmaj)mdKvWn$Gd$V02|hLI%YYi{EKn(XV+z(^mVr(v|xL2hT189}*oOs6>2{S}u0g
zE4`cn-~LF1(Ky;ageQUihKDDSBU7NDoYm6YT*&{L`=qRlA~wprCy9mEehG9rQ8ZUw
z>$js3N5+n>J~2VW3QDt?t}utrISEqGGcukcmfE$d4fx;zOZPG4XB&MHD1}yFm%CXm
z&K!xjIXm!lz2)y_@Jo!;DTPPLATRsvk0^iKx8JAd%MeI8ZJt}lQi&@Uv1&hDwntx|
zYwml9=u{-v*VVCuIu^oxuj1vLgtG@c<Ucdh$;MqaLw5OWPvr+={RHe^fBpLP?qS}M
z*Lo(p<zi10xDhSlr&zuo2R%D(r{}KliHiw2M=up!!k2o@enP;f1u1b;Dvj;i`nP9`
zZ~#x}1w-rHCTJ4~3D3rsnIHifcyh7pi=N9x8qF?k*8}9GU>;lD8h%z=IWVJ61##<Q
z?B=}|4=;N(Znnr&^#@FrbKiD8_%4Jk^4p?A$>ZN{6e13t=$8yiF57=4iDXPKMv{Tx
z!~u7EEq43`e)-|x50_;RI<WYHKI`(;_lph!amf80*B2ZH4Q3nlmTAh2{FNqwN>U~b
zenKQQz2&Zr4$Z4j^c45;X*S*Z2e?Fphgk1(2`?`HHwKKNqT)nazJqgVns$tCRH
zRVLIG(>8i+{53~^|32IUHrdi(>+eUSkGKIs5c#!EgO58MCkxB)8qd2-7Ohfz78aH)
z3D4Nc_;_hqIn82aGMkxdl~g(EBB}mJWYPy!?E8$FyRL_O3Sp-&U0v_B5+l|IW4XZR
zp8-R|prifePN}V(etiaS+pj`DJ^iu8o|c!FCu=%BG&BVL_6!FPdGgLGy+1Pldq#qb
z5;8F%D=XVrZ`S?jHj#kt%4T1W-S}kd+}O%~OJHhrxurOXi0<G`O)p2u?6+^<qHBS$
zizX%IvY|chJ6t|g{}cyx_U_($F9^@ez6(W5o8cXEI1siyj{fu?8nwQX@L=C-v@cMe
ztFM>o)V)5=_B=)+<xV=F5BqCzF@oCKG2J|JaTKP5Kka{yn{z^=NI9?0IpQdV%zl_*
zoOxYXa;FRWT)msNo#8b|Uz3YN%OR7BOl5XfLTa>Wv2Yr-Sb#o^&xlyrdhD#q3`A3@
zeEf(XIou~g?e~81=7Ot&Rg*~#d^wixZTKFcMIzv?!l0CI{rZb3&3dKsc$WCV4suVp
zUnyn%G?}e=vST7W7~k#iPcjLIVZ9ZCb9Hr9(0TtA;Q2HUlMFy=CVS|{^yBfr5!0+L
z;^CBAX`-NpcZm5uSHHQvB{a_ca)R^h1>e|W*B2CVGI3G(kInHM`}4r5sY$(`7SBkc
zFlf+$3mI$Z?6<L;y*(lixnGYUv|Fe}O+>^`SIR^gwv;gC_+XdxqIMyd{d)Cl!n=4z
zj%MTOLaE>EbzldFO_kZ|vT9Yz&$(<gIX(Dw?98_Xd1z^GY{%096^??EN_+KMt$q~A
ztQ|j*%*wCsAUjYem5lmFq7Zl6nO6u4=v-5$v$;8IVdpKoriKRA3!DOHC|LAjjHeAb
zA?GLLOq)^cO%agFIV^tDZu9f#QaGjlrxgsMHp|C&bXuh3eFPKWDq{`9rKDXs-2ho<
zi@1g|A{XV9m7hC1J2z#i0=2-7+a!=xLH-}KwdZ~Whyi&*`Lo`u!8j^D<yoM>z$c*!
zP@)bNoh&0<R?Qv(Ru71lqO*n1>NZBr8w?Z_mHXiEIE>L$O`!JUZXd5;)2%+fBzOGv
zLk0&}P8g&@mwfZ2At@%ta&Hr+SEbXD&K)ALIo7@7RO_*)gXZdHQe9;7_@GQ7k@NTO
zryT!N1y-$c^sesv+h^FgvIt(tN_$}H`MF!h?;5rKk>utY(+`mYeSh5q(*TzNe4PR}
z?1W$X=Qi?<CX)&+4-X;W*$sO$MWE|bB2JCIEyPbF5fCS7?3i*jBmFCC`}OrQ<93wU
zdT*Z8wKeC~MLl+sfGnXzGz^?gZW&5kOcUj_u4dP**$#I$1qFqsMF(sW_90S2#$0ea
z_nA*3@QK*5F_5*zeZ7Gs4f<5T)z@<45>r{(r?;2;|J;H?%x`u(E-W~d+w3=UFn$w=
zb8{?124dU}L@O)~!v^k2_ZE<N75FINo>{#-Nlr99JxjzlWy;wZ>*#th15A-Q-AO8S
ze3R)o3B*Pqk&%&+?awttiM4vfr>95E*%ccN#y<Bs;Dc~}<Jo&nfqPDooP*!9wFFhZ
zO-dq~xxZ#A2*oRT#CdzX`vrA>;pYn@g!Qjdf=BPYD&GnkLaA`%av+4%#E%Q^J$%H#
zY0|N8Efn&GSupL(moIxqVOG)CCu@T{4?}da%V+m-;`!!?>PnN=6f%hi`2yLP?T1-&
z+0iV9bE)p`;E)hgHMOnv{;0x|qPkB@qN}Pog|bx!yn)A(XlQ7DH`BlpORD3ysK!$q
zb+5fOLUXjl2i#7Lt+II$$f!{8hN5bY;8)l(<X`5g@UZW7E@j5HQiFh(7Mkgq7}SqC
zZa+Ic-UBU>ni?`K5X^&tO>{hrQ!=GsV8aU7{>{D=zIrJ>V`9XQAH4fJJGsn&D@j#5
zS28d%);qvn<L)oxy$7<QSFW>pk874W&@oBN%4&4@c^9DJ;NcAdrT1p0Bez)}7R2{0
zs1QpED<LsKkdkz8oKmF8>5u{_ZoML2^4d=Hydrmc>$S+IVt&3oG3f>$hK<_>7Tmkh
zU_F0c=$^?x%fn;6*wPU9>1<Col$5zk6E2!))kwSN7+_LlbY#nySpJfj5@xO@mb{lG
z)V|jrNxIdK`ox7YP>g56reRa$r#Lh7(Dt`=_fAt&Q^47#UP!#;PFI!&omnni92%(;
zN%i)Wl2EBOyyiO<MJ_wP84+(;4+V@1WSodxX4SjSMuNpDu1jimAo7pKFPn&N9@CP3
zYk`dAjK>MA5OgdYFS>XHWyU*zuCKG*S&)K4*qd9I`rHSL#!10h>4ByYaUUpcOt<u)
zGNt{mGyHFDqCEXIT{YP>6VN^`gfK@+X=uF6D84VioePxS_MSFFgj(J|;Npgnt4vw$
z#4o4=c5!d6>GD`0rArioffIW?sf6R-5n|lRAFESUeEa-FS;=*>eA-PzgF-;L(Pld0
z?zopq--SM5zTd6JSS+EY!@w5n^{4$=7`sV8n)}{rbglOpF4S^V2txLux&gPoCyrW;
zo@LnC2uF%G{w}E47pM*&9`e}r{-*CpBQTUwskYLscDdbmIn9?%0vFZh>M^jeYP~uL
zmi^;NI0AAtik`KdjoX+fd}Bv@!Vyi`n$D`-v07FVFnEws+$McmlYUrklP?=X8(Kz5
zni@P+#`XE5f55GM=fu`h<I&3)a#A^MvCRNj2BGytKG9LzRIo|Fbri7Y<DJyN^9h5W
zc0z-`M|>M&=@w?@ok<1b^a-|Cg99y@%GejX#2byvxPM)&fq*yk0WZ18=R%&F8<ZuS
zEdB9&vSvxb_GCZL+EA{fQkul$vp)@IV}wgT?e%+l#8*d>pJS6=`U&e7$kOwGy>u2o
zFSXG#D8+8<)p-7D^|ZfsV@L>WiNpO>tfIfVHl?6EW^sK=D0&vXWw`S6h1qQ9BR7*8
zBW-Zh?a|xAgCG;Lq3^_6Rl@PC+KiJwc98F%rl-$A<N57gnctbzenP>(xq-Uxd0$%q
zdo7T`{Jhx0?$DUcrn?W^m2Y_;jxhhuS=rY6-wIV5!*VT0PlaNgfU?>V4UIhzTJeUu
z5?IF{+xj`j!tRPE2aT-}acaGAxF+1~$T;17Z~S;lE%4-OO*_VXB#C9b^{BeTp~ZG4
zi~B)M#og(;KHKLaa)gO3{*GJ<?(!XIq!T9!5E|7WNjrXc?$@sOIC8P{P8a3C$MRzo
zSD<i%lk2U_tL=gY3NfD$U`-|}5Y&am!qFq<XPeNwEs0SfH|DQYVqiidi<Rpw+u?t^
zHY^r-U!qE^2iT=*HtVZZfpnsqi;B_V#E8DW;MYgLMBGZdx`W<w`QI2X1^5zgjLIGV
zv3-H{{Cv##W<7Cwd!hh4gAu@cC}UptzcZJ$DtP&Q$unZz<fV_6*1OKT$!WRh$oyCL
z7q%Q<PB-X)liO@=L6O1=xT5jZ=)OD{-SBRcroB!rNoXPFQbbjvZ*-<MIz2k}41Hp9
zM5riKDflSbI;W=g2op_`>Kw(WX2v!jbfDq)pdc9y)2hBBo+>^#Co^jDI#HL6TS~LN
z*)r{&sW<JlHot8^q_}QG-siyJ!-<WWKlfA9V7<&H0Tj!A(9vz5Sn|dYSSEba+L_bH
zCriD6|E^-Dl<3AtQ75ll&skI{g~D>7iQjJV2F-ln2Y9ZY@Htui<gDkz9o=*44u0L&
zM;)z^4Yp8H@ZCOChMs=*Na!mQla)sM4Lx}5D(Y>gUGL>VJDET=2b*qAq2n?Jir>RE
znBQJF)9Q5;DYfNnt;Ery_|38;0>h&<D)&u=)``mL7?0~S+)MM5RVM!wcbn<*=T_qu
zHO&tEnY9M4fz<YOrKRPonxI2IYK0Wb#I&^GrS^N54Re?JE$y0VAA4iVf^zSz8oP<J
zH#FU<U(pH*eH>r5F7{XIV{XquTi_J~l~S(>32AQ^1%`iEx2ACyWycjICCBCE1$g#}
zEx1JGEc@>6RoS}p*cxwCa=BLH6qjQcDTOUh@+~qSf>)1tUlQ^-3G<C4j~cg6uQ>gS
zMOxaDSkaEy^audt`wE6BZp@3d5H@X)0KxjFql6k1ysY_S+v<mdSa!Sb+iSnK52vG1
z;SppX3S^g^jz4`P@Vs(Dn<_b4Ss$i<&6_40@e=U4Zi1QWn6ol20ej)m-(@1|L>PgA
zIR3XYCuQ#jVi%P|u-D2ynu<z1w!y9HBkRlsG^-Z`S#>DjSwdlctKtt{VjH7zC0b>5
zZ4Wmtya$KmphP1DaG0e<)AeENqTk)gQ!1~&Y-N8l;VN$T<39r&^tVgYm^2#?>1}?8
zTkQUp+g@mX8T6EcYiuAbjSHxXdUQ?1lt)vFO0}$?w|bhOe_9)Wp6?4~dK`e?c^&(=
zJ@|ZhxLU(EY8{uJa{KTdcwYZqM(<daG5oiiCt7Nu0s#Wf>dA-D1^7@r$JNbVwk!xD
zj+4f3Pud&Gw-qDsdZKE+CYlUNCZi5On3QyK)mGcTZvWX%dO8&$(xKozYG9>Kau6mo
z)e9E~dIA%m^twI3tlo*r92$)z?U@SfL_~x$_tjeto;7?tvY8U&;(Frr4y<fTw;w_(
zYCfi0E&8l$p;-(CUfSS)yPD2xEt1ICz>`I?hd|PbE>M4|GK6!v@8ZT%wklqq66`Ni
z!>`X$um1)W8>2q)>M#=t6i+_tn3uBI#SC?=HXp>6H3=9=suIjTGOhEXcvDuUDjN%e
zcYVE>t6kx3_nCet9&eA}ysX}@_+tmo31|^s4rq;%3_3j9c`1~g0q@utA?|bVj896k
zJ{Njx^0Q#?+e?O?L+?D~j%FsQahlp-EU}9_2k?$&>Bp``w)55g4=hH)8N1U=C))+d
z-rITgzsrd+x+aSTn|!8K?e+!fK8)w6<T)&EQ#4q-<2StA*_&olw6MT(-K4PC{)Opv
zF>T4~0HV_7+jzOU+CRvwQ~3@Rl>}n?Bd4hVMXOSfMvj_o>Z<)%%3j#xF1<pR2S_SJ
z*&@31^nFLw6yy@7J-xg(=P%s%0#qI!E|~SXlSWeylrw+gK!sYJF8B3=@FYN^Q22u+
zMBM$Cd-oQ^F<+~!7W5(Nh%Xqw?RxXc-Srkrxz=?g!r1oxaKZ{;Rb$BIqdc>ItbrKY
zsNq7VRcYEx{%px9(T(ZlWvMEIc+a1|dVBAM0{q_WAt2z9B+9mn8O1+8SO)WD;(!I4
zsT$?B8ZWp;k|{UK@Y!f7`5<%P6*c@8i!D>pucx`yP|J5OZj}M3jzWb04ZI1Lk~P=*
z=<EU^LMGzD)9wQ^yT2j*{JZqZNMByS&MW1pa~U>Umnr1jb2jlyQ%&0n2tv_q4Rc!K
zIsPFJ_oS21t)nz$>^KSI<c(2COL0QO#`VJC?oxC9-y=uxVvE)FPKm$Y>F`t4Y(;v0
zejqiHlR;^S*|*_9sEOt*EUAn5v^TSKb({I@MZ^7Gh-4db$#-?KNNKgx^kO#=ZHcf%
zl_9b}Dh}IA?sG=YR>yK+ry&uisKM-kDxEUIJ^uT5PUT`e>BomC^fnd^O-k9j>oXQK
zV*)8Gz>aACO04#TXhGcEjgSNFG=Lpv?6tMz-1_`!$fPOPN<TJPyq&2s?C*~vJ&V4*
zK$ysAXe3-U4(;W*9=WbHWh!hv2I)h=5i#Tfn`8oR+aFIH3-)eQwqTv{VGfOk(G)@&
zKz!_QTxJ-Eo*PS4)1iCQ)e}yz$802=Ra*K9n-sTDHT%^_GWp>8@Imby`xqoJST-hd
zqQl=i_>FE<B{TNQ%ERw!%Kod9y`9M-AF3d8(ZS%*Kd%I9S-7~&)A%YxfI{PJT%<z3
z{l;4AgZ*=|$;f)kPK$x)@xB2P3n1cD8n}OJcJ=+$iu5bj;ZW`k<?8EasPnL=c)V6C
zXl;>cIT06#O{6AV^m$ks>1d4EF)r5<yx1ePoMDGs!~ss3Yt-tpMDU)~79rldJ6)0{
zbcUi+p$NFz=h@m<D0OuhUZ=x0rR6NCLK1d*JONjxZ|P)tkr8Le#$QFF{-Fvtb5>Q`
zOYPhe_diIX`WsVf)V!<WH>E<0t-5A6=k37xMFI(c+fvaN#n`Wo^K6THv!dpwdhQnq
zHr-2Lc|Pl#F?jMdvY+R^Ve`NFJ&be@f`N<e=k@;ljJb#)mHJg{8VxUu6<RZF<W$<5
zv5YyWG<e58Hc+X*81pVMLUih+@0ra^bUqkX3fijH=Sj%mqpK0B$+A+V4GB;xRxk#S
z#3-lQuGORtTuk1yZl%zsE*uO^;QwEqN1k3|`(=bQwh?W-e4Y~hf5;%0sWenx74w5X
z4Z)$IBNWXTv6LdR%F4<)&bVPge6}<A`1o&lc1Z-C{-R7wc*lXx{eD1E0SuUjmD<f{
z3m^@crLdfV05CZ@x$d=TD2*kV&TUz%=TdXybA4)A4Qi%6yX<5x<>KN($H3T#Ox(EK
z94YQMlbmoj8j8MIwzL^<OWNxkZfiIC7@UQzMccd=KQ8z`&hbBl)dS5K3TG-nXIVPP
zAL`xRotiv>9eAv{G#@U`ELfak`L7Sf1N_*F3hinw)@CTa)c2&MzH;q~)5Wb4NWkiL
zd@piAYdYuoH#S{DF)3Z`WU58ssd7tKpyTi)8Xjh8d)RD6EHd%Z2YZHf<ZKP~`{vP7
zJPdD9<>7QKZ>ptH>~S_f!8wh2I`pjOI^b<=A)(Z#27@9VRoZ5zn;mYP|Gk|554ZL|
z)*Sirak_q33QJ2{reR&4(e|mZk=zuYo6)*&<(%Ds1-~zzh7%Z!)|@KH56~|*I@yJP
zT?8L4*S|<%DMx;r1F!u)S)&IvM;7k@n8fby>WE?by|zaa^x@$$3JGVlkN8q0ss{!`
zEI3Cfd2EqAvDvL@B94+jBR+8x{B@)g7LGJ%uz6`%{Jgw8pE)uPK=Lg6%lRtxnz5mC
z%Cca_0=a<ov%k;>h>$J6@8tuA*Jf^@{qg&r9;wI6h5;t}<M(HnLe8td`I5hc@^I*n
zv)(+0muFlD+{Cr<uXCU%rn2^*tc5wv9dNrkc;;ooz#W+nbc9T5Fi*($@1K0!js^hi
z47flG39qOSys)@Wy#p{5baeE0;QcK@*RiQ8W^4fH5S^}Qgd_VJY-VzOQ&pSz4vG!W
z){}_J5zTSI!E8(4bP^_u6bZ-;Nfy#I)*qMsrK3V(41d0!UljI4Z$9ohc<V-k9@*bB
zV=q$QZ_7=9kN?GRYG~*<T$q6&nMJeH6zB%)0xq!E*l0`W65DSt$p@E~OKF-w;@_ZG
z*ApF@2vFkFO8wSYIbXN1OrjP5cHbH|3LWAA^9-s{V8AdkVn<ibg=(tz0N{y^-Bu$u
zi}>TX-_VU^h}C^y0WXw2rLCu?$}1RO{C30r&=#IaoazokEUC>j*EHl-w};stt^M(Z
zy`_XBmm|0Q%0pB#f4;hP+x?9SV{=o}=-B#Bd}RMS-@Lxoe2dc@Y91!{3fW&$SX4wv
z>Y2fmb0~ASfsVp&zcf-$%x1mx;~5GmC+JjPvBcrPU?2$2${zU8oKiHZyD5B-A%Xa2
zBN-mOM}kFzo>9zYbtu~Lc)~>2rco4m$G=9J#mQ7fZTAd|ghM|xKMi19L~o@WN#Kyd
zq?%62Dc+7u{`B13HzaS3#v!%{=j~%t4GoQF&v1*4+gqORKhjD{eihHJpFbGOZbV5`
z=-+WWUciC#xSwi)?@rm&)#Ibd#3E~JKUPZSDQAnun}Ej$mNy40f}Z~Rp<oUfQNjAk
z#~faSLwd63=qebnDv0`k`opIdBM7r2dUWSp=QFEe<H6sdiM4I164$WHtJ~WrwN^*s
zxKzT?Q#-43C6K-<Ydv-__>P-U<qWG+e`<h;1vIO-NC!|b@J@=sn+x*cF}IJ!ZY;*T
zh_DRzy;qEl`OLsKI<6h|jvFlgj}}0fB#O9kUIoEPsi35^RzOfNTjxp&iM|V$eL?QD
zvMRHd_J_|l*ZMUWnx0&Na5`CKaI`F%D)1F38+&YylE&iNFNo{gHwh#e|IwHE&^(#g
z_jPeT1d+&ya<b+*6`T@JOHrm9!o#7yzgQ=e1}i6!s120Jvns}x+tse~UbM+8pL6bf
zxN3LKqPr-^=WPA_=PH0soR|0a7D#HesmxaNhi3`g8S%iGEyThY^Ef%n`SA_txmOva
zj*^m?W#Q7|WPeKROa(x0oj)I`t|+R#9RDE=>_3uB!q2u3^C^0Gs>$x2tMaE$Uz+^H
zRezW3i|P8<>acSicH66w$RUpifa(c;wy<Jqx_Xt@87=Aycs!lM;O4~u-}IQXA9#d#
zMOlgH*3*%vhX}tUmI^=icZC#Onb<^)0&;ZUac5?-CL}cVvOnTO;C0`l=)DN!fM*ad
zLR*RS^?Wj$4u17H{lBwX>$n=jLLwU@mIH5nH;yOg67w}nVpn~}(s|?xnFb!K+>TNp
zz!F?}tz$WlR)b~K^;&uw8|)E3Ut|=Mfi3#6H%`u=WG92M5a|=)O;w0c)Fxj$u}q%y
z@T%;d9e?Ka1!7qb>LCUsDs{h$jV<vl;VZ;f@=P`o9M^|5PfE=Ah@O7MdY20uI(9z3
zn72&Iygx@%Sb}z`dm^VRcBXE<Q{Lvl>w<i*T{ko$iP^w{Sdf52jIpH<BI2l!qR|v`
zYr!ns0hP3M*TZu+k-Fq^s4F4|gt+L9<EhoF!lDuM*%##76EH(Mzylz04{3C_$X`wB
zbSN(q6SfzR7~*y+_mY@%+Gp3UaY)})u$I8HED?wMXQ1|s@O<VwuE0Ek%f2-+W_8i0
zQT;u8AC~exJ0E00w?FOMQ{uv$O;pTICN0xoR6$N_n;haBt92zgC(RJEB6I!wEn)Lk
zvwmzqn2_Ujqi;-gUM>Jpa=rUy^22ss)@Sum!jHbuFL$_a(vmIuptkm%RPP7>F<t2N
zuDQXyUBF*H909w9bxlFVWOnPcGT`Uj-}(!|<4UWeW~Z?c<m_k>iFi>nvE%g4ZJWH0
zQtWdu##8h+Kb{BUFI5M9(vQVfn&K@E6nL>xUoe+={gBD>nT*A5xhX{<ad4{M5(j{G
zKCBh1%>LlNZQ_{!`GllE3F<V=O*g*03=x|bNKE(Xx9T^59e@{iH?UUnU5X-Qz0AyL
znXzZO-sc=FrUIxgO2@s=%OI0myt)lD{dO1U*y=H1fH-TZCL|_K=Quwj?RxvA1mfH6
zO39z8E|m5J%`dBTt@G-6>L@IpPVU-l0iDk8di8qDP%xEYH05mOJ0j*-VA-2GQJ_6b
z3Mjp5;J|J0BB5F6qZA2un$>R&2&z+4t}_#z{VB#{1w+mCMT}D{v?hjs!`R5MX*8r>
zTmhsZzi7-`Z8k=>9bj8D493l7KmMQ)kEY5Yc$EVJDwmrku<&qki=EtfIa~jcdq7TA
z7#{v&)d8TJ5c-u7VnCsy03h1f{?yqb;w~R5dXslPO{_A^RVmfw&uVJa>4_fgPfboK
zsH$?U240`|ws;(dkFKEPes<%{C?=qy8gJGt(QgL7*c#7anp$}yF1{o<&Vy4{lAxr}
zBrPTNjX=P4!0p2jzl20W`$`Qi)%&2Ei}$XCFJG>j%e+BnOX0LkL}(stZOJX{094Xt
zX6W?VUK^UqR77O-sr5wM;qqW;OeSECp??}|xPjaOM1MWeKvb*o9E`s^Ow>$*g0B+i
z#&i=q`E%yh#9F3KyHf@-zkArvbKk8^pzKRH;eVWP&;AgGOTJa9+lUJZJmo4cUsY6N
zVC4p2nQ?$$iT3Y!CD8pDUpC+zwWgDs*@UTZ-GszP!-czSDdlOdVzOeWfWY3@gM|k2
zFx_{N`M<8d-u`22>>u`Lo96|8+Gpy7Vv320We412k~trBh$TrN9N&m7s=oiyNpnso
z7q{rvJ^O?2wn5d+jkmn~^?asl+vzhryaL#VDyNO(2dng^qkt;{OgyR)Bt^_?eiYI1
zzAUhdy!^Wco8Rvd?pd6qsV-W$LB;{YnqAqd5OW(udBiXGV&&WM5;On6*<-&dHX6}u
zW9Ix^f&d^hfX6dlbakbga8r~-L?}G>7V(wpw}R~K_;Y?8yV^kP?~xzB%$Pcp?Xln{
z2o$>gLq|)C5(hooZ$6XtLULyNp9)lDCL|?gjj{y>7Ce3Ky=RnV9nsu&n=t==JBP(1
zQ9%ht;<FJ+>}OnI0T%YEw0isD4v5>l0l5T``NxQAqYnRDemt5s|68k^qo9<v3wVX&
zbJ*GIRaMAIY)PwlMz!nRm4bOI$W9-kD=8FWobPo4s;M2%5*u6-^CdGr9>C+xSE<km
z4h}vZ$D{^AB`peGqd!)bYB2u(7|kCk0)ngC`H$Mk8$#?+z(z0$>o*>b5Lc`5*N$%h
z8kuom3Mr9PU}inn#)8-2ki8{F!j2z%Sj7w@iCUqH==4jMQA){Y#x@dmKaN!lvKR4i
z*`IYri5w&<a6SVYzio*KX47?D*RGtHEDzAx9D`kYivDTf`K-F&AYkSCV+NMVcf!H8
zG$Jx8^@Q1kGo@E9wmzVjQT*nwSZK#d`O19!f~=H4x^k9~FRoRr?t9>tcD;NxmQFZ$
zbr=SDCE|dlTMZ`7XsBo#C#TVC_K8vRv8T3)nLpSB8gX{ekM-i}m9cc_8+3sTwBH9V
zWHs(6yNIeaP1A#=c7gfIx36Ou!W8W44AM_VYeI)aJomfym#ye1$o@$;CcC-^I_wP^
zc-g=D2U^DAF@s9+A08gk#=z8Dx`_jLUpB2d)F5KETH!z?vCSNJZIt`TV~>KIe6%9S
zn4eXX2Axqt;+K1MpNTBeKmZJloVMv-ajN}1%uRudX>TAQf5N@-3=aqM(~TNWt?!`F
z*n|7i49{&Yth@WWp<@HnusDjMYR|OY>H`s=AXghW3$z^~|BTCYa*|rN^=eTEs%j4+
zn|o5A9NAOWBnZ(h8zqI6c+Fv>v?06S88q4`z-+W*e7`R+N(mUUR~ny{sOiCcUYso;
z0A{y_DYgW#8Mv>ewxu?19w`>yKj)u@$65Ix`B>xRT0Qa>d+npqfLlANQ^~D78kNeK
zXZ^>0u+&OuffF=D;<G5+;7{RL(tfX{2FkXz<pqR~7Ist;0a<6xObTj^Q%4(LWxYOr
zdV_&uA(9a5W2p2c^u^(s9wj3o4vD46uME*&Sg@UD#8e6Vl~kkNn6#Ws{+@7qPC%pt
z!l!L|IY1(tiTX%Vc_ROFi>!^!Yv|V4Mi5|eB0k@N#sdyjhh2Gx<V@RV2z#S7aWj8>
zvv7OaJ5{e+rIEfnT@G$6K2**LxM5*bIT743Xt0LAmGD;s%4}mlp;nRU8Z}3N{{3RZ
z4;37o4=LAX4N|z+(MYc%fkd`jCt`-PXy(*G0(-i2m)?_pE9RhHy(=?EN54{%S<>-$
zi|XG>)(Thv8jNMNOsSJJvw_b{VN?ApbaMV<f;8w*@VMCqtJ>FaX1(txCIN>$@V0#B
zBzE<l61-+t1l8*W-M7mdg+QuHU0oY~XTV;n0e#BpC}A?iJHCnU)B6;8ZSVvaGfwPJ
z%n2bAP#`ax^y!qx_O9fD?Cd-$jGFNIGa>3tanx`UpQ+*)g02l5(&s!T>O|#Qmsh-F
z;N<L9k1c#9;!1vUKwy}oCOYsfBr1*1iZeYuo!4ga+xNep-2Owm#-?a88}pOs`jjJR
zKl<w|SkJ7ZV#^RoZ3|<xarlskL411Tl7L+xm#9Y6BpwxQY*>d`SbkV94{xQ>)O3#2
zB9~9}NJygM3ZRx};J=rYi{W!UBhzy+QhS(&-$qrJm*8jfmCO2_&V(MV*)s>iwczVj
zkP*{xCyN^mvr#=!ky0}EH?8U5&Onq9T5OG<%47z-wu%w6r%jWZemIGv3TMSCrTPUe
zav<ci9Ks!!LR3EPa1rQ-SNt!Y;k<zH(L;4$uSF*hP0`TqfE+^4$Y9*~5$NDJK~2(#
zaY}V<9+^m@r$(b4R5E$q&|=^nKv7$!Tap4lX{pw1RJ9kU<ogCRF|<JKNE?DPu@$mK
z&@zDLC{W;5rXjwDap?FPUt7Cwo~NmE#Xm<k)Y@9~PgT^(v@w8py_k@bCMuLeRs&tT
zu=pg^qHRCG+6DkQeIOPxbQgl7Am8+D{$KzM1=6}|*~|W&g|a6-rf^fqx$T)E!UzQg
zEE=!fh(Ua*+&y7pwv-avE!8v_D*(}u=IkZ2J<?*6tOG{(-20aDr~SO7q75s5W{d`d
z(yvD3*$<m(G^?5KrZmhNO;}<ro-zpc7WbXtc@f38R{AX3`REKvM1vbHXY#jU&c3aJ
zgoA<Ix4p7vb!pDcAPQl>VAf=|VdVBB9su;5OdPLhMRI(&4Wn+cwpxJwZh3_5CD_Kn
z-d62p({*yK3B2ad)NL!a9`HINcPD)q7jpCnG(xmP?!a-9@Rr!j;jK(qXV$zvGRwU>
zvw1wz+d*j1;I7e<k&7jw<1HR|JZt%EaSM<pl(<80kLDQC2cjWvwj4Jw@$kn1^w!JS
zyUCU%q0`9L3BC=|zyr2|$LW8EP{VNDyY_=#r@G$CBK>#}51m7J!+~*G_LBc0vHuJ9
zTA)g^B$1~Vkj3f-st@ND5LTXMp%cNkZ)i2X(fQ0267hU@2Kf4gSg1>C={25XAyK4t
z^I|4?Ko(Qg(@V3P07y{9?QClQESrWs%{l}76iyT8f(ZmraF%L!(oEYRN^GibD&Z&A
z9vy>vRnK4DWgtZx(k>UzABSxv=kk^41tO+)`gSPs*U4hazKeBwF9}kpu0faYWSwsK
zcL23_%-LrFp||<KkzPjpJ6xuuI#Fo*Bd&tN@aGI!Ac{H)YUv76C};C_e;%&TDf8<+
z+g2rga)T^z9hCd!mMz?1bEBNW3UKsy5;*Pmz4nhYTn>$yOjN$-j+6yE!mte9OXA1J
zPy3y`iwp_ADE|cE{~ZwYb<*`&B$nyk-@^RDLx;SvBdNepW`yfuE<$B*E=VplwWp!6
zt>3KFWmETlrUxxpWnWVuQV{_<nJ<Vi7B{#_T9f4Smg^`)0=_*<EdIo0c(tEMFo2$z
zBQyl(3S2Q$g@^~i;Bf#>s#)V;52Svkckhsb>xJtUfa!LJs$|y9A0#oWXZ@@)BqY{G
z&xCD~W%{kqMzR?4Vt?W;yPCVh!o@ySfbX~^q@|%pT4^KDFtAZ_SmHHZ*~tzR<RrDV
z?fA*Renl0pC<ezTB=i79>|B#K4rfV}VlvyWo9+~%$nf{tP$runH}hUw2Y5|>3|~F$
zsq)<`QtHo@@|M%FXlie6&u)G->eV`N`TIy%DTl3l*q(cxT5rAuC2iHg#TOH;rx@2p
z#?A}ZC|&Os&mn+V|6<Pb*!D5cFm-$^>*R^v6Bc9sqc~h_Qda3LRm!W?DfNMKwalM(
z8{PZ|sHIwUXty2?{TAFO{Y^Dje+vk_4G7iaWnTSylrbi|s^_Agd`;8?KqWlhyS$TA
z2OalZi4i&+nLLi;j<RWoE<jsG`N#8c;G0a8P_1tmJUHbg%elzl$|r7Tr^f&i_H*Tm
zJ7qXijQcRc5o8tKHXX#Vy|6;St75fxc~H4rJ29o`Guk$7llPdboBlWZo{yC^VX(!s
z0KkQ^#Qj5MqEfwP0y|~tf%-+O3iM26O3-N=|0MM-kM)RFj^DoeE~3JAOei%3(&&rY
zp~T*w!tue&>s5xK`Qz81EHTd<fzS?E!D0(OC2<J;x9%sHr~Rope3O8rT6K8~!obX&
zMjepR{qsOXuig%W!$4x}kG7I?^8yeYfgt*&v{ZgPTLccETRRIU1lB#kE%^#4YR~<L
zd6s?Tov@I6c3;@&o<OHZ+AwnZ2^CYAyQ!f8e8JNtNT>xEY}RP?cl`LrCYw$bmo$}n
z?VHx0cJ#EgzDt*lp>3)zirAbbgi#Fh3vUgJdC3vgQtCt`>;}qVoH^?108AiPa7#|M
z*Sr`W7lv1`Mu$g-9(`*=8)QO`)>SFwN+y~su{93a+ACXr&GBF&(r9+Z={0-XZ1s6E
zpU?28z13o9%-!)C{Cs<fSVkQS8ynD-+JKsz;z@5VN^Ii>KTc{CVjrR&vew*g1fyRb
zAlXaq?x1QPVrxRQ60P!ko2E*xS4U6m<~xF}u0~5qa8OZE0UZr(n@)Ed#@yz8xakqV
z2nN^js;5kGZe^X#8DYjz1AR)lR=Gv6Z-->ZwQ00Y1x6x_dfIGdwciVpV$793`_mqC
zXK7kpoJ=S>_wT1*{G6ghtz^?9tD1_RpC-&8B&~qWezCgzoo&d!=%vCI0n9=$owF7C
zk_PxSsqF1vPqgE3SzN4zb)1REWc(}yYDQ@ij^Q7UotXWEB^k=%6@B8DpBpnDv_7hL
znbm!dPw(b|BQHFdh;qD7pMh|<RF>ritHo*!p42@JJ{`6)NBKOLFJy1nh$drtAG)9X
zd0zY+|B%Aig9cqCQ~Z-1MBIOXsaj#}Hs-)xD1|!(h$dWxB_-ra!rJ9p6ZoHJR`GyV
zvTouf&WirCa}7Vy_S=iCi9uhp7uaYrG`%_P9(o_IVDQ7bqni6PyY673XtTx87t%!R
zLz5wA%*LXdjvR^rp2wLw>J_u)yf5>nElRa;U9X>?o}OO!Ps*21Yk`A&IIKyhoD`@|
ztG`*wPypYG%i?hS(is?=GGh+846MV8NQFZ^rn#SX+|dF^XQ=BwQeqS8ISBMg((8IY
zSL^ZS0tc0Vb}JYibL~jXvnQEN`S$X*v;B0aZH|Y)fH6ZX{p;IK^sZ<VF!%Pfz<WMT
zcHuv>0*6|lY7B;#<B*FoEI&%JIoz%>3MFlY!e2+PDs?JoMsP^ql090rd+?+y0|M;V
zOS+J^;`hO4&Vyu^OCodm;_f47Rn7<dA+N&XUdPPYjWg)LGuAsywOMs(g}8EXoCixm
zWwa=rV_E%|*{#*BEu#&BS*lqcA{o_FHAq)IcB;V@pn5~YGT89K6gNF3(t1RJ?A<ua
zUf#GRvnC-G=5yTouR3#)lpWI8_p!9>?7u$`3pmc%RPvf#Ah(irjjaybFhHvET{}jD
z3-Rd@3b65St6W?;0p4*Aq3;6I9tcNuEBg*S3d+j|7dnaHBigop#u2F7EnVsHlJDQ6
zbUt1}2_F0)q+GYlcZzuc6a{MbNgjRntkU4DgzGO8fE>8RoYq;g?AlKU-dyXZyfu!-
z7iLWCiAXGYbv;}nYtWu^xT8sPu$0)Y>^9A+U5Iu_a9ED8jgOB<$H0ttXgmYt>0SVp
z*LCo;f30-y&)PGo4<DHEsbV1zYSoe`9xKuD@{wd#4X6Vp$U}rdDe2QieLt`D-0ReN
zl}tHd&C1W}b`Shns!X~ypdv8rwK#_@aMOw6BJw<{G5}N}DkuyWYSB$=!0YIviRAj5
zjHorJ-2TT#Sp*02Skh+nDgd(*o%g47FUSQjbBzY8`-s>g%gaG+ZV^dDa1;oONlbz9
zqCpdAPD$-8sO`7>5si?r>d$RY+6R7excEI0ic)IT8;%L6GM7Wl?e27u!89u%PaoEF
z{#UI5WXsD}uZDs+GN9o8msv|2Nq{o`1^I8Q-VEY@O$beQcVD(RY^7H&JW52x%S$kZ
zoKJdNI9IdwJ?cmjr;Ku4erPVxZqSAKzjt#ZvQ$i61uUB`xN8|);HCT=*v{9R$p5m{
zpy#E>+#g9&q`ad6;O=S1$M0ydiE;t#PIR_jXq4nZ5~x*T5S|x%Gp1W9FBX<-Q2`t(
z79uztUohw-n59()im*;(ItJC0I159?PJtRFWXUpizR~`JS3IrD#-wksp@GVU$ADRb
z*(Dx8Mid9$@-4P{aP@6@v^qC0ZZDbE<>r<%0s{ci;-1D8X<*no3w|0AAvlVx!V}wA
zMjbea7sV<z7&<p7@Yk*R<529`^Ff${DYW=>nwJBP@DiFD^Y+(&w9N>eN*9|d9k%EJ
zG?=?OfXq}`IbYgNCJUEI4)3tYOSw+u!FzOCs|Fs}kg<E<r<pMi8|;->zo2pCGW%U3
zdhLbR>&BM}YesjfDs@R0Wh8uf__>z@^vxNLJ^?zm)$c0!kPqeM4^`7j&<9=Z>Elxq
zggP017hC)7s*@mq_oGG<$8^|tbS279-5;ehEs;3<8huMcVNEd$?HzP$Pj&utaCf!H
z8%!6I41ns?bMl2ehq1)1_&ryQh9qFW7ET1zn7&>n^?#6!|DS2Uf1`hg*y}gj2_jjU
zko&D8<p<}+H^J{EPq2)(g?WHxwAxeAB4Rz3WvJy{M<+ZkI{G;Yr?E$GT<D=r9uug2
z^;(&alQZ`BZ!Mt*eHS@EmOovtJu2`3%r=7ByiRgp)PaG4L`H}ZqNa6+&&_X;gTkkp
zAV*BOAr@I+9Fc0r{eOx||1T$fSMx6JHQ#QF<{KsY1QAjUj0y<x*(NNL#iWj6#&44m
zxIEA5Ax+IG`L={`{VCp?jyp7&g8X<oD`3ZECJttRsy~`itouwoYILKgY%~NHc)eUk
zz<cvl_d7r;uTxns{@2js|8K2a5{D!Y6Ix%>1Y4C`F|?|<+8+8kPzjwK+<mDm$LP0p
z!$MbudBauDnS#42Gt`I0EbHFq=H*#uWMQ4W2DJOfd32Qqb$@uIl(lyjkwSC%%ba@J
zZ2l%*@rg{TsNaT$X#J3DjZN-IG5t^XS`xC+uZBl9DqLY|4$wJf6Pgr{I8kNj0Bq<N
zcj4c5J!O)PJ3ESHm0PMWHhkq=fBRF1e(@Cjatxap{TAZb`7n;<dsjC2vB?vs218;X
zF8#e!PqOG*Dm$oW7CyjrJe_)dH5LE4g;5(<udd#1*v0&|KlOHXb;*|9K`Ad&nGu&l
z;xFM$xjma3Z!rF>Rwj)PDW_W=VJ<o<wlL>@#-gihIT`tbmtU;#j*mgLYZ)W4MCFo<
zEm7|)kCm=Gq0=m;g+zxho`ICD@^ac*V@vzLf1U1wXO;Q2>LoBe*1GaU99Y#>HQ(NZ
zw@DAbOk&ojXcY@OFl;1WF^S2PUQWIFSn0*}zT2dM%$dyKDarIV>W42<jMhzju-e?p
z*=SMxreS=t?UtM)&<<iLu1h9#=v+t^>#Gc(FFbu(ZD!~Vovv>0lo8adAJ0UhE;e+&
zRx22vnGqwfpPDem)`eBX{>yXjEAtZy<?_hty(CV}KegER$^o)ME`Q%kc0*7K+Gb^Z
za*&X-jzz(t-`xHqg@jx7h?o|R_Cv=0zSV0bza9gL^EVmQ=G^IC$aW-|;OBQwFkgT#
zM)>X;93_C!hdgF%HPLB+eiR+ztC;ziXC^}=ljqRx0}>4EX8tIJ<z{5n-4^2J{{wTP
zKN9$GWd6r`Ed_C^>u0JZQ&x|aEXSl{hVnbyMZkR8h<3Tl`H=eN*h6-zF;!QWWYD$G
zh!*vk0akbVl-15uBQ*vcHh`}Y5D)-*D#OC!VzOIM+>=1c^EoOJ=jUv?Z>`o++?6vQ
z8VC+uTAKWXP-WDAxyO*+oLm0Q?LOa`$HNFQ2~1ZwdC34wmMpej<}zag`7VNxNqsys
zOh<z4K*a5DUmc-1B83NU^=<#zYScXos1N|%0SU-^#9^Gv`Y(bH(I`NmUfvojP;K{-
zec=E6JN#hYW3$DvWNLU_L9<k^tiG-Wwk#W>n7oZ#RYEUC&ABHI7a<rd5Tl|Kr0<(;
zbXw_IK$OzlUfzWaTbyow9oJoBJcA`Urt_L&>qM29$tx*YZD%@iIsN){gZHZIgl=%|
zhbiF(tJZsUBAegk{lkPz{(a@l*b%yf-YX#;0G+E`J`%X}>RYjO8}b43IR$T8z}t?#
z2Y*b5&T7ru%Y**-rXX4sr4IW5)p#0UESV7nHT?T`e41dx!j{Co>k2DrC;3_e<oiU-
z=tpQkcB}hKHL)}4YNtn(K*(8EtBWu~TqsbxLYtt)DMHILLEfF|8O=hzQCOJk0VxR;
zV}5=?^Z{y38~Z-3Hn2nKj1=~qfJ+ur?(|bpIq5ts=evdYy-m4t5d}X!_crHf8=d~c
z4<a#p1`VbmY8iWtRm2|gteYI!w!y@lAU)ALc|*v37VQVS*$!1{qR1fuyaDAa<-$D{
zq0{%_9d9p&Js-c;yui0Ms2{?&HhVm2&j;Ag>%-fXA*4*GJYm@(JppN;MiPfE|FS=|
z!?MnH55wdSi<6zyf|}cfapMR=;G6OY&A!tHrI_Qc-1jT^9`U=XFEL|}&V-8n2R)i6
zSqyj$K&tJZMZ*{>NFxCB6m;8QjTqRv&S~*l{rOA1MV=tj(*a)$vdiKyO0rlcGS}|f
zSZ7qBssi%rNbO8;!bs)0oxhGD=htX3AeNH)5M3+Oc|U%B6p*j65eax9>%3co)$vw&
z0~K`10^EBOBYHgs%u_YzJdY2>=LEwtt$z&MUOx1XxK{u}MrqnSLB@e07=d0S`Olv}
zxA<G}_6)e16R5!p+wQgFvdCiVtFiz;?p4i>N3bve)U`Trs2Yi?w&X}H-ngU4H)VB^
z*F0}_hyrxen65vzf%=DtmFTN>1CV*s&gPF~K`%sGSpSa}0H{DiqKKsW`zB>Z>s21F
zyi(XfibW&DWd`loFJANgUEPS{Ek8qfmP!cpFf9y(#nA;Stzpq?#@;)8y;=bA(DtiF
zAdyk|(#=;ZxfquD?y`cb?PPNgnDYz|jd_joPb$fpS|+<tHU)6CjDb7^Gz()!yz}z6
zi+M+h>FHnf`Oa=DmxVmm)0pmh8=WJcbIS-<@Uw5lkH`J0Uh8dg0-{JkSlrXx_g~Q{
z1ncl<+{Wp%R?nOv8U;G;kwna^-c`kEdd+r4&K<ja<#`3bAnIV-j3S8=DMx`#11m5K
zHH0PNc_<2~Y(g7O_7mZW2??tT6=3&)fD0kD3!j939Fl^)mcdlBx{vwtEDo(0-yH@?
z?)$C_Xx}maDy6~f6qlcsDVLv0=-Z89^S7BAUREaB@ritOvf~6f-sucrn!F%#z8tZJ
z{2-}LbIjVZk^NR1r4`RI8KP-{l{y3f#`(_Bfq{oLra%_JWhewsQfue#{3eI-t$R2%
z`V*$3{KngDda>h+8KN2tXI`03$P~x`PPNcO1N;>~-n0|w78wx7@3&<*23UmiPOyiC
z0L=umet99Yc1u!tbmX~^=O8C#uu&m4a5zh|(W|}#u@<F&zS)vPs{xA0*F?n1f*Pre
z{yEGJtxQwMha@&?+rInrdAE3)$-bQ4*5LAT*{kh3NVuDrOkos}DQVfN)TEFzgP1XJ
z?ZCWiP~9@QV+Uiflkb^sxDRrgZ?)rbLxd(+$t|$0l_2F)9I3+!K4<)nf{7ZzFs|AI
z@U1$NC>>Y`z7%jYi%7S4&U&QMRDw*gJgNIGg6lG{0}iD%QX}AuXECN)SY~pZu`^(q
zys)UVM9M;rr%sqclq@12+b&uNX12e|E?Gfnc~^*lm5FKE<j8zxB<R>ma{tqH;2BML
zTxT>L&%A<mk&Pv<N^icwY<%wt5e3B<PkYV)$3mY;n?-PH&Bs_(Ba#1A*j0tK(Y8&f
zP@qUD?gfIoI~0c&cZZ-wi@TLVks`(2t+=E}aVZpvYjAf95}c54-?RVhKTC35*-dsI
znVEa;*@7bx8y)(-T<{;A3gbzaTV_nQ%&v45ask(iKg@ugPeU0~v;(0KQpgSTBF?u2
z(PgYUL^D&#WD((lp(p77^*sHt;vsGH%V;P;B{}aw&wH}n^D==fe&WamEFj75ZI@1{
zKsWbKBy`m`LP2LI@X;NXld43AK|@1RkT&u0t#m>hyp*YuN$x$W9MnYDjFpyyh2?^!
zb>Hy3v5Kvg1#`-uIQeabhqMwCGjGz+NXH(+(&ls|?1jts2A8h!q`jeR_q|;I@oboI
z*QMxQK=AhIBBh65Ig0`g_3!OPst!Lc7%aH2`uFescEzWR0P)X@ms;g|wNw=qhfh(Q
z$R}KJBqDxoR{TnusJOJGA$ScH<wA8AOcx5Z9h=KB=H{LhdtCq8GNZYcO(Esskvv|%
z!?8tu*);|003+o6u1;Rx3e+C2K-9+FQ-#zge<q|>l+CJE46Edb7N%RF;ZpW?OCh4_
z6mBVlo+T~BK)1jPn>!u{bELI@WmF?5(AGpEtE$+<Dn4&}w@_q@kpB9`lWBFk6h`k@
zOfBwAkmD}0H9>Lz8T}0*W%!WeK=p6m$<J~q%E1T19W_=43=9I_#E`IsdUu6?E!Q)d
zHNu4-6J=MUl6*RP=zoX=ZKiWW?EA^Y18)aKf1KkvF3p|Ll)Af9OA&;N|MtJ4R!aog
z(?d&S*|*QeBkZR6vZkhnYs<V&=3e46=+<JTik_uD`5->>+4PSb5>{UE(9+T>)}{{l
z?uBz**1hWr>3^N3tuH@Yfic)%Y!{g9_d(eG-ZA{HUdO_2tzH~?xt;eZX%?+{7ZBSs
zcx^pxY2TOC<ansnu?+__n}Zb<76~dH3$#5S#DDQH#^p?09tr-$9SnF{5EFMgo9GGg
zh?-VPAbdKX)dSl(=+;#K3R<#i>|tMyp%P7m+H}T2w@O-C^n*J;Kb^D3W(nAg?xp~Z
zrC|&t{5CIOBfi8TI;;U&^_~ubT<!yGs~HaheW&e|A-&(jlL^wo@JtdxeC^w(II7`P
zzK%RN8z|<(yVP(vkVV2q^nG>XWvP7B?Axn@ks)`+g{(Hc@y}UFfhwn0ExVD<siRdU
z5D7N+_a3CYtZB7V(+{{8cr{g3F<NDz=e4{L-UJW^uT7^JNj7jXCkfTLnHf%h`&X%6
ze33!py6yyQ%X;9usT5`3M^cox_?2%RszalR6{6@`+ASKxs}H))ma!1X+z$Wz!UB7H
zeMX+GS>nO5p;p<D>EiUrsMzx4{phN?vn5#HR1<9B!?3&)T#o%qn&vY2qsX1@rEA+)
zU5$==&nq#Es7KzHG&youWYRaL)I*~GP>}%!i6H0GY5b7FURs1G03`{%l@+svd^M5T
zIZ6Fhf;O8tfQyfFvpiEi$tD(0S{<;exCO5cMZg08Eb2Hcv;-wRJ^pTU69%>rX69Ku
z>9YYd(=vtHQ?}c`&5o-r;fROpqq*_F?vfj0ZKvIpe#{&!rNCX{geXl9Nbn$A<kY^$
z#DHkHQ48)?gO}#=`SsYD@xJ#{=*B0$h*p|Bvgv`Yp)J+)D>rW2L#=-9Iz;bFqqFJj
zlk+^AKk=HvcUYgu&Sn-NyzcWno<Y^oVs=k+6$#sSTC#mG8zb2Lak$v=n(Cjr`us02
zB=>2D*lc|&JKdTWH!hfITSC}}Nk*nfst?WY-v4vg_?8xx3_3~RUJiJC#``^WyN?C|
zeu9_3%fj#xJ+CP5abO1(rvtQ#j+GTMrSV=0ydn6yZ^}@>K3s;i!Uc{iG<3avI$QRf
zyPmuFPucneg&%a{r;-f9zs>7dTIMoq#To62s(3FJdx3+DoLr_L@0hw%FfGld<ENUz
zhv<*$%9z}f+%uJCddM*Vgbz9AsMrmv@~KQSd8W9XE(?df@e!jE^HEe%*{Cn0lFPUd
z5}>t%(nQ1_2_8^JzUhVfF6JY<gRm&XQ1BV9spb_f(bxmKeoM-KDP%fcfN<`EAo_~6
zy00U2DXy2wigkg+nS+n65y27^*}MEsK(PzN$ar9f6OmM%)im+hT=GB3T?Rjhng>N6
zh*>z#Tu$*(!CTRj!pstU>A$=h_*wo+uoqxC-w^M=Xq*9bpMdz~&|H&~Kt<_hPv13l
zC_~_$a}M*gnEtj4_CrFE!sO@o)VqNKc_&LjQ`7E?wp%@4Cte_4UG3c2ao!HMDI&<Q
z>-E{T_lUz*D%I9jeZq+=N2TLe#pj!$f{X1!PG}qvNu}(Wt{jqiKp#8=yduAezj**H
zqBi}&vatUS@@eA+9IKXZ;x)r%Y2&15a&-bF=eH6}FSYYHnF*hnAyCa7DXOcB+{HGU
z`sqq{4HNB_KvIa*8|vidJOKu1&C-_Jt)k#{yRJs9LzWWXK3zVCRf34Vp4H`!fG#&0
zKrw`$tVPVT)DxJpfnc&!;(qYW(}ccydHn^518${W4S0CBdIV`h#qTh?eXfveakpvh
zak`oT^jY8$n}yRU#2x6$X}Y}y7DOf%HrgUD{6$Ykg@>#~t!8IRKAKwrEA10kcW-T#
zeg-3hdMmj>*wpJ$vsYPyXiIu-S5_VF6MV6j3h_t2C!$R<=u(DmWa7@J15^FOfm#Dg
z^6HgdDsdDN(F+vIb(W2bidCs30^@@1rzww!7F(lq=w%M>+O@BCBdse^2LB44eeEPy
ze^co4y3B{(%B0csYpXQ63V}aoH1&=1Ya{<ve~=#V4zrAjsCijnJMn75i_)!kPG&Ed
zSIz9TH}I`nwJzc&^sHPnI(D6QL=OeR-7!T!aHUQEs=Cela7sflQTreh=FTa``Y)_<
z>R!lmBTQXcO6}$+H(i*K5T{Xu^e1rcI<6$*eKO`>dWTC^K{otzjPo?HB(~u~6<r6!
z%5x~O&{$+ulM_ooeDOBOl?@pQGrTHH2Ra||$jOZk{1xQnq7i?u(Ia$0hPF)=@YenG
zPxV$;H@yXfQJzLnZND7gP#>E)3bm8>J%Q`PE{+`awt3y`65ZQy2{uK#;{QE_`HW>y
zdyR`ncptz_3yFY^jjY2;9HL4&ERq?uWnSmxAjFS?uQj(zkGiJbcHJSUn<)#tiscDz
zDjs6}a{Uw%9{Pe9<a~}H?7hvgm-`8p1P@SdOAAZIk7_ch)<7To(C^U8cT$dyZ2S%-
zfNTob1ozj$YwaI2ey-I?FeG8nys$pyNd+v`g)ky~QqrLbaPvfqon?@O4|;*!JEfU&
zBl5PD8qsSINHtq%DEx8{&<4m3_;353C*Ap+GCtk;s&(pfaXQ+jzX=6aTZqqwCJvPl
zIv_A-D??IMhCXFW85&|E5G>1D>bo1=orrrxVqy|7JbdnQ!W%}^Bu~$rCT97_`0OJ`
zZ8~PMYMFp3Sm(MTuUl;#JH1h?Aur<xKa~1!2WiD@q-ZzIqjNYfqz3{G2(m`y+i--f
z_N)>N>2!j<<uruc@3D*ooUE@Z8VQJr-VzeZRf{wZj*J!iec&nQ56HcyZgSizJCSrd
z-p3frlZ>zMQ~ID<`Vp%7%YHhV#Z?C|l5K9j)nzjo>p@7m-@{3PbuqGTY4Fojy7Nk#
zdTo%&%B#ZLBb4{}q4!ruc;NW==si8pCAT=(<<O<BW(X>DRVMU!y<j79sr1@9C32NS
ztuj*5y*n&Bf6nn+K7Ndfi()A9vZOsuNNx?7PQ+e<PPWTI9Bi;n66t}gvZ1Wn&VGIc
zxY;x{AchdJp@*GT*9=nVT7|=i0u+GblL?N!n<7(6sEpJZ)^q<mv2m}t?){N?qWnlD
zEtyW~t<6FeHTUeq*=7gBwalx-o)}UL3_`+jofntr?eXJ;`_=H6X13w`j+f7{3DqJc
zX<bZ>69@>b5c&<@zA5-{EI4os^K`OWuh1O!f7iQw`kQTTtC|Ns*tCWp&wfolWLE5s
zY8PXOYcsn+U77h!$EwHC=Zl>~fkl6h%_@fH(wVI37sXpCR;8aIyrm=HG#?6tiRoHt
z>+}O4^RnK1?<u8brOnvQp3Oxqes|LQ#-zM`t&nQDC3y1YUzeIS%&o-#6&p(JnXXT1
z)ng#P$~0w!UhDN*`11!`x3ztYd0Iz+rGmOj2=#RUuR{R58Q0aW1bTdKE(6}R9}H3c
z5^S%o`PL&-M@@-Alrih!v@5M<rFHqoKYGDufUV|zK!DiZcph{VFhNN}h(`CPYl-94
ze6b)40VhuW{E<(?>V-rQaqmiJmJh9e<?}JH3ox`t)xCe4a+|9Mqv4XGKM}pjhjcgh
zaly6oq~J&0BVERwUyjHRgYRIUe|D2`xjgTF*NXaF5Cv-*fr)PNgu!-jn;P{Cxt}=J
zwe?%7w<`V3lR*i?FU*pJXM^LM-6n>Zx5*}e0X6}&mg%-Sq}6I4`#CRkQUI&ajPb>*
zy!*r4K>yp}<7Rz`8#XbADTuyAB=G(Qyjgs%U9P#432%CZ^Rr#dl0i3%aJCYN<}l)c
z@UnVP_;#3{ws;DOHo<@qI)8q1_BcG9V4a~t-t6LDW~yIBG{S%%9xf*&n=q*eoQk^W
zHh0C}2aX$G?>KZ(_t=i_>$vjxd^$HMt}B=~UIO7~B27i}!mB?n2BVDWPR`VT=7mPI
z8h4p2`+b=Ro<S-A#PHQ5?vrz`uL?Q)yiUS<ciigz_nfm;^XhCK;!aT4I`8OT6CXXG
zE#4fXkzhj5CTm~)4X1^0dQ&g3FPfZ(Nb@5Xwxc6~y$V;iT7MKEy}%0Ac6d>gJkLcq
z8?9PlvT#QWIa_jjmet{>(|V+$LBJp(gldq=%*tA1H17O_W`UXX@pP(Vn4UZsi$oS#
zOEG!_6Mlchu<$}xS!-I;_hi(-<6t%8HqvLTZ4BZ$t-Bf!hy$$nDQic90e~1O_c~No
zO--#YEQ&|N^%psiA$F6y^23@m9x4MAo-%irD#%$5Co9#|+-`O+&K2@DIoRpT0&!%V
zDJcX36F^bc-_tC4vY3FMqXt`j47h!BDUqYFTCSxW1jj9&RiqOXOk>Y#a=V=XM@L5k
z`GqspdTB<^UEzOj!|JLV7}d7x0lr5d2e`Zv<D8ZtZYwNA9Qt7lFPhg}$@j){6t4UZ
zeBISdtmV<Al%ixL`<cS>om8)_$B9W~yHFLKHI>B#?p$+9ftob9uc`4*e?KaK0f8Ku
z#kHB-F%yv@%&W|0HjQnT&H4KEgkjx`>a)#i4^CtQ{qS$6b*?<KUW<a6lTulNh8uJB
z>meJ+`AKNFa^R+c_(@|1gcP8E!M+dJ?U_*T{M|B2tYdNgE}M;um+*r*Go+Z)z4k8P
zTfT`SGuwbj3T@o58J<&8ih7>$YcBjrW-Pr}Kn9|9hx8!sgm$!6gD#zbl(4}Kc=>by
z&y5xe4#k?n2Yb&h$K)uesc-CFSqJ+2zL;;vBZvf^>f?b1O>Do2tfjy4)VTPSK6Z28
zLAok{H=rP{CIpHw{IkGo%k2J&+&!ym&PT3)lgMTS3|kI%2@W%J^}HF&*Uecvu{p#a
zPfSUY`)1~kyN>7z>J~H${t8lJ8U&|?tvKV-{bui#xA5SBbzV_9Em`vd1Sh@AoQgW|
z8uf~NZn`$hW9{p_Rn{8M+bT11ZA!y>#}ZYy5DuVUul(Z~*W_mScMVW~8Tp>aA?THU
zlNUFY>z0am3?9HELHH2c4m?oE!pMBWjb<T{=_JjrhY&LhrpOq`U|=gvjb)It+X4fx
z?FZ<;y~DbM)41i0{67}g9+`K0CU?f|E*eO{-W&D3(oJr7`#1vveyfO0&i~WHPNk>!
zX6<D^JCI~p@S5Tj2d{)c`fRF-IOB1GPNe5qrV~Ip)1IC`6q#Fq(I<0M8R=lZv-E7o
zK_ax7)ZR~F<cyy?BqNBhjWp$%$;f!#_C;VTqy|Xiyt<~AJr^7Y0{GO8NeGC}lYM=*
zYHQus=KY!72sJb{dpm!0<E?QTkF>a20xQaS5##<^*6sQWAmxgq?a!w-2(*yU-l;d9
z6C=mU+BJ@w%niM^)X3N&T=CFRqU;A@3W<YKiIw{ZyIun)UER-yQigtHx(ok`AE+!A
z5_ft!;54yys#o_tkP~#j<>j35y+E%8b=7EEcXMP9mx9GfYa^@Ea=7)LFcmG@LcP{2
zlSP*NAJ5h4s?m!6#!iWL9%r-V?Ej6=la=m0@NDqc<I0{J<fij^$PxD8!*w22kx!EM
z(Z-D>w^=8kuQGzf^+m+p!>2jx;{x)=3A3lKW>IZAgp(3(V)C_F`z;#Ze2y(MMYMKm
zg`1`{R?gAB$^IwW-~0J1k8!*tLQ)Vkgk6~SHLQ5}`k8)pC*)9Xn(H_rL0hdnZkF@x
z9qQkB%D)F58UUr#O5gsWLp}O3wB=T)qD&i?ra~GFra@um*Yt?84go$4e6DQGWeGm1
z@C4g=9?g}*@AH~q6Ug|uI0hDmf08K)LUpkGTqwW`X-^XFHH;(r%J=B$^_~YzO;E)v
z!0JxPgvHo!a^8eam#JR<p;J4*E4yR~EX{1RFT77Xzqowe8zvBeMHw(;uLuBlAZLqU
zn|R$VMe~@==DJD{Zkv6zTM42EnRr%-e%*&ccyXkprFA$=&k~L!s+YLH1=`vHV?uST
zs<Yb8U6xET5i1|od6W|O*WBL$#AD+K)<w(SF+l%>awb=jaWybf*G}S|ZEAX2V;n6;
z+<Q^Gt;hqmIXP^kDOO?}ql(vZ*Oi9ntC|K#XV+BF6}Y$rs~_HuwFU2qR__s4&No`7
z<wDy&bMo<BsOI(_ogY%_o`^K|DmrU9!jET+pe6C&zgy0qEk#wo9e(T@-CO^X_f_I$
zd%6|Aq$9=$E@*{GQgJ%SWS<8GJGP+D*?R1_`OL4wakVZ(&+Og|<&gj9-@!@mwK|Ys
zw(CxQArIVJ=4u>3)5__5s96?$3`>ze=pvJg<~vx^4z%+AR_OOdI`A+veK^B4NzuXi
zs0R$4Ev7v0a^|4B$BVP|IgIBn4>`?rU|P*MjEC3ew&-B|3f+F^W&n}_?yTvr??ctd
zNHH9vW2u{*Ur)E@0rMs74TW$3pRDh!a_kHl7*_9cPMfoRMK-hbc=ybv>xpc-R0Uh~
zXiWq{Xu8oKBV3i;`TSlpK{lM2(|gr30v24GR!?_2dr9RRz+?0MZorF=9F9X1xx0~J
zrxspaKbC<YDjzzC$>T~$D3aG=XO3!OUp$|V7p?gujbxa{#h9zXh4u&A-4EY?klHtU
zUwZ#uF(kCADWg;Mqd>jY##G%y%<T%(Mcf`Ya+tG)xe^dBdReD|<c}mPi_3xhWVZ*L
z6S(?R=Kw>*OPRBlVV6(z1a03lc=y;<^DyO??2;$Z3oxt#P<fv1Pj`8T%L4t)O->Nc
z2;*EyBGaVOSW>}@PKT%V7H}&vv$X9DLoony%j&l}^HELGf47Q?6cmzLZK&O9d}ZWM
z2H*kD{z?l=OQkrK{sO}A*e8=#&x?%R#J5C`g-&gb0MJxMz~#X=2P?ILThuMMfa{rA
zK%i!cMEaiuxL7;<-?|YR&s~osstbIdTN8gOFyh<e+=p&z4JftCrsFxY$w#Vyw<9|{
zt<SXR2~$$5jSjOc0m}(>(7gms_5L>~7FVQcO+AXo{(SDUJ%iKV4p>V~j(s~9qmxbo
ztV%;lSZxj@MSo-33tu(cRC^rG){>VZ0vpfzX_KDHy@i5~U&N9LRv7<P(K%R!kh56(
zIbOyvkOHcoi`E{nwe$gMi`~q;!m$#rWH*eFRXJ!M?}G^k?Xj199BI!lW-ZsBIC&`{
zG1EJ=MM`L@5lNj0|Dx!3%={`vejhwn{a`AkDj9LZ2Y>7#Jo~9qF_e<u+ML_SL2HXL
zCnB9pEuz7Ta=`4Vlau|S&cD&%S#C!*YRESmwc2SJ_w*<ju+*ucV_wVG@G&urmNi+3
zBKbbe3oxuEXe9U*wN8mEM`K_d5dYnLD{ZbDfGtv&a+ww&sD7G45pT**BB`2p^v&@Z
zW`|lx9bo>})MEzB@<2jlU3G22IHh0h#2&8D1<)?SQTDoW^n-`TFK4-Ar1N^D%X)bS
zVa)Cq9@-s6!(aBJI9(%`<=+`2<KjA9hn#+Po30Gj*t6i_3D8RK9_x{Z)2+uMZHI9^
zaJpu;AL)0{PE9)LFawbnbxSAnh=S68N88~N-go}tL34buk&)=)K}Pwi!)oEugch6;
z04~!qHvXdk3$99fOo5s)a0r|ddnbQ8aPIErkDQv)JZ-amW-Hxmb3;C#hJm|m`(ycC
z>zAXKnH8?Eu=*C`64`k=K<_LJ9uG?XNsk7s_~b37=JCHS_mkqoN`(0$q%PdJ<|eU8
zxFVeae*SeI4eS4Xxvb9RHGo(YuVl0A4@EJqI5J!9Aw)-j-uRs50~%{)w4t3xsu{jC
zI&CsKs9R-0hZ={+{0*Y(?QmMnI>K9`*nalKYC1xs#E|*r-OJWn0L%bMsbPbP{`!nA
z`+G9k;z_WFNo5A$+QtEW{1y8Dc?GT&MJ-|EG7(ve_$9KFpVv`niwgGv!RK3i%a&j;
z44ppofjV7Qsz+Epo>D2zIVW}iIOV;}diHG*E)uuRd*>p}d#!kclfDBmZwxGpt9E$}
zRB~PbW46#PljG++a#xB%)D0^HP(i$omsBkpWeErXx-W{-dULD_ob@^8THscau&QEY
z8$+b2JSn?Tvw3g&10ZLoRf`IkBx!t#6>j|!T&<v^%yt*V44lxS(cr+<UC>0JJirop
zIdhPX2&_}$=0<+M6B+6$qoWQX=IkPui18nYEe-nWN0>~(agSo2z&>v+n&1Y)b9>Ri
z_Ns9UVS!7u5NaSmr<54fVbWxor~-yQOSguY3Gr%n+3tI;PkS7#LPT}5QdUI0As$1T
z%F*0o0Hz)@?-}VR3alTnhnA~UlQptXi~_lb;l!^qgpYy~tVwmWNF{fJ#9?ngqDuKy
zduUE;MaBFwyCwxGNhi{miXpHXk7nq3SiWU9oftIRbf|h6q~3v#1WJ`N{{6BG-aTFj
z$j#tMDYj<@>8V+tzM=-}RJB%oFZ$E-k=L};qz&hfZd9Lhe5p>Q@@^0pv7CknFhgZe
zei^pP&U6$pnD<!lVF5ceu^aB`4iK%3@}E{jm7OnUnzuDgn};}b%vL3K(5)`1tNW3j
zPQsu}W+wH@0z8msX0fO>UVjpCk{0-_J;}G2ac&n7K;7VH+<?Cf2&*o!Ei)D4+W-T9
zBh^aa^K$RycxSh}8ZraeLAVhM#DnAGd&dWmWl=vdUfZz?yZs@7QZeonM(V)KQSxSW
zZT*<x9gDy5qAlL=7k-`x_!1RGZsYSBDE|Sc+`Jny?7nyRSK(h@;T+;c0N|CBA(JG!
z4fgx{ZJ19SwP??VjDnBTu4C96ENWMU@Wt(d#Obu@fcRuO(s>82J<Kr;`Wj5uInE;G
z79YRGB%oIedb7}AcSL_vuRlT|)|t2R>g%NOF49Q?NR%kh05nO{;3NLHT=6iy7G@^l
zz7G6>XLZh_tz37uIzQpZD>BFcbEH%e7`We^E*!}0r(ndlE0$T(W{VPU$-EhOB$!UW
zrF&1CTr}vot5_x1{}ZnN&8_r1@6)-c$H#!(;JcTjMosUSg#l0c?R6H63x73340+*E
zQN}Rc;*5`mbuA~?;Zy&n0Asl#OX2B^(Xh$Bhn2Ja-$X&P`%PU!w#W{|m`~-0_l0(P
ze16c^`yLTyCN8pp+8sKcigl>p#Y7<r%Jbp>o)Pdx5EPWD(_|_jHjy+gHpb{KX{}Z>
z{>v$6A#+ZJ=Pq6C>EI%s`nAYqXC{*0Y!rE~S!s!sBIvCS&1{^)lg?K%1-v01y_heE
z`g?)M$DZ%wqbTo0L@+C7D-?dBp852KQ?h?>F+IJikniJ)3<{(fpl*5d4(3RugI}pD
zA}v_4@?!}`@^3Hii82KF@Yq)weds4F8$L)SDY3)t{IF7u&R!gVQD|x5g1?HIqUVc!
zM1q5;N(wk*|C)Zat1M2SV>p}n1Pfa`n#pL#Jo*y?ZZ#tYNIo>G0VrNtm0whN+jEqP
zBiX(>lMW(qRX~mMhg}p;%#|vo5;srx-)o~{V@rIucKPg=BY*BQ0s8rSSB-P}9pjpr
z(8v4(ljb1G3dWK*l=!1s@e(?&mPuA~4|Km36J-rYJ%Afb2aGzPIDH_plNZbrZ-j<h
zK6A+t_Ku%DC4Mgu{+Zbb{mZuj;@3}O`4TeojqZZ6_WV6o7uBAX6(0-?NX<C!6Jug3
z;~l3eR6Zh*8N4=ti^6?EI~LtvHXO>z6mUJXCo-AU{K<w7%Mut%=7LKZ@Vd|90q4E+
z1j3=`e}G!S(1^24QH=a2ZHiqbAaBhO+px)5sZxObir*7E_8Teo(8w65z&=z0NAS~3
z2tc2v9PZdKY7a|*a@Gjq_<4hS0{~)$a#VrbSkJ?`-KPkKsVX_%7omUbqkt^Xm^p*$
zKs;^4&P<I~mViOE-CQAaA_y#yVF2+ZBvs*gBfaj{3KT#ya~^_00FcO7aEermgsv4#
zXbeutb#})lQFT(cEjaVVm|@5(mYP8t=%a*;blOchzv~7l02&S)9>Q$4&@|C)L$v1k
zY&Mt)RM$u}VVS0vOw>XY&~{>F1p>Y3oRn}OK|^O|B&qHv{-7i;eI3ZStsYIw%EWX}
zL#Cs{$`*`R#D*U(zJe@_ywvN#rrv-{uC7W;^N`Zr*M^~>G6JFwFqO87hvu!@a?9)w
z>1pErdB5#B`c((-_LlK3NKt!T{OUq1GlHiY3lI^9C<bu@b0=W-nVcPuL%cd#SN<=b
zsiktRaV*^d2SHcia-Dk7A%R8W4nK0{BbMgcDk1i{DlrJjm+f%vy%W)XB^tDds_4Vi
zU)>t}oE^}m%o@_{tnbSk$;%4>NF>9G^WeA+{R(a+=Q4k@5`y|1?Z2)5|GYe7?RoK*
zzMr0ei0Dgwja_&56>sz~4^9L==*KrKQrc(a^Z*`k2?QQb0f_KO+%NK7XX$k7$-vV<
OP?Y3UWox9(L;nXf09F_P
literal 0
HcmV?d00001
diff --git a/doc/images/olena_layers.svg b/doc/images/olena_layers.svg
new file mode 100644
index 0000000..8ced02b
--- /dev/null
+++ b/doc/images/olena_layers.svg
@@ -0,0 +1,368 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!-- Created with Inkscape (http://www.inkscape.org/) -->
+
+<svg
+ xmlns:dc="http://purl.org/dc/elements/1.1/"
+ xmlns:cc="http://creativecommons.org/ns#"
+ xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+ xmlns:svg="http://www.w3.org/2000/svg"
+ xmlns="http://www.w3.org/2000/svg"
+ xmlns:xlink="http://www.w3.org/1999/xlink"
+ xmlns:sodipodi="http://sodipodi.sourceforge.net/DTD/sodipodi-0.dtd"
+ xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"
+ width="783.75262"
+ height="786.33826"
+ id="svg5663"
+ version="1.1"
+ inkscape:version="0.47 r22583"
+ sodipodi:docname="olena_layers.png">
+ <defs
+ id="defs5665">
+ <marker
+ inkscape:stockid="Arrow1Lstart"
+ orient="auto"
+ refY="0"
+ refX="0"
+ id="Arrow1Lstart"
+ style="overflow:visible">
+ <path
+ id="path11891"
+ d="M 0,0 5,-5 -12.5,0 5,5 0,0 z"
+ style="fill-rule:evenodd;stroke:#000000;stroke-width:1pt;marker-start:none"
+ transform="matrix(0.8,0,0,0.8,10,0)" />
+ </marker>
+ <inkscape:perspective
+ sodipodi:type="inkscape:persp3d"
+ inkscape:vp_x="191.42857 : -725.24748 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_z="921.23734 : -648.10462 : 1"
+ inkscape:persp3d-origin="372.04724 : 350.78739 : 1"
+ id="perspective5671" />
+ <inkscape:perspective
+ id="perspective5620"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <linearGradient
+ inkscape:collect="always"
+ xlink:href="#linearGradient5430"
+ id="linearGradient5446"
+ gradientUnits="userSpaceOnUse"
+ x1="1765.6388"
+ y1="84.601952"
+ x2="2334.2168"
+ y2="84.601952" />
+ <linearGradient
+ inkscape:collect="always"
+ id="linearGradient5430">
+ <stop
+ style="stop-color:#000000;stop-opacity:1;"
+ offset="0"
+ id="stop5432" />
+ <stop
+ style="stop-color:#000000;stop-opacity:0;"
+ offset="1"
+ id="stop5434" />
+ </linearGradient>
+ <inkscape:perspective
+ id="perspective10885"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective11022"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective11679"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective11679-9"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective11708"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective11729"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ <inkscape:perspective
+ id="perspective12619"
+ inkscape:persp3d-origin="0.5 : 0.33333333 : 1"
+ inkscape:vp_z="1 : 0.5 : 1"
+ inkscape:vp_y="0 : 1000 : 0"
+ inkscape:vp_x="0 : 0.5 : 1"
+ sodipodi:type="inkscape:persp3d" />
+ </defs>
+ <sodipodi:namedview
+ id="base"
+ pagecolor="#ffffff"
+ bordercolor="#666666"
+ borderopacity="1.0"
+ inkscape:pageopacity="0.0"
+ inkscape:pageshadow="2"
+ inkscape:zoom="0.35"
+ inkscape:cx="379.83489"
+ inkscape:cy="193.79466"
+ inkscape:document-units="px"
+ inkscape:current-layer="layer1"
+ showgrid="false"
+ inkscape:window-width="968"
+ inkscape:window-height="805"
+ inkscape:window-x="302"
+ inkscape:window-y="265"
+ inkscape:window-maximized="0"
+ showguides="true"
+ inkscape:guide-bbox="true"
+ inkscape:snap-bbox="true" />
+ <metadata
+ id="metadata5668">
+ <rdf:RDF>
+ <cc:Work
+ rdf:about="">
+ <dc:format>image/svg+xml</dc:format>
+ <dc:type
+ rdf:resource="http://purl.org/dc/dcmitype/StillImage" />
+ <dc:title></dc:title>
+ </cc:Work>
+ </rdf:RDF>
+ </metadata>
+ <g
+ inkscape:label="Calque 1"
+ inkscape:groupmode="layer"
+ id="layer1"
+ transform="translate(23.332153,-163.39523)">
+ <text
+ id="text5361"
+ y="279.38019"
+ x="669.5874"
+ style="font-size:12px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ xml:space="preserve"><tspan
+ y="279.38019"
+ x="669.5874"
+ id="tspan5363"
+ sodipodi:role="line"
+ style="fill:#000000;fill-opacity:1">`</tspan></text>
+ <flowRoot
+ xml:space="preserve"
+ id="flowRoot11041"
+ style="font-size:12px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ transform="translate(-29.759501,70.733897)"><flowRegion
+ id="flowRegion11043"><rect
+ id="rect11045"
+ width="354.28571"
+ height="14.285714"
+ x="174.28572"
+ y="831.828" /></flowRegion><flowPara
+ id="flowPara11047" /></flowRoot> <g
+ id="g12633">
+ <rect
+ style="opacity:0.67790264;color:#000000;fill:#effcea;fill-opacity:1;fill-rule:evenodd;stroke:#000000;stroke-width:4.09148645;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;stroke-dashoffset:0;marker:none;visibility:visible;display:inline;overflow:visible;enable-background:accumulate"
+ id="rect5292"
+ width="779.66113"
+ height="270.79654"
+ x="-21.286409"
+ y="676.89117" />
+ <text
+ id="text10963"
+ y="759.70477"
+ x="368.92697"
+ style="font-size:12px;font-style:normal;font-weight:normal;text-align:end;text-anchor:end;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ xml:space="preserve"><tspan
+ style="font-size:56px;font-style:italic;font-weight:bold;text-align:center;text-anchor:middle;-inkscape-font-specification:Bitstream Vera Sans Bold Oblique"
+ y="759.70477"
+ x="368.92697"
+ id="tspan10965"
+ sodipodi:role="line">Image Processing</tspan><tspan
+ id="tspan11039"
+ style="font-size:56px;font-style:italic;font-weight:bold;text-align:center;text-anchor:middle;-inkscape-font-specification:Bitstream Vera Sans Bold Oblique"
+ y="829.70477"
+ x="368.92697"
+ sodipodi:role="line"> C++ Library</tspan></text>
+ <text
+ id="text11051"
+ y="913.99048"
+ x="256.60666"
+ style="font-size:12px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ xml:space="preserve"><tspan
+ style="font-size:48px;font-style:normal;font-weight:bold;-inkscape-font-specification:Bitstream Vera Sans Bold"
+ y="913.99048"
+ x="256.60666"
+ id="tspan11053"
+ sodipodi:role="line">(Milena)</tspan></text>
+ </g>
+ <g
+ id="g3086">
+ <g
+ transform="translate(-1.3471178,19.316265)"
+ id="g12641">
+ <rect
+ style="opacity:0.67790264;color:#000000;fill:#f7efe8;fill-opacity:1;fill-rule:evenodd;stroke:#000000;stroke-width:3.14199305;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;stroke-dashoffset:0;marker:none;visibility:visible;display:inline;overflow:visible;enable-background:accumulate"
+ id="rect5294"
+ width="550.55963"
+ height="226.127"
+ x="-16.004351"
+ y="399.29413" />
+ <text
+ xml:space="preserve"
+ style="font-size:12px;font-style:normal;font-weight:normal;line-height:125%;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ x="136.32234"
+ y="579.04132"
+ id="text5341"
+ sodipodi:linespacing="125%"><tspan
+ sodipodi:role="line"
+ id="tspan5343"
+ x="136.32234"
+ y="579.04132"
+ style="font-size:48px;font-style:normal;font-variant:normal;font-weight:bold;font-stretch:normal;text-align:start;line-height:125%;writing-mode:lr-tb;text-anchor:start;font-family:Bitstream Vera Sans;-inkscape-font-specification:Bitstream Vera Sans Bold">(SCRIBO)</tspan></text>
+ <text
+ xml:space="preserve"
+ style="font-size:12px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ x="21.480543"
+ y="484.03006"
+ id="text10991"><tspan
+ sodipodi:role="line"
+ id="tspan10993"
+ x="21.480543"
+ y="484.03006"
+ style="font-size:56px;font-style:italic;font-weight:bold;-inkscape-font-specification:Bitstream Vera Sans Bold Oblique">DIA framework</tspan></text>
+ </g>
+ <g
+ transform="translate(10.327175,22.857142)"
+ id="g12669">
+ <rect
+ y="395.1326"
+ x="548.69751"
+ height="227.3683"
+ width="200.37236"
+ id="rect12650"
+ style="opacity:0.67790264;color:#000000;fill:#f7efe8;fill-opacity:1;fill-rule:evenodd;stroke:#000000;stroke-width:1.90068746;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;stroke-dashoffset:0;marker:none;visibility:visible;display:inline;overflow:visible;enable-background:accumulate" />
+ <text
+ xml:space="preserve"
+ style="font-size:12px;font-style:normal;font-weight:normal;fill:#000000;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ x="631.42859"
+ y="447.61859"
+ id="text12665"
+ transform="translate(-33.978474,66.489174)"><tspan
+ sodipodi:role="line"
+ id="tspan12667"
+ x="631.42859"
+ y="447.61859"
+ style="font-size:56px;font-weight:bold;-inkscape-font-specification:Bitstream Vera Sans Bold">. . .</tspan></text>
+ </g>
+ </g>
+ <g
+ id="g3056">
+ <g
+ id="g3041">
+ <g
+ id="g3025">
+ <rect
+ y="164.45264"
+ x="254.95554"
+ height="224.35983"
+ width="251.51389"
+ id="rect5300"
+ style="opacity:0.67790264;color:#000000;fill:#ffffff;fill-opacity:1;fill-rule:evenodd;stroke:#000000;stroke-width:2.11477399;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;stroke-dashoffset:0;marker:none;visibility:visible;display:inline;overflow:visible;enable-background:accumulate" />
+ <text
+ xml:space="preserve"
+ style="font-size:12px;font-style:normal;font-weight:normal;line-height:125%;fill:#000080;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ x="347.98691"
+ y="347.86737"
+ id="text5357"
+ sodipodi:linespacing="125%"><tspan
+ sodipodi:role="line"
+ id="tspan5359"
+ x="347.98691"
+ y="347.86737"
+ style="font-size:40px;font-style:normal;font-variant:normal;font-weight:normal;font-stretch:normal;text-align:start;line-height:125%;writing-mode:lr-tb;text-anchor:start;fill:#000080;fill-opacity:1;font-family:Bitstream Vera Sans;-inkscape-font-specification:Bitstream Vera Sans">GUI</tspan></text>
+ </g>
+ <g
+ id="g3036">
+ <rect
+ style="opacity:0.67790264;color:#000000;fill:#ffffff;fill-opacity:1;fill-rule:evenodd;stroke:#000000;stroke-width:2.02970839;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;stroke-dashoffset:0;marker:none;visibility:visible;display:inline;overflow:visible;enable-background:accumulate"
+ id="rect12725"
+ width="231.59895"
+ height="224.44492"
+ x="527.73358"
+ y="164.41008" />
+ <text
+ sodipodi:linespacing="125%"
+ id="text5388"
+ y="347.86734"
+ x="604.16968"
+ style="font-size:12px;font-style:normal;font-weight:normal;line-height:125%;fill:#000080;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ xml:space="preserve"><tspan
+ style="font-size:40px;font-style:normal;font-variant:normal;font-weight:normal;font-stretch:normal;text-align:start;line-height:125%;writing-mode:lr-tb;text-anchor:start;fill:#000080;fill-opacity:1;font-family:Bitstream Vera Sans;-inkscape-font-specification:Bitstream Vera Sans"
+ y="347.86734"
+ x="604.16968"
+ id="tspan5390"
+ sodipodi:role="line">CLI</tspan></text>
+ </g>
+ <g
+ id="g3030">
+ <rect
+ style="opacity:0.67790264;color:#000000;fill:#ffffff;fill-opacity:1;fill-rule:evenodd;stroke:#000000;stroke-width:2.11477399;stroke-linecap:butt;stroke-linejoin:miter;stroke-miterlimit:4;stroke-opacity:1;stroke-dasharray:none;stroke-dashoffset:0;marker:none;visibility:visible;display:inline;overflow:visible;enable-background:accumulate"
+ id="rect12723"
+ width="251.51389"
+ height="224.35983"
+ x="-17.865078"
+ y="164.45264" />
+ <text
+ sodipodi:linespacing="125%"
+ id="text5392"
+ y="306.07605"
+ x="102.67558"
+ style="font-size:12px;font-style:normal;font-weight:normal;text-align:center;line-height:125%;text-anchor:middle;fill:#000080;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ xml:space="preserve"><tspan
+ style="font-size:40px;font-style:normal;font-variant:normal;font-weight:normal;font-stretch:normal;text-align:center;line-height:125%;writing-mode:lr-tb;text-anchor:middle;fill:#000080;fill-opacity:1;font-family:Bitstream Vera Sans;-inkscape-font-specification:Bitstream Vera Sans"
+ y="306.07605"
+ x="102.67558"
+ id="tspan5394"
+ sodipodi:role="line">Web</tspan><tspan
+ id="tspan12708"
+ style="font-size:40px;font-style:normal;font-variant:normal;font-weight:normal;font-stretch:normal;text-align:center;line-height:125%;writing-mode:lr-tb;text-anchor:middle;fill:#000080;fill-opacity:1;font-family:Bitstream Vera Sans;-inkscape-font-specification:Bitstream Vera Sans"
+ y="356.07605"
+ x="102.67558"
+ sodipodi:role="line">services</tspan></text>
+ </g>
+ </g>
+ <flowRoot
+ xml:space="preserve"
+ id="flowRoot5349"
+ style="font-size:12px;font-style:normal;font-weight:normal;fill:#241c1c;fill-opacity:1;stroke:none;font-family:Bitstream Vera Sans"
+ transform="translate(-1579.3038,-1749.8142)"><flowRegion
+ style="fill:url(#linearGradient5446);fill-opacity:1"
+ id="flowRegion5351"><rect
+ style="fill:#241c1c;fill-opacity:1"
+ id="rect5353"
+ width="560.05524"
+ height="110.27995"
+ x="1765.7467"
+ y="1933.4031" /></flowRegion><flowPara
+ id="flowPara5355"
+ style="font-size:56px;font-style:normal;font-variant:normal;font-weight:bold;font-stretch:normal;text-align:start;line-height:125%;writing-mode:lr-tb;text-anchor:start;fill:#241c1c;fill-opacity:1;font-family:Bitstream Vera Sans;-inkscape-font-specification:Bitstream Vera Sans Bold">Applications</flowPara></flowRoot> </g>
+ </g>
+</svg>
diff --git a/doc/mainpage.dox b/doc/mainpage.dox
index 9ebda3b..4d72915 100644
--- a/doc/mainpage.dox
+++ b/doc/mainpage.dox
@@ -12,7 +12,7 @@
</tr>
<tr>
<td>
- \li What is Olena ?
+ \li \ref about
\li <a class="el" href="http://www.lrde.epita.fr/cgi-bin/twiki/view/Olena/Publications">Bibliography</a>
\li <a class="el" href="a00028.html">Contributors</a>
</td>
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-70-g596d824 Add more content in general documentation.
by Guillaume Lazzara
* doc/Makefile.am: New dependencies.
* doc/license_commercial.dox,
* doc/mainpage.dox,
* doc/modules_list.dox: More content.
* doc/get_sources.dox,
* doc/online_resources.dox: New.
---
ChangeLog | 13 +++++++++++++
doc/Makefile.am | 2 ++
doc/get_sources.dox | 37 +++++++++++++++++++++++++++++++++++++
doc/license_commercial.dox | 5 +++--
doc/mainpage.dox | 4 ++--
doc/modules_list.dox | 2 +-
doc/online_resources.dox | 34 ++++++++++++++++++++++++++++++++++
7 files changed, 92 insertions(+), 5 deletions(-)
create mode 100644 doc/get_sources.dox
create mode 100644 doc/online_resources.dox
diff --git a/ChangeLog b/ChangeLog
index 2344721..5379982 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,16 @@
+2011-10-28 Guillaume Lazzara <z(a)lrde.epita.fr>
+
+ Add more content in general documentation.
+
+ * doc/Makefile.am: New dependencies.
+
+ * doc/license_commercial.dox,
+ * doc/mainpage.dox,
+ * doc/modules_list.dox: More content.
+
+ * doc/get_sources.dox,
+ * doc/online_resources.dox: New.
+
2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
Add details about licensing.
diff --git a/doc/Makefile.am b/doc/Makefile.am
index 410316b..81c34b1 100644
--- a/doc/Makefile.am
+++ b/doc/Makefile.am
@@ -64,6 +64,8 @@ REFMAN_deps = $(srcdir)/footer.html \
$(srcdir)/bibliography.dox \
$(srcdir)/license.dox \
$(srcdir)/license_commercial.dox \
+ $(srcdir)/online_resources.dox \
+ $(srcdir)/get_sources.dox \
contributors.html \
lrde_olena.html
diff --git a/doc/get_sources.dox b/doc/get_sources.dox
new file mode 100644
index 0000000..23f1446
--- /dev/null
+++ b/doc/get_sources.dox
@@ -0,0 +1,37 @@
+/*! \page getsources Get the Sources
+
+If you are interested in developping for Olena, you are definitely
+interested in using the development version from our Git Repository.
+
+To successfuly compile and use the development version, you will need the following additional tools:
+
+\li Git 1.5 or higher
+\li Autoconf 2.61 or higher
+\li Automake 1.10 or higher
+\li Libtool 1.5.22 or higher
+\li Doxygen 1.5.6 or higher
+\li ImageMagick 6.3.7 or higher
+\li HeVeA 1.10
+\li LaTeX2HTML (or at minimum, its html.sty style file)
+
+The Git repository is located at git://git.lrde.epita.fr/olena.
+
+For those who are not used to Git, here are the required commands to get and build a fresh source tree:
+\verbatim
+ git clone git://git.lrde.epita.fr/olena
+ cd olena
+ ./bootstrap
+ mkdir _build
+ cd _build
+ ../configure
+ make
+\endverbatim
+
+The last command may take some time.
+
+If you want to install Olena, type thereafter (as root):
+\verbatim
+ make install
+\endverbatim
+
+*/
\ No newline at end of file
diff --git a/doc/license_commercial.dox b/doc/license_commercial.dox
index 0ece2b5..af2f094 100644
--- a/doc/license_commercial.dox
+++ b/doc/license_commercial.dox
@@ -1,6 +1,7 @@
/*! \page commercial "Commercial Use"
- < Under construction >
+ Even though Olena is released under GPL v2, using Olena in a close
+ source commercial products may be possible.
- Please contact us : olena(a)lrde.epita.fr
+ If you are interested, please contact us at olena(a)lrde.epita.fr
*/
\ No newline at end of file
diff --git a/doc/mainpage.dox b/doc/mainpage.dox
index ea15c32..9ebda3b 100644
--- a/doc/mainpage.dox
+++ b/doc/mainpage.dox
@@ -38,8 +38,8 @@
<tr>
<td>
\li Report Bugs and Make Suggestions
- \li Get the Sources
- \li Online Resources
+ \li \ref getsources
+ \li \ref onlineres
</td>
<td>
\li \ref gplv2 "GNU GPL"
diff --git a/doc/modules_list.dox b/doc/modules_list.dox
index e3f2b8d..6b872ef 100644
--- a/doc/modules_list.dox
+++ b/doc/modules_list.dox
@@ -1,4 +1,4 @@
-/** \page moduleslist List of available modules
+/** \page moduleslist List of Available Modules
As a platform, Olena is divided in several modules:
diff --git a/doc/online_resources.dox b/doc/online_resources.dox
new file mode 100644
index 0000000..32cd18d
--- /dev/null
+++ b/doc/online_resources.dox
@@ -0,0 +1,34 @@
+/*! \page onlineres Online Resources
+
+\section Trac
+
+In order to keep traces of bugs and feature requests, we provide a
+Trac at <a
+href="https://trac.lrde.org/olena">https://trac.lrde.org/olena</a>.
+
+\section Gitweb
+
+Source code is freely browseable through a web interface at this
+address : <a
+href="http://git.lrde.epita.fr/?p=olena.git">http://git.lrde.epita.fr/?p=olena.git</a>
+
+It is also a way to keep an eye on current develop.
+
+\section BuildBot
+
+During the development process, we need to be sure that changes don't
+break the whole code. To do so, we have a buildfarm running several
+architectures and compilers. A monitoring interface is available here
+: <a href="https://buildfarm.lrde.org/buildfarm/oln/">https://buildfarm.lrde.org/buildfarm/oln/</a>
+
+\section Mailing-list
+
+A general mailing-list is available and open for any questions related
+to Olena. You can :
+
+\li <a href="https://www.lrde.epita.fr/mailman/listinfo/olena">subscribe to our mailing-list</a>
+\li <a href="https://www.lrde.epita.fr/pipermail/olena/">fetch the archives of the list</a>
+\li send any questions to: <a href="mailto::olena@lrde.epita.fr">olena(a)lrde.epita.fr</a>
+
+
+*/
\ No newline at end of file
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-69-g397df18 Add details about licensing.
by Guillaume Lazzara
* doc/Doxyfile.in
* doc/Makefile.am
* doc/license.dox
* doc/license_commercial.dox
* doc/mainpage.dox
---
ChangeLog | 10 ++++++++++
doc/Doxyfile.in | 3 ++-
doc/Makefile.am | 2 ++
doc/license.dox | 15 +++++++++++++++
doc/license_commercial.dox | 6 ++++++
doc/mainpage.dox | 4 ++--
6 files changed, 37 insertions(+), 3 deletions(-)
create mode 100644 doc/license.dox
create mode 100644 doc/license_commercial.dox
diff --git a/ChangeLog b/ChangeLog
index 98ad275..2344721 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,15 @@
2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+ Add details about licensing.
+
+ * doc/Doxyfile.in
+ * doc/Makefile.am
+ * doc/license.dox
+ * doc/license_commercial.dox
+ * doc/mainpage.dox
+
+2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+
* doc/Doxyfile.in: Do not warn if undocumented.
2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
diff --git a/doc/Doxyfile.in b/doc/Doxyfile.in
index 8474dc6..b53bbe8 100644
--- a/doc/Doxyfile.in
+++ b/doc/Doxyfile.in
@@ -100,7 +100,8 @@ EXCLUDE_SYMLINKS = YES
EXCLUDE_PATTERNS = *spe.hh
EXCLUDE_SYMBOLS = *::internal* *_ mln::trait::*
EXAMPLE_PATH = @top_srcdir@/doc/ \
- @builddir@
+ @builddir@ \
+ @top_srcdir@
EXAMPLE_PATTERNS = *.cc \
*.cc.raw \
*.txt
diff --git a/doc/Makefile.am b/doc/Makefile.am
index 1bf95d5..410316b 100644
--- a/doc/Makefile.am
+++ b/doc/Makefile.am
@@ -62,6 +62,8 @@ REFMAN_deps = $(srcdir)/footer.html \
$(srcdir)/olena.bib \
$(srcdir)/contributors.dox \
$(srcdir)/bibliography.dox \
+ $(srcdir)/license.dox \
+ $(srcdir)/license_commercial.dox \
contributors.html \
lrde_olena.html
diff --git a/doc/license.dox b/doc/license.dox
new file mode 100644
index 0000000..eb9a8e4
--- /dev/null
+++ b/doc/license.dox
@@ -0,0 +1,15 @@
+/*! \page gplv2 "GNU General Public License (GPL)"
+
+Olena and its modules are Copyright (C) 2011 EPITA Research and
+Development Laboratory (LRDE).
+<br>Olena and its modules are released under GNU GPL v2.
+
+
+\section gplv2sec "The GNU General Public License (Version 2)"
+
+Reference : <a href="http://www.gnu.org/licenses/gpl.html">GNU General
+Public License</a>
+
+\verbinclude COPYING
+
+*/
\ No newline at end of file
diff --git a/doc/license_commercial.dox b/doc/license_commercial.dox
new file mode 100644
index 0000000..0ece2b5
--- /dev/null
+++ b/doc/license_commercial.dox
@@ -0,0 +1,6 @@
+/*! \page commercial "Commercial Use"
+
+ < Under construction >
+
+ Please contact us : olena(a)lrde.epita.fr
+*/
\ No newline at end of file
diff --git a/doc/mainpage.dox b/doc/mainpage.dox
index 8f49ca3..ea15c32 100644
--- a/doc/mainpage.dox
+++ b/doc/mainpage.dox
@@ -42,8 +42,8 @@
\li Online Resources
</td>
<td>
- \li GNU GPL
- \li Commercial Use
+ \li \ref gplv2 "GNU GPL"
+ \li \ref commercial "Commercial Use"
</td>
<td> </td>
</tr>
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-68-g5375506 doc/Doxyfile.in: Do not warn about undocumented code.
by Guillaume Lazzara
---
ChangeLog | 4 ++++
doc/Doxyfile.in | 2 +-
2 files changed, 5 insertions(+), 1 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index d796c0b..98ad275 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2012-02-27 Guillaume Lazzara <z(a)lrde.epita.fr>
+
+ * doc/Doxyfile.in: Do not warn if undocumented.
+
2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
Add new groups in Scribo documentation.
diff --git a/doc/Doxyfile.in b/doc/Doxyfile.in
index 86377ef..8474dc6 100644
--- a/doc/Doxyfile.in
+++ b/doc/Doxyfile.in
@@ -83,7 +83,7 @@ FILE_VERSION_FILTER =
#---------------------------------------------------------------------------
QUIET = YES
WARNINGS = YES
-WARN_IF_UNDOCUMENTED = YES
+WARN_IF_UNDOCUMENTED = NO
WARN_IF_DOC_ERROR = YES
WARN_NO_PARAMDOC = NO
WARN_FORMAT = "$file:$line: $text"
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-67-g3970078 Fix and improve documentation in Scribo.
by Guillaume Lazzara
* doc/Doxyfile.in: Cleanup doxygen output.
* doc/groups.dox: Set groups.
* scribo/binarization/global_threshold.hh,
* scribo/binarization/internal/niblack_formula.hh,
* scribo/binarization/internal/sauvola_formula.hh,
* scribo/binarization/local_threshold.hh,
* scribo/binarization/niblack.hh,
* scribo/binarization/niblack_threshold.hh,
* scribo/binarization/otsu.hh,
* scribo/binarization/otsu_threshold.hh,
* scribo/binarization/sauvola.hh,
* scribo/binarization/sauvola_ms.hh,
* scribo/binarization/sauvola_ms_split.hh,
* scribo/binarization/sauvola_threshold.hh,
* scribo/core/erase_objects.hh,
* scribo/debug/alignment_decision_image.hh,
* scribo/debug/bboxes_enlarged_image.hh,
* scribo/debug/bboxes_image.hh,
* scribo/debug/char_space_image.hh,
* scribo/debug/decision_image.hh,
* scribo/debug/highlight_text_area.hh,
* scribo/debug/line_info_image.hh,
* scribo/debug/linked_bboxes_image.hh,
* scribo/debug/links_decision_image.hh,
* scribo/debug/links_image.hh,
* scribo/debug/logger.hh,
* scribo/debug/looks_like_a_text_line_image.hh,
* scribo/debug/mean_and_base_lines_image.hh,
* scribo/debug/save_comp_diff.hh,
* scribo/debug/save_label_image.hh,
* scribo/debug/save_table_image.hh,
* scribo/debug/text_areas_image.hh,
* scribo/debug/text_color_image.hh,
* scribo/draw/bounding_box_links.hh,
* scribo/filter/images_in_paragraph.hh,
* scribo/filter/line_links_x_height.hh,
* scribo/filter/object_groups_mean_width.hh,
* scribo/filter/object_groups_size_ratio.hh,
* scribo/filter/object_groups_small.hh,
* scribo/filter/object_groups_with_holes.hh,
* scribo/filter/object_links_aligned.hh,
* scribo/filter/object_links_bbox_h_ratio.hh,
* scribo/filter/object_links_bbox_overlap.hh,
* scribo/filter/object_links_bbox_ratio.hh,
* scribo/filter/object_links_bbox_w_ratio.hh,
* scribo/filter/object_links_bottom_aligned.hh,
* scribo/filter/object_links_center_aligned.hh,
* scribo/filter/object_links_left_aligned.hh,
* scribo/filter/object_links_non_aligned_simple.hh,
* scribo/filter/object_links_right_aligned.hh,
* scribo/filter/object_links_top_aligned.hh,
* scribo/filter/objects_h_thick.hh,
* scribo/filter/objects_h_thin.hh,
* scribo/filter/objects_large.hh,
* scribo/filter/objects_on_border.hh,
* scribo/filter/objects_size_ratio.hh,
* scribo/filter/objects_small.hh,
* scribo/filter/objects_thick.hh,
* scribo/filter/objects_thin.hh,
* scribo/filter/objects_v_thick.hh,
* scribo/filter/objects_v_thin.hh,
* scribo/filter/objects_with_holes.hh,
* scribo/filter/paragraphs_bbox_overlap.hh,
* scribo/filter/paragraphs_in_borders.hh,
* scribo/filter/paragraphs_in_image.hh,
* scribo/filter/separators_in_borders.hh,
* scribo/filter/separators_in_element.hh,
* scribo/filter/separators_in_paragraph.hh,
* scribo/filter/separators_vert_in_borders.hh,
* scribo/fun/v2b/objects_on_border_filter.hh,
* scribo/postprocessing/images_to_drop_capital.hh,
* scribo/preprocessing/denoise.hh,
* scribo/preprocessing/denoise_bg.hh,
* scribo/preprocessing/denoise_fg.hh,
* scribo/preprocessing/deskew_crop.hh,
* scribo/preprocessing/homogeneous_contrast.hh,
* scribo/preprocessing/split_bg_fg.hh,
* scribo/primitive/extract/alignments.hh,
* scribo/primitive/extract/cells.hh,
* scribo/primitive/extract/components.hh,
* scribo/primitive/extract/lines_h_single.hh,
* scribo/primitive/extract/lines_h_thick_and_single.hh,
* scribo/primitive/extract/lines_v_pattern.hh,
* scribo/primitive/extract/lines_v_single.hh,
* scribo/primitive/extract/lines_v_thick_and_single.hh,
* scribo/primitive/group/from_double_link.hh,
* scribo/primitive/identify.hh,
* scribo/primitive/link/compute.hh,
* scribo/primitive/link/with_single_left_link.hh,
* scribo/primitive/link/with_single_left_link_dmax_ratio_aligned.hh,
* scribo/primitive/link/with_single_right_link.hh,
* scribo/primitive/link/with_single_right_link_bottom.hh,
* scribo/primitive/link/with_single_right_link_dmax_ratio_aligned.hh,
* scribo/primitive/link/with_single_right_link_top.hh,
* scribo/table/align_lines_horizontaly.hh,
* scribo/table/rebuild.hh,
* scribo/text/extract_lines.hh,
* scribo/text/extract_lines_wo_merge.hh,
* scribo/text/recognition.hh,
* scribo/toolchain/content_in_doc.hh,
* scribo/toolchain/content_in_hdoc.hh,
* scribo/toolchain/internal/content_in_doc_functor.hh,
* scribo/toolchain/internal/content_in_hdoc_functor.hh,
* scribo/toolchain/internal/text_in_doc_functor.hh,
* scribo/toolchain/internal/text_in_doc_preprocess_functor.hh,
* scribo/toolchain/internal/text_in_picture_functor.hh,
* scribo/toolchain/internal/toolchain_functor.hh,
* scribo/toolchain/nepomuk/text_extraction.hh,
* scribo/toolchain/text_in_doc.hh,
* scribo/toolchain/text_in_doc_preprocess.hh,
* scribo/toolchain/text_in_picture.hh: Fix comments.
---
scribo/ChangeLog | 118 +++++++++++++++++
scribo/doc/Doxyfile.in | 10 +-
scribo/doc/groups.dox | 137 +++++++++++++++++++-
scribo/scribo/binarization/global_threshold.hh | 4 +-
.../binarization/internal/niblack_formula.hh | 5 +-
.../binarization/internal/sauvola_formula.hh | 5 +-
scribo/scribo/binarization/local_threshold.hh | 8 +-
scribo/scribo/binarization/niblack.hh | 13 +-
scribo/scribo/binarization/niblack_threshold.hh | 15 ++-
scribo/scribo/binarization/otsu.hh | 2 +
scribo/scribo/binarization/otsu_threshold.hh | 2 +
scribo/scribo/binarization/sauvola.hh | 22 ++-
scribo/scribo/binarization/sauvola_ms.hh | 12 +-
scribo/scribo/binarization/sauvola_ms_split.hh | 4 +
scribo/scribo/binarization/sauvola_threshold.hh | 13 ++-
scribo/scribo/core/erase_objects.hh | 6 +-
scribo/scribo/debug/alignment_decision_image.hh | 13 +-
scribo/scribo/debug/bboxes_enlarged_image.hh | 8 +-
scribo/scribo/debug/bboxes_image.hh | 17 ++-
scribo/scribo/debug/char_space_image.hh | 5 +-
scribo/scribo/debug/decision_image.hh | 22 ++-
scribo/scribo/debug/highlight_text_area.hh | 12 +-
scribo/scribo/debug/line_info_image.hh | 4 +
scribo/scribo/debug/linked_bboxes_image.hh | 29 +++-
scribo/scribo/debug/links_decision_image.hh | 12 +-
scribo/scribo/debug/links_image.hh | 7 +-
scribo/scribo/debug/logger.hh | 9 +-
.../scribo/debug/looks_like_a_text_line_image.hh | 12 ++-
scribo/scribo/debug/mean_and_base_lines_image.hh | 8 +-
scribo/scribo/debug/save_comp_diff.hh | 31 +++--
scribo/scribo/debug/save_label_image.hh | 9 +-
scribo/scribo/debug/save_table_image.hh | 11 ++-
scribo/scribo/debug/text_areas_image.hh | 4 +-
scribo/scribo/debug/text_color_image.hh | 4 +
scribo/scribo/draw/bounding_box_links.hh | 38 +++---
scribo/scribo/filter/images_in_paragraph.hh | 14 +--
scribo/scribo/filter/line_links_x_height.hh | 9 ++
scribo/scribo/filter/object_groups_mean_width.hh | 1 +
scribo/scribo/filter/object_groups_size_ratio.hh | 16 +++
scribo/scribo/filter/object_groups_small.hh | 9 +-
scribo/scribo/filter/object_groups_with_holes.hh | 13 ++-
scribo/scribo/filter/object_links_aligned.hh | 6 +-
scribo/scribo/filter/object_links_bbox_h_ratio.hh | 8 +-
scribo/scribo/filter/object_links_bbox_overlap.hh | 6 +-
scribo/scribo/filter/object_links_bbox_ratio.hh | 6 +-
scribo/scribo/filter/object_links_bbox_w_ratio.hh | 9 +-
.../scribo/filter/object_links_bottom_aligned.hh | 5 +-
.../scribo/filter/object_links_center_aligned.hh | 11 +-
scribo/scribo/filter/object_links_left_aligned.hh | 10 +-
.../filter/object_links_non_aligned_simple.hh | 4 +-
scribo/scribo/filter/object_links_right_aligned.hh | 5 +-
scribo/scribo/filter/object_links_top_aligned.hh | 8 +-
scribo/scribo/filter/objects_h_thick.hh | 4 +
scribo/scribo/filter/objects_h_thin.hh | 4 +
scribo/scribo/filter/objects_large.hh | 17 ++-
scribo/scribo/filter/objects_on_border.hh | 8 +-
scribo/scribo/filter/objects_size_ratio.hh | 6 +-
scribo/scribo/filter/objects_small.hh | 17 ++-
scribo/scribo/filter/objects_thick.hh | 14 ++-
scribo/scribo/filter/objects_thin.hh | 15 ++-
scribo/scribo/filter/objects_v_thick.hh | 12 ++-
scribo/scribo/filter/objects_v_thin.hh | 21 ++-
scribo/scribo/filter/objects_with_holes.hh | 26 ++++-
scribo/scribo/filter/paragraphs_bbox_overlap.hh | 11 +-
scribo/scribo/filter/paragraphs_in_borders.hh | 8 +-
scribo/scribo/filter/paragraphs_in_image.hh | 11 +-
scribo/scribo/filter/separators_in_borders.hh | 15 ++-
scribo/scribo/filter/separators_in_element.hh | 12 +-
scribo/scribo/filter/separators_in_paragraph.hh | 26 +++--
scribo/scribo/filter/separators_vert_in_borders.hh | 10 +-
scribo/scribo/fun/v2b/objects_on_border_filter.hh | 5 +-
.../postprocessing/images_to_drop_capital.hh | 10 +-
scribo/scribo/preprocessing/denoise.hh | 11 +-
scribo/scribo/preprocessing/denoise_bg.hh | 6 +-
scribo/scribo/preprocessing/denoise_fg.hh | 9 +-
scribo/scribo/preprocessing/deskew_crop.hh | 13 +-
.../scribo/preprocessing/homogeneous_contrast.hh | 4 +-
scribo/scribo/preprocessing/split_bg_fg.hh | 8 +-
scribo/scribo/primitive/extract/alignments.hh | 14 ++-
scribo/scribo/primitive/extract/cells.hh | 14 +-
scribo/scribo/primitive/extract/components.hh | 8 +-
scribo/scribo/primitive/extract/lines_h_single.hh | 22 ++--
.../primitive/extract/lines_h_thick_and_single.hh | 23 ++--
scribo/scribo/primitive/extract/lines_v_pattern.hh | 6 +-
scribo/scribo/primitive/extract/lines_v_single.hh | 22 ++--
.../primitive/extract/lines_v_thick_and_single.hh | 20 ++--
scribo/scribo/primitive/group/from_double_link.hh | 1 -
scribo/scribo/primitive/identify.hh | 2 +-
scribo/scribo/primitive/link/compute.hh | 6 +-
.../scribo/primitive/link/with_single_left_link.hh | 10 +-
.../with_single_left_link_dmax_ratio_aligned.hh | 10 +-
.../primitive/link/with_single_right_link.hh | 8 +-
.../link/with_single_right_link_bottom.hh | 10 +-
.../with_single_right_link_dmax_ratio_aligned.hh | 8 +-
.../primitive/link/with_single_right_link_top.hh | 6 +-
scribo/scribo/table/align_lines_horizontaly.hh | 9 +-
scribo/scribo/table/rebuild.hh | 20 ++--
scribo/scribo/text/extract_lines.hh | 10 +-
scribo/scribo/text/extract_lines_wo_merge.hh | 8 +-
scribo/scribo/text/recognition.hh | 15 ++-
scribo/scribo/toolchain/content_in_doc.hh | 23 +++-
scribo/scribo/toolchain/content_in_hdoc.hh | 21 +++-
.../toolchain/internal/content_in_doc_functor.hh | 6 +
.../toolchain/internal/content_in_hdoc_functor.hh | 9 ++
.../toolchain/internal/text_in_doc_functor.hh | 6 +
.../internal/text_in_doc_preprocess_functor.hh | 5 +
.../toolchain/internal/text_in_picture_functor.hh | 4 +-
.../scribo/toolchain/internal/toolchain_functor.hh | 5 +
scribo/scribo/toolchain/nepomuk/text_extraction.hh | 6 +-
scribo/scribo/toolchain/text_in_doc.hh | 22 +++-
scribo/scribo/toolchain/text_in_doc_preprocess.hh | 14 ++
scribo/scribo/toolchain/text_in_picture.hh | 21 +++-
112 files changed, 1087 insertions(+), 381 deletions(-)
diff --git a/scribo/ChangeLog b/scribo/ChangeLog
index 1470f32..75ff89d 100644
--- a/scribo/ChangeLog
+++ b/scribo/ChangeLog
@@ -1,5 +1,123 @@
2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
+ Fix and improve documentation in Scribo.
+
+ * doc/Doxyfile.in: Cleanup doxygen output.
+
+ * doc/groups.dox: Set groups.
+
+ * scribo/binarization/global_threshold.hh,
+ * scribo/binarization/internal/niblack_formula.hh,
+ * scribo/binarization/internal/sauvola_formula.hh,
+ * scribo/binarization/local_threshold.hh,
+ * scribo/binarization/niblack.hh,
+ * scribo/binarization/niblack_threshold.hh,
+ * scribo/binarization/otsu.hh,
+ * scribo/binarization/otsu_threshold.hh,
+ * scribo/binarization/sauvola.hh,
+ * scribo/binarization/sauvola_ms.hh,
+ * scribo/binarization/sauvola_ms_split.hh,
+ * scribo/binarization/sauvola_threshold.hh,
+ * scribo/core/erase_objects.hh,
+ * scribo/debug/alignment_decision_image.hh,
+ * scribo/debug/bboxes_enlarged_image.hh,
+ * scribo/debug/bboxes_image.hh,
+ * scribo/debug/char_space_image.hh,
+ * scribo/debug/decision_image.hh,
+ * scribo/debug/highlight_text_area.hh,
+ * scribo/debug/line_info_image.hh,
+ * scribo/debug/linked_bboxes_image.hh,
+ * scribo/debug/links_decision_image.hh,
+ * scribo/debug/links_image.hh,
+ * scribo/debug/logger.hh,
+ * scribo/debug/looks_like_a_text_line_image.hh,
+ * scribo/debug/mean_and_base_lines_image.hh,
+ * scribo/debug/save_comp_diff.hh,
+ * scribo/debug/save_label_image.hh,
+ * scribo/debug/save_table_image.hh,
+ * scribo/debug/text_areas_image.hh,
+ * scribo/debug/text_color_image.hh,
+ * scribo/draw/bounding_box_links.hh,
+ * scribo/filter/images_in_paragraph.hh,
+ * scribo/filter/line_links_x_height.hh,
+ * scribo/filter/object_groups_mean_width.hh,
+ * scribo/filter/object_groups_size_ratio.hh,
+ * scribo/filter/object_groups_small.hh,
+ * scribo/filter/object_groups_with_holes.hh,
+ * scribo/filter/object_links_aligned.hh,
+ * scribo/filter/object_links_bbox_h_ratio.hh,
+ * scribo/filter/object_links_bbox_overlap.hh,
+ * scribo/filter/object_links_bbox_ratio.hh,
+ * scribo/filter/object_links_bbox_w_ratio.hh,
+ * scribo/filter/object_links_bottom_aligned.hh,
+ * scribo/filter/object_links_center_aligned.hh,
+ * scribo/filter/object_links_left_aligned.hh,
+ * scribo/filter/object_links_non_aligned_simple.hh,
+ * scribo/filter/object_links_right_aligned.hh,
+ * scribo/filter/object_links_top_aligned.hh,
+ * scribo/filter/objects_h_thick.hh,
+ * scribo/filter/objects_h_thin.hh,
+ * scribo/filter/objects_large.hh,
+ * scribo/filter/objects_on_border.hh,
+ * scribo/filter/objects_size_ratio.hh,
+ * scribo/filter/objects_small.hh,
+ * scribo/filter/objects_thick.hh,
+ * scribo/filter/objects_thin.hh,
+ * scribo/filter/objects_v_thick.hh,
+ * scribo/filter/objects_v_thin.hh,
+ * scribo/filter/objects_with_holes.hh,
+ * scribo/filter/paragraphs_bbox_overlap.hh,
+ * scribo/filter/paragraphs_in_borders.hh,
+ * scribo/filter/paragraphs_in_image.hh,
+ * scribo/filter/separators_in_borders.hh,
+ * scribo/filter/separators_in_element.hh,
+ * scribo/filter/separators_in_paragraph.hh,
+ * scribo/filter/separators_vert_in_borders.hh,
+ * scribo/fun/v2b/objects_on_border_filter.hh,
+ * scribo/postprocessing/images_to_drop_capital.hh,
+ * scribo/preprocessing/denoise.hh,
+ * scribo/preprocessing/denoise_bg.hh,
+ * scribo/preprocessing/denoise_fg.hh,
+ * scribo/preprocessing/deskew_crop.hh,
+ * scribo/preprocessing/homogeneous_contrast.hh,
+ * scribo/preprocessing/split_bg_fg.hh,
+ * scribo/primitive/extract/alignments.hh,
+ * scribo/primitive/extract/cells.hh,
+ * scribo/primitive/extract/components.hh,
+ * scribo/primitive/extract/lines_h_single.hh,
+ * scribo/primitive/extract/lines_h_thick_and_single.hh,
+ * scribo/primitive/extract/lines_v_pattern.hh,
+ * scribo/primitive/extract/lines_v_single.hh,
+ * scribo/primitive/extract/lines_v_thick_and_single.hh,
+ * scribo/primitive/group/from_double_link.hh,
+ * scribo/primitive/identify.hh,
+ * scribo/primitive/link/compute.hh,
+ * scribo/primitive/link/with_single_left_link.hh,
+ * scribo/primitive/link/with_single_left_link_dmax_ratio_aligned.hh,
+ * scribo/primitive/link/with_single_right_link.hh,
+ * scribo/primitive/link/with_single_right_link_bottom.hh,
+ * scribo/primitive/link/with_single_right_link_dmax_ratio_aligned.hh,
+ * scribo/primitive/link/with_single_right_link_top.hh,
+ * scribo/table/align_lines_horizontaly.hh,
+ * scribo/table/rebuild.hh,
+ * scribo/text/extract_lines.hh,
+ * scribo/text/extract_lines_wo_merge.hh,
+ * scribo/text/recognition.hh,
+ * scribo/toolchain/content_in_doc.hh,
+ * scribo/toolchain/content_in_hdoc.hh,
+ * scribo/toolchain/internal/content_in_doc_functor.hh,
+ * scribo/toolchain/internal/content_in_hdoc_functor.hh,
+ * scribo/toolchain/internal/text_in_doc_functor.hh,
+ * scribo/toolchain/internal/text_in_doc_preprocess_functor.hh,
+ * scribo/toolchain/internal/text_in_picture_functor.hh,
+ * scribo/toolchain/internal/toolchain_functor.hh,
+ * scribo/toolchain/nepomuk/text_extraction.hh,
+ * scribo/toolchain/text_in_doc.hh,
+ * scribo/toolchain/text_in_doc_preprocess.hh,
+ * scribo/toolchain/text_in_picture.hh: Fix comments.
+
+2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
+
Fix scribo doc (un)installation.
* Makefile.am: Add doc in SUBDIRS.
diff --git a/scribo/doc/Doxyfile.in b/scribo/doc/Doxyfile.in
index 7ad639f..a6d8c3e 100644
--- a/scribo/doc/Doxyfile.in
+++ b/scribo/doc/Doxyfile.in
@@ -49,6 +49,7 @@ DISTRIBUTE_GROUP_DOC = NO
SUBGROUPING = YES
INLINE_GROUPED_CLASSES = NO
SORT_GROUP_NAMES = YES
+TYPEDEF_HIDES_STRUCT = YES
#---------------------------------------------------------------------------
# Build related configuration options
#---------------------------------------------------------------------------
@@ -85,7 +86,7 @@ FILE_VERSION_FILTER =
#---------------------------------------------------------------------------
QUIET = YES
WARNINGS = YES
-WARN_IF_UNDOCUMENTED = YES
+WARN_IF_UNDOCUMENTED = NO
WARN_IF_DOC_ERROR = YES
WARN_NO_PARAMDOC = NO
WARN_FORMAT = "$file:$line: $text"
@@ -107,7 +108,7 @@ EXCLUDE = @top_srcdir@/scribo/demo \
@top_srcdir@/scribo/tests/unit_test
EXCLUDE_SYMLINKS = YES
EXCLUDE_PATTERNS = *spe.hh
-EXCLUDE_SYMBOLS = *::internal* *_ mln::trait::*
+EXCLUDE_SYMBOLS = *_ mln::trait::*
EXAMPLE_PATH = @top_srcdir@/scribo/doc/examples \
@top_srcdir@/scribo/doc/outputs \
@top_srcdir@/scribo/doc/examples/split \
@@ -260,7 +261,6 @@ PREDEFINED = "for_all(x)=for(x.start(); x.is_valid(); x.next())" \
"mln_trait_op_plus(L, R)=typename mln::trait::op::plus< L , R >::ret" \
"mln_trait_op_times(L, R)=typename mln::trait::op::times< L , R >::ret" \
"mln_trait_op_uminus(T)=typename mln::trait::op::uminus< T >::ret" \
- "mln_ch_value(I, V)=typename mln::trait::ch_value< I, V >::ret" \
"mlc_unqualif_(T)=mln::metal::unqualif<T>::ret" \
"mln_piter_(T)=T::piter" \
"mln_fwd_piter_(T)=T::fwd_piter" \
@@ -295,8 +295,8 @@ PREDEFINED = "for_all(x)=for(x.start(); x.is_valid(); x.next())" \
"mln_trait_op_div_(L, R)=mln::trait::op::div< L , R >::ret" \
"mln_trait_op_mod_(L, R)=mln::trait::op::mod< L , R >::ret" \
"mln_trait_op_uminus_(T)=mln::trait::op::uminus< T >::ret" \
- "mln_ch_value_(I, V)=typename mln::trait::ch_value< I, V >::ret" \
- "mln_ch_value(I, V)_=mln::trait::ch_value< I, V >::ret" \
+ "mln_ch_value_(I, V)=mln::trait::ch_value< I, V >::ret" \
+ "mln_ch_value(I, V)=typename mln::trait::ch_value< I, V >::ret" \
"mln_morpher_lvalue(I)=typename internal::morpher_lvalue_<I>::ret" \
"mln_concrete(I)=typename mln::trait::concrete< I >::ret" \
"mln_concrete_ch_value(I, V)=typename mln::trait::ch_value< typename mln::trait::concrete< I >::ret, V >::ret" \
diff --git a/scribo/doc/groups.dox b/scribo/doc/groups.dox
index 73d842e..f3db20b 100644
--- a/scribo/doc/groups.dox
+++ b/scribo/doc/groups.dox
@@ -4,7 +4,6 @@
*
*/
-
/*! \defgroup grproutine Routines
*
* \brief All routines/algorithms provided in Scribo.
@@ -12,8 +11,144 @@
*/
+
/*! \defgroup grpalgobin Binarization
*
* \brief Binarization algorithms.
*
+ * \ingroup grproutine
+ */
+
+/*! \defgroup grpalgobinsauvola Sauvola
+ *
+ * \brief Sauvola's Binarization implementations.
+ *
+ * \ingroup grpalgobin
+ */
+
+/*! \defgroup grpalgobinniblack Niblack
+ *
+ * \brief Niblack's Binarization implementations.
+ *
+ * \ingroup grpalgobin
+ */
+
+/*! \defgroup grpalgobinotsu Otsu
+ *
+ * \brief Otsu's Binarization implementations.
+ *
+ * \ingroup grpalgobin
+ */
+
+
+/*! \defgroup grpalgodebug Debug
+ *
+ * \brief Debug routines.
+ *
+ * \ingroup grproutine
+ */
+
+
+
+/*! \defgroup grpalgofilter Element Filtering
+ *
+ * \brief Element filtering routines.
+ *
+ * \ingroup grproutine
+ */
+
+/*! \defgroup grpalgofiltercomp Components
+ *
+ * \brief Component filtering routines.
+ *
+ * \ingroup grpalgofilter
+ */
+
+/*! \defgroup grpalgofiltercomplink Component Links
+ *
+ * \brief Component links filtering routines.
+ *
+ * \ingroup grpalgofilter
+ */
+
+/*! \defgroup grpalgofiltercompgroup Component Groups
+ *
+ * \brief Component groups filtering routines.
+ *
+ * \ingroup grpalgofilter
+ */
+
+/*! \defgroup grpalgofilterlinelink Line Links
+ *
+ * \brief Line links filtering routines.
+ *
+ * \ingroup grpalgofilter
+ */
+
+/*! \defgroup grpalgofilterpar Paragraphs
+ *
+ * \brief Paragraphs filtering routines.
+ *
+ * \ingroup grpalgofilter
+ */
+
+/*! \defgroup grpalgofilterelt Elements
+ *
+ * \brief Elements filtering routines.
+ *
+ * \ingroup grpalgofilter
+ */
+
+
+
+
+/*! \defgroup grptoolchain Toolchains
+ *
+ * \brief Full toolchains performing content analysis and extraction.
+ *
+ * \ingroup grproutine
+ */
+
+/*! \defgroup grptoolchainpic Pictures
+ *
+ * \brief Full toolchains performing content analysis and extraction
+ * in pictures.
+ *
+ * \ingroup grptoolchain
+ */
+
+/*! \defgroup grptoolchaindoc Documents
+ *
+ * \brief Full toolchains performing content analysis and extraction
+ * in document images.
+ *
+ * \ingroup grptoolchain
+ */
+
+/*! \defgroup grptoolchainpicpreproc Preprocessing
+ *
+ * \brief Classical preprocessing toolchains for pictures.
+ *
+ * \ingroup grptoolchainpic
+ */
+
+/*! \defgroup grptoolchaindocpreproc Preprocessing
+ *
+ * \brief Classical preprocessing toolchains for document image.
+ *
+ * \ingroup grptoolchaindoc
+ */
+
+/*! \defgroup grptoolchainpicproc Processing
+ *
+ * \brief Classical processing toolchains for pictures.
+ *
+ * \ingroup grptoolchainpic
+ */
+
+/*! \defgroup grptoolchaindocproc Processing
+ *
+ * \brief Classical processing toolchains for document image.
+ *
+ * \ingroup grptoolchaindoc
*/
diff --git a/scribo/scribo/binarization/global_threshold.hh b/scribo/scribo/binarization/global_threshold.hh
index aefbe00..46c25ef 100644
--- a/scribo/scribo/binarization/global_threshold.hh
+++ b/scribo/scribo/binarization/global_threshold.hh
@@ -55,8 +55,8 @@ namespace scribo
///
/// output(p) = input(p) >= threshold
///
- /// \input[in] input A grayscale image.
- /// \input[in] threshold A value.
+ /// \param[in] input A grayscale image.
+ /// \param[in] threshold A value.
///
/// \return A boolean image.
//
diff --git a/scribo/scribo/binarization/internal/niblack_formula.hh b/scribo/scribo/binarization/internal/niblack_formula.hh
index 54dbc9b..81432ae 100644
--- a/scribo/scribo/binarization/internal/niblack_formula.hh
+++ b/scribo/scribo/binarization/internal/niblack_formula.hh
@@ -56,7 +56,7 @@ namespace scribo
\param[in] m_x_y Mean value.
\param[in] s_x_y Standard deviation.
- \param[in] k Control the threshold value in the local
+ \param[in] K Control the threshold value in the local
window. The higher, the lower the threshold
form the local mean m(x, y).
\param[in] R Maximum value of the standard deviation (128
@@ -68,7 +68,8 @@ namespace scribo
const double K, const double R) const;
/*!
- \overload K = 0.34.
+ \overload
+ K is set to 0.34.
*/
double operator()(const double m_x_y, const double s_x_y) const;
diff --git a/scribo/scribo/binarization/internal/sauvola_formula.hh b/scribo/scribo/binarization/internal/sauvola_formula.hh
index adbef13..6eb49a0 100644
--- a/scribo/scribo/binarization/internal/sauvola_formula.hh
+++ b/scribo/scribo/binarization/internal/sauvola_formula.hh
@@ -62,7 +62,7 @@ namespace scribo
\param[in] m_x_y Mean value.
\param[in] s_x_y Standard deviation.
- \param[in] k Control the threshold value in the local
+ \param[in] K Controls the threshold value in the local
window. The higher, the lower the threshold
form the local mean m(x, y).
\param[in] R Maximum value of the standard deviation (128
@@ -74,7 +74,8 @@ namespace scribo
const double K, const double R) const;
/*!
- \overload K = 0.34 and R = 128.
+ \overload
+ K is set to 0.34 and R to 128.
*/
double operator()(const double m_x_y, const double s_x_y) const;
diff --git a/scribo/scribo/binarization/local_threshold.hh b/scribo/scribo/binarization/local_threshold.hh
index 9d116ed..8ac3276 100644
--- a/scribo/scribo/binarization/local_threshold.hh
+++ b/scribo/scribo/binarization/local_threshold.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -53,8 +53,8 @@ namespace scribo
///
/// \pre input.domain() == threshold.domain()
///
- /// \input[in] input A grayscale image.
- /// \input[in] threshold A grayscale image.
+ /// \param[in] input A grayscale image.
+ /// \param[in] threshold A grayscale image.
///
/// \return A boolean image.
//
diff --git a/scribo/scribo/binarization/niblack.hh b/scribo/scribo/binarization/niblack.hh
index e66e7b4..cc78808 100644
--- a/scribo/scribo/binarization/niblack.hh
+++ b/scribo/scribo/binarization/niblack.hh
@@ -55,12 +55,13 @@ namespace scribo
/*! \brief Convert an image into a binary image.
- \input[in] input An image.
- \input[in] window_size The window size.
- \input[in] K Niblack's formulae constant.
+ \param[in] input An image.
+ \param[in] window_size The window size.
+ \param[in] K Niblack's formulae constant.
\return A binary image.
+ \ingroup grpalgobinniblack
*/
template <typename I>
mln_ch_value(I, bool)
@@ -72,11 +73,12 @@ namespace scribo
Niblack's formulae constant K is set to 0.34.
- \input[in] input An image.
- \input[in] window_size The window size.
+ \param[in] input An image.
+ \param[in] window_size The window size.
\return A binary image.
+ \ingroup grpalgobinniblack
*/
template <typename I>
mln_ch_value(I, bool)
@@ -85,6 +87,7 @@ namespace scribo
/// \overload
/// The window size is set to 11.
+ /// \ingroup grpalgobinniblack
//
template <typename I>
mln_ch_value(I, bool)
diff --git a/scribo/scribo/binarization/niblack_threshold.hh b/scribo/scribo/binarization/niblack_threshold.hh
index db4a74e..d9829f4 100644
--- a/scribo/scribo/binarization/niblack_threshold.hh
+++ b/scribo/scribo/binarization/niblack_threshold.hh
@@ -56,14 +56,16 @@ namespace scribo
/*! \brief Compute an image of local threshold using Niblack algorithm.
- \input[in] input A gray level image.
- \input[in] window_size The window size.
- \input[out] simple The sum of all intensities of \p input.
- \input[out] squared The sum of all squared intensities of \p
+ \param[in] input A gray level image.
+ \param[in] window_size The window size.
+ \param[in] K Controls the threshold value in the local window.
+ \param[out] simple The sum of all intensities of \p input.
+ \param[out] squared The sum of all squared intensities of \p
input.
\return An image of local thresholds.
+ \ingroup grpalgobinniblack
*/
template <typename I, typename J>
mln_ch_value(I, value::int_u8)
@@ -73,6 +75,8 @@ namespace scribo
Image<J>& squared);
/// \overload
+ /// \ingroup grpalgobinniblack
+ //
template <typename I>
mln_ch_value(I, value::int_u8)
niblack_threshold(const Image<I>& input, unsigned window_size,
@@ -80,6 +84,8 @@ namespace scribo
/// \overload
/// K is set to 0.34
+ /// \ingroup grpalgobinniblack
+ //
template <typename I>
mln_ch_value(I, value::int_u8)
niblack_threshold(const Image<I>& input, unsigned window_size);
@@ -87,6 +93,7 @@ namespace scribo
/// \overload
/// The window size is set to 11.
+ /// \ingroup grpalgobinniblack
//
template <typename I>
mln_ch_value(I, value::int_u8)
diff --git a/scribo/scribo/binarization/otsu.hh b/scribo/scribo/binarization/otsu.hh
index 6b9d492..2b8e97a 100644
--- a/scribo/scribo/binarization/otsu.hh
+++ b/scribo/scribo/binarization/otsu.hh
@@ -48,6 +48,8 @@ namespace scribo
\return A binary image. True for foreground, False for
background.
+
+ \ingroup grpalgobinotsu
*/
template <typename I>
mln_ch_value(I,bool)
diff --git a/scribo/scribo/binarization/otsu_threshold.hh b/scribo/scribo/binarization/otsu_threshold.hh
index ab663d0..7ba458f 100644
--- a/scribo/scribo/binarization/otsu_threshold.hh
+++ b/scribo/scribo/binarization/otsu_threshold.hh
@@ -49,6 +49,8 @@ namespace scribo
\param[in] input A gray-scale image.
\return A global threshold value.
+
+ \ingroup grpalgobinotsu
*/
template <typename I>
mln_value(I)
diff --git a/scribo/scribo/binarization/sauvola.hh b/scribo/scribo/binarization/sauvola.hh
index fc3e104..35757d6 100644
--- a/scribo/scribo/binarization/sauvola.hh
+++ b/scribo/scribo/binarization/sauvola.hh
@@ -56,12 +56,13 @@ namespace scribo
/*! \brief Convert an image into a binary image.
- \input[in] input An image.
- \input[in] window_size The window size.
- \input[in] K Sauvola's formulae constant.
+ \param[in] input An image.
+ \param[in] window_size The window size.
+ \param[in] K Sauvola's formulae constant.
\return A binary image.
+ \ingroup grpalgobinsauvola
*/
template <typename I>
mln_ch_value(I, bool)
@@ -71,22 +72,27 @@ namespace scribo
/*! \brief Convert an image into a binary image.
+ \overload
+
Sauvola's formulae constant K is set to 0.34.
- \input[in] input An image.
- \input[in] window_size The window size.
+ \param[in] input An image.
+ \param[in] window_size The window size.
\return A binary image.
+ \ingroup grpalgobinsauvola
*/
template <typename I>
mln_ch_value(I, bool)
sauvola(const Image<I>& input, unsigned window_size);
- /// \overload
- /// The window size is set to 11.
- //
+ /*! \overload
+ The window size is set to 11.
+
+ \ingroup grpalgobinsauvola
+ */
template <typename I>
mln_ch_value(I, bool)
sauvola(const Image<I>& input);
diff --git a/scribo/scribo/binarization/sauvola_ms.hh b/scribo/scribo/binarization/sauvola_ms.hh
index 36629f9..c06cea8 100644
--- a/scribo/scribo/binarization/sauvola_ms.hh
+++ b/scribo/scribo/binarization/sauvola_ms.hh
@@ -88,21 +88,23 @@ namespace scribo
\param[in] input_1 A grayscale or a color image.
\param[in] w_1 The window size used to compute stats.
\param[in] s The scale factor used for the first subscaling.
- \param[in] lambda_min_1 Size of the objects kept at scale 1.
\param[in] K Sauvola's formulae parameter.
-
- \p w_1 and \p lambda_min_1 are expressed according to the image
- at scale 0, i.e. the original size.
+ \p w_1 is expressed according to the image at scale 0, i.e. the
+ original size.
\return A Boolean image.
+
+ \ingroup grpalgobinsauvola
*/
template <typename I>
mln_ch_value(I,bool)
- sauvola_ms(const Image<I>& input_1_, unsigned w_1, unsigned s, double K);
+ sauvola_ms(const Image<I>& input_1, unsigned w_1, unsigned s, double K);
/// \overload
/// K is set to 0.34.
+ ///
+ /// \ingroup grpalgobinsauvola
//
template <typename I>
mln_ch_value(I,bool)
diff --git a/scribo/scribo/binarization/sauvola_ms_split.hh b/scribo/scribo/binarization/sauvola_ms_split.hh
index cbef524..b7a44ac 100644
--- a/scribo/scribo/binarization/sauvola_ms_split.hh
+++ b/scribo/scribo/binarization/sauvola_ms_split.hh
@@ -65,6 +65,8 @@ namespace scribo
original size.
\return A Boolean image.
+
+ \ingroup grpalgobinsauvola
*/
template <typename I>
mln_ch_value(I, bool)
@@ -75,6 +77,8 @@ namespace scribo
/// \overload
/// K is set to 0.34.
+ /// \ingroup grpalgobinsauvola
+ //
template <typename I>
mln_ch_value(I, bool)
sauvola_ms_split(const Image<I>& input_1_, unsigned w_1,
diff --git a/scribo/scribo/binarization/sauvola_threshold.hh b/scribo/scribo/binarization/sauvola_threshold.hh
index df46e95..41d70f7 100644
--- a/scribo/scribo/binarization/sauvola_threshold.hh
+++ b/scribo/scribo/binarization/sauvola_threshold.hh
@@ -56,14 +56,16 @@ namespace scribo
/*! \brief Compute an image of local threshold using Sauvola algorithm.
- \input[in] input A gray level image.
- \input[in] window_size The window size.
- \input[out] simple The sum of all intensities of \p input.
- \input[out] squared The sum of all squared intensities of \p
+ \param[in] input A gray level image.
+ \param[in] window_size The window size.
+ \param[in] K Controls the threshold value in the local window.
+ \param[out] simple The sum of all intensities of \p input.
+ \param[out] squared The sum of all squared intensities of \p
input.
\return An image of local thresholds.
+ \ingroup grpalgobinsauvola
*/
template <typename I, typename J>
mln_ch_value(I, value::int_u8)
@@ -73,6 +75,7 @@ namespace scribo
Image<J>& squared);
/// \overload
+ /// \ingroup grpalgobinsauvola
template <typename I>
mln_ch_value(I, value::int_u8)
sauvola_threshold(const Image<I>& input, unsigned window_size,
@@ -80,6 +83,7 @@ namespace scribo
/// \overload
/// K is set to 0.34
+ /// \ingroup grpalgobinsauvola
template <typename I>
mln_ch_value(I, value::int_u8)
sauvola_threshold(const Image<I>& input, unsigned window_size);
@@ -87,6 +91,7 @@ namespace scribo
/// \overload
/// The window size is set to 11.
+ /// \ingroup grpalgobinsauvola
//
template <typename I>
mln_ch_value(I, value::int_u8)
diff --git a/scribo/scribo/core/erase_objects.hh b/scribo/scribo/core/erase_objects.hh
index 931024e..78fd293 100644
--- a/scribo/scribo/core/erase_objects.hh
+++ b/scribo/scribo/core/erase_objects.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -47,7 +47,7 @@ namespace scribo
/// Remove labeled components from a binary image.
///
- /// \param[in,out] input_ A binary image.
+ /// \param[in,out] input A binary image.
/// \param[in] comps A set of components to be erased.
///
//
diff --git a/scribo/scribo/debug/alignment_decision_image.hh b/scribo/scribo/debug/alignment_decision_image.hh
index 450091d..cc7937b 100644
--- a/scribo/scribo/debug/alignment_decision_image.hh
+++ b/scribo/scribo/debug/alignment_decision_image.hh
@@ -55,18 +55,21 @@ namespace scribo
/*! \brief Save a color image showing the difference between to
object links.
- \input[in] input An image. It's value type must be convertible
+ \param[in] input An image. It's value type must be convertible
towards rgb8.
- \input[in] links Object links information.
- \input[in] filtered_links A copy of \p links which have been
+ \param[in] links Object links information.
+ \param[in] filtered_links A copy of \p links which have been
filtered.
+ \param[in] anchor Anchor from where the links are drawn.
\return A color image. Non filtered links are drawn in
green. Others are drawn in red.
- */
+
+ \ingroup grpalgodebug
+ */
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
- alignment_decision_image(const Image<I>& input_,
+ alignment_decision_image(const Image<I>& input,
const object_links<L>& links,
const object_links<L>& filtered_links,
const anchor::Type& anchor);
diff --git a/scribo/scribo/debug/bboxes_enlarged_image.hh b/scribo/scribo/debug/bboxes_enlarged_image.hh
index 24fea2f..01afd2c 100644
--- a/scribo/scribo/debug/bboxes_enlarged_image.hh
+++ b/scribo/scribo/debug/bboxes_enlarged_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -70,6 +71,8 @@ namespace scribo
/// a text line.
///
/// \return A color image.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
@@ -81,6 +84,9 @@ namespace scribo
/// \overload
/// text_value is set to literal::green.
/// non_text_value is set to literal::red.
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
bboxes_enlarged_image(const Image<I>& input,
diff --git a/scribo/scribo/debug/bboxes_image.hh b/scribo/scribo/debug/bboxes_image.hh
index d3dc64b..1b2ee4f 100644
--- a/scribo/scribo/debug/bboxes_image.hh
+++ b/scribo/scribo/debug/bboxes_image.hh
@@ -51,14 +51,20 @@ namespace scribo
using namespace mln;
- /// Draw a list of bounding boxes
+ /// \brief Draw a list of bounding boxes
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I>
mln_ch_value(I, value::rgb8)
bboxes_image(const Image<I>& input,
const mln::util::array< box<mln_site(I)> >& bboxes,
const value::rgb8& value);
-
+ /// \overload
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I, typename L>
mln_ch_value(I, value::rgb8)
bboxes_image(const Image<I>& input,
@@ -67,6 +73,8 @@ namespace scribo
/// \overload
/// value is set to literal::red.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
inline
@@ -74,7 +82,10 @@ namespace scribo
bboxes_image(const Image<I>& input,
const line_set<L>& lines);
- /// \overload.
+ /// \overload
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I, typename L>
inline
mln_ch_value(I, value::rgb8)
diff --git a/scribo/scribo/debug/char_space_image.hh b/scribo/scribo/debug/char_space_image.hh
index 53fc303..ffb2e82 100644
--- a/scribo/scribo/debug/char_space_image.hh
+++ b/scribo/scribo/debug/char_space_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -53,6 +54,8 @@ namespace scribo
/// \brief Draw inter character space.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
diff --git a/scribo/scribo/debug/decision_image.hh b/scribo/scribo/debug/decision_image.hh
index ff298e6..9857c66 100644
--- a/scribo/scribo/debug/decision_image.hh
+++ b/scribo/scribo/debug/decision_image.hh
@@ -56,19 +56,22 @@ namespace scribo
/*! \brief Save a color image showing the difference between to
object groups.
- \input[in] input An image. It's value type must be convertible
+ \param[in] input An image. It's value type must be convertible
towards rgb8.
- \input[in] groups Object groups information.
- \input[in] filtered_groups A copy of \p groups which have been
+ \param[in] groups Object groups information.
+ \param[in] filtered_groups A copy of \p groups which have been
filtered.
+ \param[in] anchor Anchor from where the links are drawn.
\return A color image. Components part of a validated group are
drawn in green with their bounding box. Otherwise,
they are drawn in red.
+
+ \ingroup grpalgodebug
*/
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
- decision_image(const Image<I>& input_,
+ decision_image(const Image<I>& input,
const object_groups<L>& groups,
const object_groups<L>& filtered_groups,
anchor::Type anchor);
@@ -76,18 +79,21 @@ namespace scribo
/*! \brief Save a color image showing the difference between to
object links.
- \input[in] input An image. It's value type must be convertible
+ \param[in] input An image. It's value type must be convertible
towards rgb8.
- \input[in] links Object links information.
- \input[in] filtered_links A copy of \p links which have been
+ \param[in] links Object links information.
+ \param[in] filtered_links A copy of \p links which have been
filtered.
+ \param[in] anchor Anchor from where the links are drawn.
\return A color image. Non filtered links are drawn in
green. Others are drawn in red.
+
+ \ingroup grpalgodebug
*/
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
- decision_image(const Image<I>& input_,
+ decision_image(const Image<I>& input,
const object_links<L>& links,
const object_links<L>& filtered_links,
anchor::Type anchor);
diff --git a/scribo/scribo/debug/highlight_text_area.hh b/scribo/scribo/debug/highlight_text_area.hh
index 5fc9329..2a7ed8a 100644
--- a/scribo/scribo/debug/highlight_text_area.hh
+++ b/scribo/scribo/debug/highlight_text_area.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -62,10 +63,11 @@ namespace scribo
\return a color image with highlighted text areas.
+ \ingroup grpalgodebug
*/
template <typename I>
mln_ch_value(I, value::rgb8)
- highlight_text_area(const Image<I>& input_,
+ highlight_text_area(const Image<I>& input,
const mln::util::array<box<mln_site(I)> >& bbox);
@@ -76,10 +78,11 @@ namespace scribo
\return a color image with highlighted text areas.
+ \ingroup grpalgodebug
*/
template <typename I, typename L>
mln_ch_value(I, value::rgb8)
- highlight_text_area(const Image<I>& input_,
+ highlight_text_area(const Image<I>& input,
const line_set<L>& lines);
/*! \brief Darken an image and highlight valid lines.
@@ -92,10 +95,11 @@ namespace scribo
\return a color image with highlighted text areas.
+ \ingroup grpalgodebug
*/
template <typename I, typename L>
mln_ch_value(I, value::rgb8)
- highlight_text_area(const Image<I>& input_,
+ highlight_text_area(const Image<I>& input,
const scribo::component_set<L>& components);
diff --git a/scribo/scribo/debug/line_info_image.hh b/scribo/scribo/debug/line_info_image.hh
index f3d7ddf..ff277ef 100644
--- a/scribo/scribo/debug/line_info_image.hh
+++ b/scribo/scribo/debug/line_info_image.hh
@@ -53,6 +53,10 @@ namespace scribo
using namespace mln;
+ /*! \brief Draw typographic information from lines.
+
+ \ingroup grpalgodebug
+ */
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
line_info_image(const Image<I>& input, const line_set<L>& line);
diff --git a/scribo/scribo/debug/linked_bboxes_image.hh b/scribo/scribo/debug/linked_bboxes_image.hh
index 2373405..1b502b0 100644
--- a/scribo/scribo/debug/linked_bboxes_image.hh
+++ b/scribo/scribo/debug/linked_bboxes_image.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Compute the bounding box links image.
+/// \brief Compute the bounding box links image.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/graph.hh>
@@ -59,10 +59,12 @@ namespace scribo
/// Compute the line of components links image.
///
/// \param[in,out] input The binary from where the components are extracted.
- /// \param[in] link_array Lines of components links.
+ /// \param[in] array Components links.
/// \param[in] box_value Value used to draw line bounding boxes.
/// \param[in] link_value Value used to draw line links.
/// \param[in] anchor Anchor from where the links are drawn.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
@@ -74,6 +76,8 @@ namespace scribo
/// \overload
/// The default anchor type is set to anchor::Center.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
@@ -85,10 +89,13 @@ namespace scribo
/// Compute the line of components left and right links image.
///
/// \param[in,out] input The binary from where the components are extracted.
- /// \param[in] left_link Lines of components left links.
- /// \param[in] right_link Lines of components right links.
+ /// \param[in] left_link Components left links.
+ /// \param[in] right_link Components right links.
/// \param[in] box_value Value used to draw line bounding boxes.
/// \param[in] link_value Value used to draw line links.
+ /// \param[in] anchor Anchor from where the links are drawn.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
@@ -103,12 +110,15 @@ namespace scribo
/// Draw also validated links.
///
/// \param[in,out] input The binary from where the components are extracted.
- /// \param[in] left_link Lines of components left links.
- /// \param[in] right_link Lines of components right links.
+ /// \param[in] left_link Components left links.
+ /// \param[in] right_link Components right links.
/// \param[in] box_value Value used to draw line bounding boxes.
/// \param[in] left_link_value Value used to draw line left links.
/// \param[in] right_link_value Value used to draw line left links.
/// \param[in] validated_link_value Value used to draw line validated links.
+ /// \param[in] anchor Anchor from where the links are drawn.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
inline
@@ -129,6 +139,9 @@ namespace scribo
/// \param[in] g The link graph.
/// \param[in] box_value Value used to draw line bounding boxes.
/// \param[in] link_value Value used to draw line links.
+ /// \param[in] anchor Anchor from where the links are drawn.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L, typename G>
mln_ch_value(I,value::rgb8)
diff --git a/scribo/scribo/debug/links_decision_image.hh b/scribo/scribo/debug/links_decision_image.hh
index 387cd7f..f62eca4 100644
--- a/scribo/scribo/debug/links_decision_image.hh
+++ b/scribo/scribo/debug/links_decision_image.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Save a color image showing the difference between to object groups.
+/// \brief Save a color image showing the difference between to object groups.
# include <mln/core/concept/image.hh>
# include <mln/data/convert.hh>
@@ -49,7 +49,11 @@ namespace scribo
using namespace mln;
- /// FIXME: DOC!
+ /*! \brief Save a color image showing the difference between to
+ object groups.
+
+ \ingroup grpalgodebug
+ */
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
links_decision_image(const Image<I>& input_,
diff --git a/scribo/scribo/debug/links_image.hh b/scribo/scribo/debug/links_image.hh
index 47394be..ec747f6 100644
--- a/scribo/scribo/debug/links_image.hh
+++ b/scribo/scribo/debug/links_image.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Save a color image showing components links.
+/// \brief Save a color image showing components links.
# include <mln/core/concept/image.hh>
# include <mln/accu/center.hh>
@@ -50,7 +50,10 @@ namespace scribo
using namespace mln;
- /// FIXME: DOC!
+ /*! \brief Save a color image showing components links.
+
+ \ingroup grpalgodebug
+ */
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
links_image(const Image<I>& input_,
diff --git a/scribo/scribo/debug/logger.hh b/scribo/scribo/debug/logger.hh
index f16dd89..ef5fd74 100644
--- a/scribo/scribo/debug/logger.hh
+++ b/scribo/scribo/debug/logger.hh
@@ -27,6 +27,10 @@
#ifndef SCRIBO_CORE_LOGGER_HH
# define SCRIBO_CORE_LOGGER_HH
+/// \file
+///
+/// \brief Logger class used to save debug images.
+
# include <mln/core/concept/image.hh>
# include <mln/data/wrap.hh>
# include <mln/value/int_u8.hh>
@@ -115,7 +119,10 @@ namespace scribo
} // end of namespace scribo::debug::internal
- // Return a reference to the logger.
+ /*! \brief Return a reference to the logger.
+
+ \ingroup grpalgodebug
+ */
scribo::debug::internal::logger_& logger();
diff --git a/scribo/scribo/debug/looks_like_a_text_line_image.hh b/scribo/scribo/debug/looks_like_a_text_line_image.hh
index 376baae..5bfe61e 100644
--- a/scribo/scribo/debug/looks_like_a_text_line_image.hh
+++ b/scribo/scribo/debug/looks_like_a_text_line_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -28,8 +29,8 @@
/// \file
///
-/// Compute an image where components are drawn differently whether
-/// they look like a line or not.
+/// \brief Compute an image where components are drawn differently
+/// whether they look like a line or not.
# include <mln/core/concept/image.hh>
@@ -64,6 +65,8 @@ namespace scribo
/// a text line.
///
/// \return A color image.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
@@ -75,6 +78,9 @@ namespace scribo
/// \overload
/// text_value is set to literal::green.
/// non_text_value is set to literal::red.
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
looks_like_a_text_line_image(const Image<I>& input,
diff --git a/scribo/scribo/debug/mean_and_base_lines_image.hh b/scribo/scribo/debug/mean_and_base_lines_image.hh
index 56055c9..89ee198 100644
--- a/scribo/scribo/debug/mean_and_base_lines_image.hh
+++ b/scribo/scribo/debug/mean_and_base_lines_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -65,6 +66,8 @@ namespace scribo
/// \param[in] baseline_value Value used to draw base lines.
///
/// \return A color image.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
@@ -77,6 +80,9 @@ namespace scribo
/// \overload
/// text_value is set to literal::green.
/// non_text_value is set to literal::red.
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I, typename L>
mln_ch_value(I,value::rgb8)
mean_and_base_lines_image(const Image<I>& input,
diff --git a/scribo/scribo/debug/save_comp_diff.hh b/scribo/scribo/debug/save_comp_diff.hh
index f1f93f0..6b2bc3b 100644
--- a/scribo/scribo/debug/save_comp_diff.hh
+++ b/scribo/scribo/debug/save_comp_diff.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -55,15 +55,16 @@ namespace scribo
/*! \brief Show the difference between two object images.
- \param[in] lbl A component set.
- \param[in] lbl_2 Another component set.
+ \param[in] comps_ref A component set.
+ \param[in] comps_new Another component set.
\param[in] filename The output filename.
+ \ingroup grpalgodebug
*/
template <typename L, typename L2>
void
- save_comp_diff(const component_set<L>& cset_1,
- const component_set<L2>& cset_2,
+ save_comp_diff(const component_set<L>& comps_ref,
+ const component_set<L2>& comps_new,
const std::string& filename);
@@ -72,22 +73,26 @@ namespace scribo
template <typename L, typename L2>
void
- save_comp_diff(const component_set<L>& cset_1,
- const component_set<L2>& cset_2,
+ save_comp_diff(const component_set<L>& comps_ref,
+ const component_set<L2>& comps_new,
const std::string& filename)
{
trace::entering("scribo::debug::save_comp_diff");
image2d<value::rgb8> output;
- initialize(output, cset_1.labeled_image());
+ initialize(output, comps_ref.labeled_image());
data::fill(output, literal::black);
- for_all_comps(i, cset_1)
- data::fill(((output | cset_1(i).bbox()).rw() | (pw::value(cset_1.labeled_image()) == i)).rw(), literal::red);
+ for_all_comps(i, comps_ref)
+ data::fill(((output | comps_ref(i).bbox()).rw()
+ | (pw::value(comps_ref.labeled_image()) == i)).rw(),
+ literal::red);
- for_all_comps(i, cset_2)
- data::fill(((output | cset_2(i).bbox()).rw() | (pw::value(cset_2.labeled_image()) == i)).rw(), literal::green);
+ for_all_comps(i, comps_new)
+ data::fill(((output | comps_new(i).bbox()).rw()
+ | (pw::value(comps_new.labeled_image()) == i)).rw(),
+ literal::green);
io::ppm::save(output, filename);
diff --git a/scribo/scribo/debug/save_label_image.hh b/scribo/scribo/debug/save_label_image.hh
index b41325e..af138d9 100644
--- a/scribo/scribo/debug/save_label_image.hh
+++ b/scribo/scribo/debug/save_label_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2009 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2009, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -28,7 +29,7 @@
/// \file
///
-/// Save a labeled image in a color image.
+/// \brief Save a labeled image in a color image.
# include <mln/core/concept/image.hh>
# include <mln/labeling/colorize.hh>
@@ -43,11 +44,13 @@ namespace scribo
using namespace mln;
- /// Save a labeled image in a color image.
+ /// \brief Save a labeled image in a color image.
///
/// \param[in] lbl A label image.
/// \param[in] nlabels The number of labels.
/// \param[in] filename The output file name.
+ ///
+ /// \ingroup grpalgodebug
//
template <typename I>
void
diff --git a/scribo/scribo/debug/save_table_image.hh b/scribo/scribo/debug/save_table_image.hh
index 746d162..8a5c01f 100644
--- a/scribo/scribo/debug/save_table_image.hh
+++ b/scribo/scribo/debug/save_table_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2009 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2009, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -28,7 +29,7 @@
/// \file
///
-/// Save table line bounding boxes in an image.
+/// \brief Save table line bounding boxes in an image.
# include <string>
@@ -53,6 +54,9 @@ namespace scribo
/// Save lines bounding boxes in a copy of \p input_.
/// Bounding boxes are displayed with \p bbox_color.
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename I>
void
save_table_image(const Image<I>& input_,
@@ -64,6 +68,9 @@ namespace scribo
/// Save lines bounding boxes in an image defined on \p input_domain
/// filled with \p bg_color.
/// Bounding boxes are displayed with \p bbox_color.
+ ///
+ /// \ingroup grpalgodebug
+ //
template <typename S>
void
save_table_image(const Site_Set<S>& input_domain,
diff --git a/scribo/scribo/debug/text_areas_image.hh b/scribo/scribo/debug/text_areas_image.hh
index 5b9f086..6aaa64f 100644
--- a/scribo/scribo/debug/text_areas_image.hh
+++ b/scribo/scribo/debug/text_areas_image.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -59,6 +60,7 @@ namespace scribo
/*! \brief Compute an image including detected text areas only.
+ \ingroup grpalgodebug
*/
template <typename I, typename L>
mln_concrete(I)
diff --git a/scribo/scribo/debug/text_color_image.hh b/scribo/scribo/debug/text_color_image.hh
index f8d76c8..182b71b 100644
--- a/scribo/scribo/debug/text_color_image.hh
+++ b/scribo/scribo/debug/text_color_image.hh
@@ -49,6 +49,10 @@ namespace scribo
{
using namespace mln;
+ /*! \brief Draw text components with their respective colors.
+
+ \ingroup grpalgodebug
+ */
template <typename L>
image2d<value::rgb8>
text_color_image(const document<L>& doc);
diff --git a/scribo/scribo/draw/bounding_box_links.hh b/scribo/scribo/draw/bounding_box_links.hh
index 830da25..2816d4e 100644
--- a/scribo/scribo/draw/bounding_box_links.hh
+++ b/scribo/scribo/draw/bounding_box_links.hh
@@ -50,18 +50,16 @@ namespace scribo
using namespace mln;
- /// Draw a list of bounding box links. Draw links from the bounding
- /// box centers.
+ /// \brief Draw a list of bounding box links.
///
- /// \param[in,out] input_ An image where to draw.
- /// \param[in] bboxes Bounding boxes.
- /// \param[in] links Bounding box links.
+ /// \param[in,out] input An image where to draw.
+ /// \param[in] link component links.
/// \param[in] value Value used to draw links.
/// \param[in] anchor Anchor from where the links are drawn.
//
template <typename I, typename L>
void
- bounding_box_links(Image<I>& input_,
+ bounding_box_links(Image<I>& input,
const object_links<L>& link,
const mln_value(I)& value,
anchor::Type anchor);
@@ -72,24 +70,26 @@ namespace scribo
//
template <typename I, typename L>
void
- bounding_box_links(Image<I>& input_,
+ bounding_box_links(Image<I>& input,
const object_links<L>& link,
const mln_value(I)& value);
- /// Draw left, right and validated lists of bounding box links.
- /// Draw from anchors.
+ /// \brief Draw left, right and validated lists of bounding box
+ /// links.
///
- /// \param[in,out] input_ An image where to draw.
- /// \param[in] left_link Bounding box left links.
- /// \param[in] right_link Bounding box right links.
+ /// \param[in,out] input An image where to draw.
+ /// \param[in] left_link Component's left links.
+ /// \param[in] right_link Component's right links.
/// \param[in] left_link_value Value used to draw left links.
/// \param[in] right_link_value Value used to draw right links.
/// \param[in] validated_link_value Value used to draw validated links.
+ /// \param[in] anchor Anchor from where the links are drawn.
+ //
template <typename I, typename L>
inline
void
- bounding_box_links(Image<I>& input_,
+ bounding_box_links(Image<I>& input,
const object_links<L>& left_link,
const object_links<L>& right_link,
const mln_value(I)& left_link_value,
@@ -98,18 +98,18 @@ namespace scribo
anchor::Type anchor);
- /// Draw a graph of bounding box links.
+ /// \brief Draw a graph of bounding box links.
/// Draw from bounding box centers.
///
- /// \param[in,out] input_ An image where to draw.
- /// \param[in] bboxes Bounding boxes.
- /// \param[in] g_ The link graph.
+ /// \param[in,out] input An image where to draw.
+ /// \param[in] g The link graph.
/// \param[in] link_value The value used to draw the links.
+ //
template <typename I, typename G>
inline
void
- bounding_box_links(Image<I>& input_,
- const Graph<G>& g_,
+ bounding_box_links(Image<I>& input,
+ const Graph<G>& g,
const mln_value(I)& link_value);
diff --git a/scribo/scribo/filter/images_in_paragraph.hh b/scribo/scribo/filter/images_in_paragraph.hh
index cf77892..9a88c51 100644
--- a/scribo/scribo/filter/images_in_paragraph.hh
+++ b/scribo/scribo/filter/images_in_paragraph.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Invalidate false positive separators.
+/// Invalidate false positive images.
/// \fixme Share same test canvas as text::merging.
@@ -47,16 +47,12 @@ namespace scribo
using namespace mln;
- /// Invalidate false positive separators.
+ /// Invalidate false positive images.
///
- /// \param[in] separators A paragraph set.
- ///
- /// \return A doc with invalidated separators.
- ///
- /// Warning: it does not remove separators from separator
- /// image. It only invalidate separator components in their
- /// respective component_set.
+ /// \param[in,out] doc A document structure.
///
+ /// \ingroup grpalgofilterelt
+ //
template <typename L>
void
images_in_paragraph(document<L>& doc);
diff --git a/scribo/scribo/filter/line_links_x_height.hh b/scribo/scribo/filter/line_links_x_height.hh
index 0fcc341..c3b2909 100644
--- a/scribo/scribo/filter/line_links_x_height.hh
+++ b/scribo/scribo/filter/line_links_x_height.hh
@@ -41,6 +41,15 @@ namespace scribo
namespace filter
{
+ /*! \brief Filter line links according to character x height.
+
+ \param[in] links Line links information.
+
+ \return A new line_links structure where some links may have
+ been invalidated.
+
+ \ingroup grpalgofilterlinelink
+ */
template <typename L>
line_links<L>
line_links_x_height(const line_links<L>& links);
diff --git a/scribo/scribo/filter/object_groups_mean_width.hh b/scribo/scribo/filter/object_groups_mean_width.hh
index b9c7b8a..2386517 100644
--- a/scribo/scribo/filter/object_groups_mean_width.hh
+++ b/scribo/scribo/filter/object_groups_mean_width.hh
@@ -55,6 +55,7 @@ namespace scribo
\return Filtered object group information.
+ \ingroup grpalgofiltercompgroup
*/
template <typename L>
object_groups<L>
diff --git a/scribo/scribo/filter/object_groups_size_ratio.hh b/scribo/scribo/filter/object_groups_size_ratio.hh
index 6f2e33e..2607c86 100644
--- a/scribo/scribo/filter/object_groups_size_ratio.hh
+++ b/scribo/scribo/filter/object_groups_size_ratio.hh
@@ -29,6 +29,7 @@
/// \file
///
+/// \brief Invalidates groups with too much thin and high components.
# include <mln/util/array.hh>
@@ -44,6 +45,21 @@ namespace scribo
using namespace mln;
+ /*! \brief Invalidates groups with too much thin and high
+ components.
+
+ For each components in the group, it computes the height/width
+ ratio. If it is higher or equal to \p max_size_ratio, the
+ component is counted as invalid.
+
+ If there are \p max_invalid_ratio_per_group invalid
+ components, the group is invalidated.
+
+ \return An object_group structure potentially with invalided
+ groups.
+
+ \ingroup grpalgofiltercompgroup
+ */
template <typename L>
object_groups<L>
object_groups_size_ratio(const object_groups<L>& groups,
diff --git a/scribo/scribo/filter/object_groups_small.hh b/scribo/scribo/filter/object_groups_small.hh
index dd92852..5a698cb 100644
--- a/scribo/scribo/filter/object_groups_small.hh
+++ b/scribo/scribo/filter/object_groups_small.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove small objects groups.
+/// \brief Invalidate groups with few components.
# include <mln/util/array.hh>
@@ -45,14 +45,15 @@ namespace scribo
using namespace mln;
- /*! \brief Remove objects within a group with less than \p n
- links.
+ /*! \brief Invalidate groups with few components.
\param[in] groups Information about object groups.
\param[in] n_links The minimum number of links per group.
\return A copy of object group in which small groups have been
- removed.
+ invalidated.
+
+ \ingroup grpalgofiltercompgroup
*/
template <typename L>
object_groups<L>
diff --git a/scribo/scribo/filter/object_groups_with_holes.hh b/scribo/scribo/filter/object_groups_with_holes.hh
index a59beec..ca55980 100644
--- a/scribo/scribo/filter/object_groups_with_holes.hh
+++ b/scribo/scribo/filter/object_groups_with_holes.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -69,8 +70,16 @@ namespace scribo
using namespace mln;
/*! \brief Remove groups not having at least two background
- * components of \p min_size pixels.
+ components of \p min_size pixels.
+ In order to verify the property of "having two background
+ components", a component group must have at least one
+ component with a hole.
+
+ Example: the letter 'o' has two background components: outside
+ and inside the letter.
+
+ \ingroup grpalgofiltercompgroup
*/
template <typename L>
inline
diff --git a/scribo/scribo/filter/object_links_aligned.hh b/scribo/scribo/filter/object_links_aligned.hh
index e37bcb6..6d76d95 100644
--- a/scribo/scribo/filter/object_links_aligned.hh
+++ b/scribo/scribo/filter/object_links_aligned.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Invalidate links between two components if their not aligned
+/// \brief Invalidate links between two components if their not aligned
/// according to an anchor.
# include <scribo/core/macros.hh>
@@ -44,7 +44,8 @@ namespace scribo
using namespace mln;
- /*! \brief Invalidate links between two components according to a specific anchor.
+ /*! \brief Invalidate links between two components according to a
+ specific anchor.
\param[in] links Object links information.
\param[in] max_alpha Maximum angle value (degrees).
@@ -74,6 +75,7 @@ namespace scribo
Example with an anchor set to the bottom right :
The angle between the two bottoms must be lower than \p max_alpha.
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_bbox_h_ratio.hh b/scribo/scribo/filter/object_links_bbox_h_ratio.hh
index 5573f16..2c3256a 100644
--- a/scribo/scribo/filter/object_links_bbox_h_ratio.hh
+++ b/scribo/scribo/filter/object_links_bbox_h_ratio.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2009 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2009, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -28,7 +29,8 @@
/// \file
///
-/// Invalidate links between two objects with too different height.
+/// \brief Invalidate links between two objects with too different
+/// height.
///
/// \todo rename to object_links_bbox_v_ratio (v for vertical) to be
/// consistent with other routine names.
@@ -57,6 +59,8 @@ namespace scribo
bounding boxes.
\result A filtered object link information.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_bbox_overlap.hh b/scribo/scribo/filter/object_links_bbox_overlap.hh
index 29b71ba..d09839a 100644
--- a/scribo/scribo/filter/object_links_bbox_overlap.hh
+++ b/scribo/scribo/filter/object_links_bbox_overlap.hh
@@ -29,8 +29,8 @@
/// \file
///
-/// Invalidate links between two components having their bounding box
-/// overlapping too much.
+/// \brief Invalidate links between two components having their
+/// bounding box overlapping too much.
# include <mln/math/min.hh>
@@ -59,6 +59,8 @@ namespace scribo
areas.
\result A filtered object link information.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_bbox_ratio.hh b/scribo/scribo/filter/object_links_bbox_ratio.hh
index d6c9286..2a2afa1 100644
--- a/scribo/scribo/filter/object_links_bbox_ratio.hh
+++ b/scribo/scribo/filter/object_links_bbox_ratio.hh
@@ -29,8 +29,8 @@
/// \file
///
-/// Invalidate links between two objects with too different height or
-/// width.
+/// \brief Invalidate links between two objects with too different
+/// height or width.
# include <mln/util/array.hh>
@@ -56,6 +56,8 @@ namespace scribo
bounding boxes.
\result A filtered object link information.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_bbox_w_ratio.hh b/scribo/scribo/filter/object_links_bbox_w_ratio.hh
index 07adcee..7bcb3d3 100644
--- a/scribo/scribo/filter/object_links_bbox_w_ratio.hh
+++ b/scribo/scribo/filter/object_links_bbox_w_ratio.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,8 @@
/// \file
///
-/// Invalidate links between two objects with too different width.
+/// \brief Invalidate links between two objects with too different
+/// width.
///
/// \todo rename to object_links_bbox_h_ratio (h for horizontal) to be
/// consistent with other routine names.
@@ -57,6 +58,8 @@ namespace scribo
bounding boxes.
\result A filtered object link information.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_bottom_aligned.hh b/scribo/scribo/filter/object_links_bottom_aligned.hh
index 2329f37..353c2f6 100644
--- a/scribo/scribo/filter/object_links_bottom_aligned.hh
+++ b/scribo/scribo/filter/object_links_bottom_aligned.hh
@@ -29,8 +29,8 @@
/// \file
///
-/// Invalidate links between two components if their bottom are not
-/// aligned.
+/// \brief Invalidate links between two components if their bottom are
+/// not aligned.
# include <mln/util/array.hh>
@@ -77,6 +77,7 @@ namespace scribo
The angle between the two bottoms must be lower than \p max_alpha.
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_center_aligned.hh b/scribo/scribo/filter/object_links_center_aligned.hh
index da18b00..adbfad8 100644
--- a/scribo/scribo/filter/object_links_center_aligned.hh
+++ b/scribo/scribo/filter/object_links_center_aligned.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2009 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2009, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -28,8 +29,8 @@
/// \file
///
-/// Invalidate links between two objects if their center are not
-/// aligned.
+/// \brief Invalidate links between two objects if their center are
+/// not aligned.
# include <mln/util/array.hh>
@@ -50,10 +51,11 @@ namespace scribo
/*! \brief Invalidate links between two objects if their center are not
aligned.
- \param[in] objects An object image.
\param[in] links Object links information.
\param[in] max_alpha Maximum angle value (degrees).
+ \return New link data.
+
\verbatim
------ ------
@@ -70,6 +72,7 @@ namespace scribo
The angle between the two bottoms must be lower than \p max_alpha.
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_left_aligned.hh b/scribo/scribo/filter/object_links_left_aligned.hh
index 830678a..9a50a20 100644
--- a/scribo/scribo/filter/object_links_left_aligned.hh
+++ b/scribo/scribo/filter/object_links_left_aligned.hh
@@ -29,10 +29,10 @@
/// \file
///
-/// Invalidate links between two objects if their left are not
+/// \brief Invalidate links between two objects if their left are not
/// aligned.
///
-/// \fixme UPDATE DOC!
+/// \todo UPDATE DOC!
# include <mln/util/array.hh>
@@ -49,8 +49,8 @@ namespace scribo
using namespace mln;
- /*! \brief Invalidate links between two objects if their left are not
- aligned.
+ /*! \brief Invalidate links between two objects if their left are
+ not aligned.
\param[in] links Object links information.
\param[in] max_alpha Maximum angle value (degrees).
@@ -77,6 +77,8 @@ namespace scribo
\endverbatim
The angle between the two lefts must be lower than \p max_alpha.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_non_aligned_simple.hh b/scribo/scribo/filter/object_links_non_aligned_simple.hh
index f8db5cf..77c5b1c 100644
--- a/scribo/scribo/filter/object_links_non_aligned_simple.hh
+++ b/scribo/scribo/filter/object_links_non_aligned_simple.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Invalidate links between two non aligned components.
+/// \brief Invalidate links between two non aligned components.
# include <mln/util/array.hh>
@@ -52,6 +52,7 @@ namespace scribo
Alignment is based on a given anchor of object bounding boxes.
\param[in] links Object links information.
+ \param[in] anchor Anchor from where the links are made.
\param[in] max_alpha Maximum angle value (degrees).
@@ -79,6 +80,7 @@ namespace scribo
The angle between the two bottoms must be lower than \p alpha.
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_right_aligned.hh b/scribo/scribo/filter/object_links_right_aligned.hh
index 4f854e2..018cf97 100644
--- a/scribo/scribo/filter/object_links_right_aligned.hh
+++ b/scribo/scribo/filter/object_links_right_aligned.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Invalidate links between two objects if their right are not
+/// \brief Invalidate links between two objects if their right are not
/// aligned.
@@ -51,7 +51,6 @@ namespace scribo
/*! \brief Invalidate links between two objects if their right are not
aligned.
- \param[in] objects An object image.
\param[in] links Object links information.
\param[in] max_alpha Maximum angle value (degrees).
@@ -77,6 +76,8 @@ namespace scribo
\endverbatim
The angle between the two rights must be lower than \p max_alpha.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/object_links_top_aligned.hh b/scribo/scribo/filter/object_links_top_aligned.hh
index 48598c6..6489ea4 100644
--- a/scribo/scribo/filter/object_links_top_aligned.hh
+++ b/scribo/scribo/filter/object_links_top_aligned.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Invalidate links between two objects if their top are not
+/// \brief Invalidate links between two objects if their top are not
/// aligned.
@@ -48,8 +48,8 @@ namespace scribo
using namespace mln;
- /*! \brief Invalidate links between two objects if their top are not
- aligned.
+ /*! \brief Invalidate links between two objects if their top are
+ not aligned.
\param[in] links Object links information.
\param[in] max_alpha Maximum angle value (degrees).
@@ -77,6 +77,8 @@ namespace scribo
\endverbatim
The angle between the two tops must be lower than \p max_alpha.
+
+ \ingroup grpalgofiltercomplink
*/
template <typename L>
object_links<L>
diff --git a/scribo/scribo/filter/objects_h_thick.hh b/scribo/scribo/filter/objects_h_thick.hh
index 06b75dc..7dc7fa5 100644
--- a/scribo/scribo/filter/objects_h_thick.hh
+++ b/scribo/scribo/filter/objects_h_thick.hh
@@ -54,6 +54,8 @@ namespace scribo
/// \param[in] max_thickness The maximum thickness value.
///
/// \result A binary image without thick objects.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename I, typename N, typename V>
inline
@@ -70,6 +72,8 @@ namespace scribo
/// \param[in] max_thickness The minimum thickness value.
///
/// \result A component data set without too thick components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
inline
diff --git a/scribo/scribo/filter/objects_h_thin.hh b/scribo/scribo/filter/objects_h_thin.hh
index 8caa2cc..7b1d18a 100644
--- a/scribo/scribo/filter/objects_h_thin.hh
+++ b/scribo/scribo/filter/objects_h_thin.hh
@@ -53,6 +53,8 @@ namespace scribo
/// \param[in] min_thinness the minimum thinness value.
///
/// \result A binary image without h_thin components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename I, typename N, typename V>
inline
@@ -68,6 +70,8 @@ namespace scribo
/// \param[in] min_thinness the minimum thinness value.
///
/// \result An object image without too thin vertical components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
inline
diff --git a/scribo/scribo/filter/objects_large.hh b/scribo/scribo/filter/objects_large.hh
index 8dfec9d..96d1817 100644
--- a/scribo/scribo/filter/objects_large.hh
+++ b/scribo/scribo/filter/objects_large.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove large objects in a binary image.
+/// \brief Remove large objects in a binary image.
# include <mln/core/concept/image.hh>
@@ -56,7 +56,8 @@ namespace scribo
using namespace mln;
- /// Remove large objects in a binary image.
+ /// \brief Remove large objects in a binary image.
+ ///
/// Set to 'false' all the removed objects.
///
/// \param[in] input_ A binary image.
@@ -65,6 +66,9 @@ namespace scribo
/// \param[in] max_size The minimum cardinality of an object.
///
/// \return A binary image without large objects.
+ ///
+ /// \ingroup grpalgofiltercomp
+ //
template <typename I, typename N, typename V>
mln_concrete(I)
components_large(const Image<I>& input_,
@@ -72,13 +76,16 @@ namespace scribo
const V& label_type,
unsigned max_size);
- /// Remove too large components.
+ /// \brief Remove too large components.
///
/// \param[in] components An object image.
/// \param[in] max_size The maximum cardinality of an object.
///
/// \return A component set with large components set to
/// component::Ignored.
+ ///
+ /// \ingroup grpalgofiltercomp
+ //
template <typename L>
inline
component_set<L>
diff --git a/scribo/scribo/filter/objects_on_border.hh b/scribo/scribo/filter/objects_on_border.hh
index d6c4cf5..55578a4 100644
--- a/scribo/scribo/filter/objects_on_border.hh
+++ b/scribo/scribo/filter/objects_on_border.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Remove large objects in a binary image.
+/// \brief Remove components located on image borders.
# include <mln/core/concept/image.hh>
@@ -55,13 +55,15 @@ namespace scribo
using namespace mln;
- /// Remove too large components.
+ /// \brief Remove components located on image borders.
///
/// \param[in] components An object image.
- /// \param[in] max_size The maximum cardinality of an object.
///
/// \return A component set with large components set to
/// component::Ignored.
+ ///
+ /// \ingroup grpalgofiltercomp
+ //
template <typename L>
inline
component_set<L>
diff --git a/scribo/scribo/filter/objects_size_ratio.hh b/scribo/scribo/filter/objects_size_ratio.hh
index 6edb35f..9efa8bf 100644
--- a/scribo/scribo/filter/objects_size_ratio.hh
+++ b/scribo/scribo/filter/objects_size_ratio.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -50,6 +50,8 @@ namespace scribo
///
/// If the height/width ratio is lower than min_size_ratio then
/// the component is invalidated.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
component_set<L>
diff --git a/scribo/scribo/filter/objects_small.hh b/scribo/scribo/filter/objects_small.hh
index c18c86f..921aea2 100644
--- a/scribo/scribo/filter/objects_small.hh
+++ b/scribo/scribo/filter/objects_small.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove small components in a binary image.
+/// \brief Remove small components in a binary image.
# include <mln/core/concept/image.hh>
@@ -57,7 +57,8 @@ namespace scribo
using namespace mln;
- /// Remove small components in a binary image.
+ /// \brief Remove small components in a binary image.
+ ///
/// Set to 'false' all the removed components.
///
/// \param[in] input_ A binary image.
@@ -66,6 +67,9 @@ namespace scribo
/// \param[in] min_size The minimum cardinality of an object.
///
/// \return A binary image without small components.
+ ///
+ /// \ingroup grpalgofiltercomp
+ //
template <typename I, typename N, typename V>
mln_concrete(I)
components_small(const Image<I>& input_,
@@ -74,13 +78,16 @@ namespace scribo
unsigned min_size);
- /// Remove too small components.
+ /// \brief Remove too small components.
///
/// \param[in] components An object image.
/// \param[in] min_size The minimum cardinality of an object.
///
/// \return A component set with small components set to
/// component::Ignored.
+ ///
+ /// \ingroup grpalgofiltercomp
+ //
template <typename L>
component_set<L>
components_small(const component_set<L>& components,
diff --git a/scribo/scribo/filter/objects_thick.hh b/scribo/scribo/filter/objects_thick.hh
index faf13e7..8c3a03b 100644
--- a/scribo/scribo/filter/objects_thick.hh
+++ b/scribo/scribo/filter/objects_thick.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove too thick components.
+/// \brief Remove too thick components.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -49,7 +49,7 @@ namespace scribo
using namespace mln;
- /// Remove components thicker or equal to \p max_thickness.
+ /// \brief Remove components thicker or equal to \p max_thickness.
///
/// \param[in] input_ A binary image.
/// \param[in] nbh_ A neighborhood used in labeling algorithms.
@@ -57,6 +57,8 @@ namespace scribo
/// \param[in] max_thickness The maximum thickness value.
///
/// \result A binary image without thick components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename I, typename N, typename V>
inline
@@ -67,12 +69,14 @@ namespace scribo
unsigned max_thickness);
- /// Remove components thicker or equal to \p max_thickness.
+ /// \brief Remove components thicker or equal to \p max_thickness.
///
/// \param[in] components An object image.
/// \param[in] max_thickness The maximum thickness value.
///
/// \result An object image without too thick components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
inline
diff --git a/scribo/scribo/filter/objects_thin.hh b/scribo/scribo/filter/objects_thin.hh
index f58d0db..d376856 100644
--- a/scribo/scribo/filter/objects_thin.hh
+++ b/scribo/scribo/filter/objects_thin.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove too thin components.
+/// \brief Remove too thin components.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -52,7 +52,7 @@ namespace scribo
using namespace mln;
- /// Remove components thinner or equal to \p min_thickness.
+ /// \brief Remove components thinner or equal to \p min_thickness.
///
/// \param[in] input_ a binary image.
/// \param[in] nbh_ a neighborhood used in labeling algorithms.
@@ -60,6 +60,8 @@ namespace scribo
/// \param[in] min_thickness the minimum thickness value.
///
/// \result A binary image without thin components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename I, typename N, typename V>
inline
@@ -69,12 +71,15 @@ namespace scribo
const V& label_type,
unsigned min_thickness);
- /// Remove lines of text thinner or equal to \p min_thickness.
+ /// \brief Remove lines of text thinner or equal to \p
+ /// min_thickness.
///
/// \param[in] components An object image.
/// \param[in] min_thickness the minimum thickness value.
///
/// \result An object image without too thin components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
inline
diff --git a/scribo/scribo/filter/objects_v_thick.hh b/scribo/scribo/filter/objects_v_thick.hh
index ad1a4df..e9e85a1 100644
--- a/scribo/scribo/filter/objects_v_thick.hh
+++ b/scribo/scribo/filter/objects_v_thick.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove too verticaly thick components.
+/// \brief Remove too verticaly thick components.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -46,7 +46,8 @@ namespace scribo
using namespace mln;
- /// Remove components verticaly thicker or equal to \p max_thickness.
+ /// \brief Remove components verticaly thicker or equal to \p
+ /// max_thickness.
///
/// \param[in] input_ A binary image.
/// \param[in] nbh_ A neighborhood used in labeling algorithms.
@@ -54,6 +55,8 @@ namespace scribo
/// \param[in] max_thickness The maximum thickness value.
///
/// \result A binary image without thick components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename I, typename N, typename V>
inline
@@ -64,12 +67,15 @@ namespace scribo
unsigned max_thickness);
- /// Remove components verticaly thicker or equal to \p max_thickness.
+ /// \brief Remove components verticaly thicker or equal to \p
+ /// max_thickness.
///
/// \param[in] comps A component set.
/// \param[in] max_thickness The maximum thickness value.
///
/// \result An object image without too thick components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
inline
diff --git a/scribo/scribo/filter/objects_v_thin.hh b/scribo/scribo/filter/objects_v_thin.hh
index f2280eb..c5364b3 100644
--- a/scribo/scribo/filter/objects_v_thin.hh
+++ b/scribo/scribo/filter/objects_v_thin.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Remove too thin components.
+/// \brief Remove too thin components.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -46,34 +46,39 @@ namespace scribo
using namespace mln;
- /// Remove components thinner or equal to \p min_thinness.
+ /// \brief Remove components thinner or equal to \p min_thinness.
///
- /// \param[in] input_ a binary image.
- /// \param[in] nbh_ a neighborhood used in labeling algorithms.
+ /// \param[in] input a binary image.
+ /// \param[in] nbh a neighborhood used in labeling algorithms.
/// \param[in] label_type the label type used for labeling.
/// \param[in] min_thinness the minimum thinness value.
///
/// \result A binary image without v_thin components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename I, typename N, typename V>
inline
mln_concrete(I)
- objects_v_thin(const Image<I>& input_,
- const Neighborhood<N>& nbh_,
+ objects_v_thin(const Image<I>& input,
+ const Neighborhood<N>& nbh,
const V& label_type,
unsigned min_thinness);
- /// Remove lines of text thinner or equal to \p min_thinness.
+ /// \brief Remove lines of text thinner or equal to \p
+ /// min_thinness.
///
/// \param[in] comps A component set.
/// \param[in] min_thinness the minimum thinness value.
///
/// \result An object image without too thin vertical components.
+ ///
+ /// \ingroup grpalgofiltercomp
//
template <typename L>
inline
component_set<L>
- components_v_thin(const component_set<L>& text,
+ components_v_thin(const component_set<L>& comps,
unsigned min_thinness);
diff --git a/scribo/scribo/filter/objects_with_holes.hh b/scribo/scribo/filter/objects_with_holes.hh
index 09eaad5..83b9755 100644
--- a/scribo/scribo/filter/objects_with_holes.hh
+++ b/scribo/scribo/filter/objects_with_holes.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -81,8 +81,17 @@ namespace scribo
/*! \brief Remove components having a minimum number of holes.
+ \param[in] components A component set.
+ \param[in] min_holes_count If a component have at least \p
+ min_holes_count holes it is
+ invalidated.
+ \param[in] min_size The minimum hole area to take a hole into
+ account.
+ \return A component where the component having too much holes
+ are invalidated.
+ \ingroup grpalgofiltercomp
*/
template <typename L>
component_set<L>
@@ -90,7 +99,20 @@ namespace scribo
unsigned min_holes_count,
unsigned min_size);
+ /*! \brief Remove components having at least two holes.
+ This is a fastest version since it is optimized for 2 holes
+ detection.
+
+ \param[in] components A component set.
+ \param[in] min_size The minimum hole area to take a hole into
+ account.
+
+ \return A component where the component having at least two
+ holes are invalidated.
+
+ \ingroup grpalgofiltercomp
+ */
template <typename L>
inline
component_set<L>
diff --git a/scribo/scribo/filter/paragraphs_bbox_overlap.hh b/scribo/scribo/filter/paragraphs_bbox_overlap.hh
index 1cf9793..744b1ee 100644
--- a/scribo/scribo/filter/paragraphs_bbox_overlap.hh
+++ b/scribo/scribo/filter/paragraphs_bbox_overlap.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Remove invalid paragraphs.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Remove invalid paragraphs.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -52,12 +52,15 @@ namespace scribo
using namespace mln;
- /// Remove invalid paragraphs.
+ /// \brief Remove invalid paragraphs.
///
- /// \param[in] paragraphs A paragraph set.
+ /// \param[in] parset A paragraph set.
///
/// \return A paragraph set with invalid paragraphs tag set to
/// Paragraph::Ignored.
+ ///
+ /// \ingroup grpalgofilterpar
+ //
template <typename L>
paragraph_set<L>
paragraphs_bbox_overlap(const paragraph_set<L>& parset);
diff --git a/scribo/scribo/filter/paragraphs_in_borders.hh b/scribo/scribo/filter/paragraphs_in_borders.hh
index 8953282..c3b9623 100644
--- a/scribo/scribo/filter/paragraphs_in_borders.hh
+++ b/scribo/scribo/filter/paragraphs_in_borders.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Invalidate false positive paragraphs.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Invalidate false positive paragraphs.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -46,7 +46,7 @@ namespace scribo
using namespace mln;
- /// Invalidate paragraphs located close to the image borders.
+ /// \brief Invalidate paragraphs located close to the image borders.
///
/// \param[in,out] doc A document structure.
///
@@ -66,6 +66,8 @@ namespace scribo
/// -----------
///
/// \endverbatim
+ ///
+ /// \ingroup grpalgofilterpar
//
template <typename L>
void
diff --git a/scribo/scribo/filter/paragraphs_in_image.hh b/scribo/scribo/filter/paragraphs_in_image.hh
index be15b06..d95864d 100644
--- a/scribo/scribo/filter/paragraphs_in_image.hh
+++ b/scribo/scribo/filter/paragraphs_in_image.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Remove invalid paragraphs.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Remove invalid paragraphs.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -59,12 +59,15 @@ namespace scribo
using namespace mln;
- /// Remove invalid paragraphs.
+ /// \brief Remove invalid paragraphs.
///
- /// \param[in] paragraphs A paragraph set.
+ /// \param[in,out] doc A document structure.
///
/// \return A paragraph set with invalid paragraphs tag set to
/// Paragraph::Ignored.
+ ///
+ /// \ingroup grpalgofilterpar
+ //
template <typename L>
void
paragraphs_in_image(document<L>& doc);
diff --git a/scribo/scribo/filter/separators_in_borders.hh b/scribo/scribo/filter/separators_in_borders.hh
index 8ccb6b1..f5048d2 100644
--- a/scribo/scribo/filter/separators_in_borders.hh
+++ b/scribo/scribo/filter/separators_in_borders.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Invalidate false positive separators.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Invalidate false positive separators.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -46,9 +46,16 @@ namespace scribo
using namespace mln;
- /// Invalidate separators located close to the image borders.
+ /// \brief Invalidate separators located close to the image
+ /// borders.
///
/// \param[in,out] doc A document structure.
+ /// \param[in] vratio Ratio to be used for evaluating the inner
+ /// border size in which vertical separators
+ /// will be invalidated.
+ /// \param[in] hratio Ratio to be used for evaluating the inner
+ /// border size in which horizontal separators
+ /// will be invalidated.
///
/// Warning: it does not remove separators from separator
/// image. It only invalidate separator components in their
@@ -66,6 +73,8 @@ namespace scribo
/// -----------
///
/// \endverbatim
+ ///
+ /// \ingroup grpalgofilterelt
//
template <typename L>
void
diff --git a/scribo/scribo/filter/separators_in_element.hh b/scribo/scribo/filter/separators_in_element.hh
index 7e8a437..d4f587b 100644
--- a/scribo/scribo/filter/separators_in_element.hh
+++ b/scribo/scribo/filter/separators_in_element.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Invalidate false positive separators.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Invalidate false positive separators.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -47,16 +47,16 @@ namespace scribo
using namespace mln;
- /// Invalidate false positive separators.
+ /// \brief Invalidate false positive separators.
///
- /// \param[in] separators A paragraph set.
- ///
- /// \return A doc with invalidated separators.
+ /// \param[in,out] doc A document structure.
///
/// Warning: it does not remove separators from separator
/// image. It only invalidate separator components in their
/// respective component_set.
///
+ /// \ingroup grpalgofilterelt
+ //
template <typename L>
void
separators_in_element(document<L>& doc);
diff --git a/scribo/scribo/filter/separators_in_paragraph.hh b/scribo/scribo/filter/separators_in_paragraph.hh
index ac399c3..eae12e9 100644
--- a/scribo/scribo/filter/separators_in_paragraph.hh
+++ b/scribo/scribo/filter/separators_in_paragraph.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Invalidate false positive separators.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Invalidate false positive separators.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -47,26 +47,32 @@ namespace scribo
using namespace mln;
- /// Invalidate false positive separators.
+ /// \brief Invalidate false positive separators.
///
- /// \param[in] separators A paragraph set.
- ///
- /// \return A doc with invalidated separators.
+ /// \param[in,out] doc A document structure.
+ /// \param[in] hmin_size Minimum width of a line to be considered
+ /// as separators.
+ /// \param[in] vmin_size Minimum height of a line to be considered
+ /// as separators.
///
/// Warning: it does not remove separators from separator
/// image. It only invalidate separator components in their
/// respective component_set.
///
+ /// \ingroup grpalgofilterelt
+ //
template <typename L>
void
- separators_in_paragraph(document<L>& doc, unsigned hmax_size, unsigned vmax_size);
+ separators_in_paragraph(document<L>& doc,
+ unsigned hmin_size, unsigned vmin_size);
# ifndef MLN_INCLUDE_ONLY
template <typename L>
void
- separators_in_paragraph(document<L>& doc, unsigned hmax_size, unsigned vmax_size)
+ separators_in_paragraph(document<L>& doc,
+ unsigned hmin_size, unsigned vmin_size)
{
trace::entering("scribo::filter::separators_in_paragraph");
@@ -106,7 +112,7 @@ namespace scribo
// This separator is included in an element (picture, drawing...)
// => Ignore it.
if (tl && tr && ml && mc && mr && bl && br
- && hline(c).bbox().width() < hmax_size)
+ && hline(c).bbox().width() < hmin_size)
hline(c).update_tag(component::Ignored);
}
@@ -135,7 +141,7 @@ namespace scribo
// This separator is included in an element (picture, drawing...)
// => Ignore it.
if (tl && tr && ml && mc && mr && bl && br
- && vline(c).bbox().height() < vmax_size)
+ && vline(c).bbox().height() < vmin_size)
vline(c).update_tag(component::Ignored);
}
diff --git a/scribo/scribo/filter/separators_vert_in_borders.hh b/scribo/scribo/filter/separators_vert_in_borders.hh
index 4a9e806..4f7c070 100644
--- a/scribo/scribo/filter/separators_vert_in_borders.hh
+++ b/scribo/scribo/filter/separators_vert_in_borders.hh
@@ -28,8 +28,9 @@
/// \file
///
-/// Invalidate false positive separators.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Invalidate vertical separators located close to the image
+/// borders.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -47,7 +48,8 @@ namespace scribo
using namespace mln;
- /// Invalidate separators located close to the image borders.
+ /// \brief Invalidate vertical separators located close to the image
+ /// borders.
///
/// \param[in,out] doc A document structure.
///
@@ -67,6 +69,8 @@ namespace scribo
/// -----------
///
/// \endverbatim
+ ///
+ /// \ingroup grpalgofilterelt
//
template <typename L>
void
diff --git a/scribo/scribo/fun/v2b/objects_on_border_filter.hh b/scribo/scribo/fun/v2b/objects_on_border_filter.hh
index e6f1738..60655ae 100644
--- a/scribo/scribo/fun/v2b/objects_on_border_filter.hh
+++ b/scribo/scribo/fun/v2b/objects_on_border_filter.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Remove large components in a binary image.
+/// \brief Remove components in image inner border.
@@ -66,14 +66,13 @@ namespace scribo
/// Constructor
///
/// \param[in] components Component bounding boxes.
- /// \param[in] max_size Maximum component size.
//
components_on_border_filter(const component_set<L>& components);
/// Check if the component is large enough.
///
- /// \param l A label.
+ /// \param[in] l A label.
///
/// \return false if the component area is strictly inferion to
/// \p max_size_.
diff --git a/scribo/scribo/postprocessing/images_to_drop_capital.hh b/scribo/scribo/postprocessing/images_to_drop_capital.hh
index f7d18a6..727f23c 100644
--- a/scribo/scribo/postprocessing/images_to_drop_capital.hh
+++ b/scribo/scribo/postprocessing/images_to_drop_capital.hh
@@ -28,8 +28,8 @@
/// \file
///
-/// Set type for specific images to Drop Capital component.
-/// \fixme Share same test canvas as text::merging.
+/// \brief Set type for specific images to Drop Capital component.
+/// \todo Share same test canvas as text::merging.
# include <mln/core/concept/image.hh>
@@ -47,12 +47,10 @@ namespace scribo
using namespace mln;
- /// Set type for specific images to Drop Capital component.
+ /// \brief Set type for specific images to Drop Capital component.
///
- /// \param[in] separators A paragraph set.
+ /// \param[in] doc A document structure.
///
- /// \return A doc with images tagged as dropped capital is such
- /// images have been found.
//
template <typename L>
void
diff --git a/scribo/scribo/preprocessing/denoise.hh b/scribo/scribo/preprocessing/denoise.hh
index 1877c06..115ac1d 100644
--- a/scribo/scribo/preprocessing/denoise.hh
+++ b/scribo/scribo/preprocessing/denoise.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -48,10 +49,12 @@ namespace scribo
/// \param[in] input A binary image. True for objects, False for
/// background.
/// \param[in] nbh Neighborhood to use for denoising.
- /// \param[in] min_card Minimum component cardinality to not be
- /// considered as noise.
+ /// \param[in] fg_min_card Minimum component cardinality to not be
+ /// considered as noise in the foreground.
+ /// \param[in] bg_min_card Minimum component cardinality to not be
+ /// considered as noise in the foreground.
///
- /// \output A binary image with the same domain as \p input. All
+ /// \return A binary image with the same domain as \p input. All
/// small components have been removed.
//
template <typename I, typename N>
diff --git a/scribo/scribo/preprocessing/denoise_bg.hh b/scribo/scribo/preprocessing/denoise_bg.hh
index eda3803..1c02f8e 100644
--- a/scribo/scribo/preprocessing/denoise_bg.hh
+++ b/scribo/scribo/preprocessing/denoise_bg.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Denoise image background.
+/// \brief Denoise image background.
# include <mln/core/image/image2d.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -57,7 +57,7 @@ namespace scribo
using namespace mln;
- /// Denoise image background.
+ /// \brief Denoise image background.
///
/// \param[in] input A binary image. True for objects, False for
/// background.
@@ -65,7 +65,7 @@ namespace scribo
/// \param[in] min_card Minimum component cardinality to not be
/// considered as noise.
///
- /// \output A binary image with the same domain as \p input. All
+ /// \return A binary image with the same domain as \p input. All
/// small components have been removed and merged with the
/// background.
//
diff --git a/scribo/scribo/preprocessing/denoise_fg.hh b/scribo/scribo/preprocessing/denoise_fg.hh
index 9701d06..23b838e 100644
--- a/scribo/scribo/preprocessing/denoise_fg.hh
+++ b/scribo/scribo/preprocessing/denoise_fg.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -28,7 +29,7 @@
/// \file
///
-/// Denoise image foreground.
+/// \brief Denoise image foreground.
# include <mln/core/image/image2d.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -56,7 +57,7 @@ namespace scribo
using namespace mln;
- /// Denoise image foreground.
+ /// \brief Denoise image foreground.
///
/// \param[in] input A binary image. True for objects, False for
/// background.
@@ -64,7 +65,7 @@ namespace scribo
/// \param[in] min_card Minimum component cardinality to not be
/// considered as noise.
///
- /// \output A binary image with the same domain as \p input. All
+ /// \return A binary image with the same domain as \p input. All
/// small components have been removed and merged with the
/// background.
//
diff --git a/scribo/scribo/preprocessing/deskew_crop.hh b/scribo/scribo/preprocessing/deskew_crop.hh
index 988e803..2f35932 100644
--- a/scribo/scribo/preprocessing/deskew_crop.hh
+++ b/scribo/scribo/preprocessing/deskew_crop.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2010 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2010, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -40,16 +41,14 @@ namespace scribo
/*! \brief Deskew a region of interest.
- \param[in] input_bin A binary image.
- \param[in] input_gray A gray-level image.
+ \param[in] crop_gl A gray-level image.
+ \param[in] input_gl A gray-level image.
\return A deskewed binary image.
-
-
Handles skew angles from -25 to +25 degrees.
- \p input_bin and \p input_gray must be 2D images and must be
+ \p crop_gl and \p input_gl must be 2D images and must be
identical (e.g. only the value differs).
This algorithm is designed for images created from a region of
@@ -58,7 +57,7 @@ namespace scribo
*/
template <typename I, typename J>
mln_concrete(I)
- deskew(const Image<I>& crop_gl_, const Image<I>& input_gl_);
+ deskew(const Image<I>& crop_gl, const Image<I>& input_gl);
# ifndef MLN_INCLUDE_ONLY
diff --git a/scribo/scribo/preprocessing/homogeneous_contrast.hh b/scribo/scribo/preprocessing/homogeneous_contrast.hh
index 82b2ac6..b510ecd 100644
--- a/scribo/scribo/preprocessing/homogeneous_contrast.hh
+++ b/scribo/scribo/preprocessing/homogeneous_contrast.hh
@@ -52,14 +52,14 @@ namespace scribo
/// \param[in] input A gray-level image.
/// \param[in] h Height attribute value for leveling closing.
///
- /// \output A gray-level image with better contrast homogeneity.
+ /// \return A gray-level image with better contrast homogeneity.
//
template <typename I>
mln_concrete(I)
homogeneous_contrast(const Image<I>& input, unsigned h);
- /// \overload.
+ /// \overload
/// Default height attribute value is set to 75.
//
template <typename I>
diff --git a/scribo/scribo/preprocessing/split_bg_fg.hh b/scribo/scribo/preprocessing/split_bg_fg.hh
index e1fae7d..b35d04c 100644
--- a/scribo/scribo/preprocessing/split_bg_fg.hh
+++ b/scribo/scribo/preprocessing/split_bg_fg.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -30,7 +30,7 @@
/// \file
///
-/// Split the background and the foreground.
+/// \brief Split the background and the foreground.
# include <mln/core/macros.hh>
# include <mln/core/image/image2d.hh>
@@ -89,7 +89,7 @@ namespace scribo
*/
template <typename I>
mln::util::couple<mln_concrete(I), mln_concrete(I)>
- split_bg_fg(const Image<I>& input_, unsigned lambda, unsigned delta);
+ split_bg_fg(const Image<I>& input, unsigned lambda, unsigned delta);
diff --git a/scribo/scribo/primitive/extract/alignments.hh b/scribo/scribo/primitive/extract/alignments.hh
index b2c7c0c..32285ea 100644
--- a/scribo/scribo/primitive/extract/alignments.hh
+++ b/scribo/scribo/primitive/extract/alignments.hh
@@ -26,6 +26,10 @@
#ifndef SCRIBO_PRIMITIVE_EXTRACT_ALIGNMENTS_HH
# define SCRIBO_PRIMITIVE_EXTRACT_ALIGNMENTS_HH
+/// \file
+///
+///
+
# include <iostream>
# include <mln/core/image/image2d.hh>
@@ -112,21 +116,24 @@ namespace scribo
This method handles skewed alignments and draw skew lines if
possible. Examples :
+ \verbatim
|x |x
| x \x
| x -> \x
| x \x
+ \endverbatim
-
- \fixme Because the first link step at step 2.a check aligment
+ \todo Because the first link step at step 2.a check aligment
between only 2 components and not the whole group, we may
retrieve alignment patterns such this one :
+ \verbatim
|x
| x
| x
| x
|x
+ \endverbatim
Here, the 'x' are aligned pair by pair but globally they are
not. Here we cannot draw skewed lines without processing every
@@ -137,12 +144,13 @@ namespace scribo
with that line. Here, we would like to split links/alignement
in two groups in order to get something like that :
+ \verbatim
\x
\x
\x
/x
/x
-
+ \endverbatim
\param[in] doc A document information with text lines.
diff --git a/scribo/scribo/primitive/extract/cells.hh b/scribo/scribo/primitive/extract/cells.hh
index 08b3fd6..4ae3e49 100644
--- a/scribo/scribo/primitive/extract/cells.hh
+++ b/scribo/scribo/primitive/extract/cells.hh
@@ -1,4 +1,5 @@
-// Copyright (C) 2009 EPITA Research and Development Laboratory (LRDE)
+// Copyright (C) 2009, 2011 EPITA Research and Development Laboratory
+// (LRDE)
//
// This file is part of Olena.
//
@@ -60,19 +61,20 @@ namespace scribo
using namespace mln;
- /// Extract canvas cells from a binary image.
+ /// \brief Extract canvas cells from a binary image.
/// Use arbitrary criterions.
///
- /// \param[in] input_ A binary image.
- /// \param[in,out] ncells Will store the number of cells found.
+ /// \param[in] input A binary image.
+ /// \param[in] nbh A neighborhood.
+ /// \param[in,out] label_type Type of the labeled image.
///
/// \return A list of cell bounding boxes.
//
template <typename I, typename N, typename V>
mln::util::couple<mln::util::array<box<mln_site(I)> >,
mln::util::array<box<mln_site(I)> > >
- cells(const Image<I>& input_,
- const Neighborhood<N>& nbh_, const V& label_type);
+ cells(const Image<I>& input,
+ const Neighborhood<N>& nbh, const V& label_type);
# ifndef MLN_INCLUDE_ONLY
diff --git a/scribo/scribo/primitive/extract/components.hh b/scribo/scribo/primitive/extract/components.hh
index fe2b523..851193e 100644
--- a/scribo/scribo/primitive/extract/components.hh
+++ b/scribo/scribo/primitive/extract/components.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Extract components in a binary image.
+/// \brief Extract components in a binary image.
# include <mln/core/concept/neighborhood.hh>
@@ -63,10 +63,10 @@ namespace scribo
using namespace mln;
- /// Extract components in a binary image.
+ /// \brief Extract components in a binary image.
///
- /// \param[in] input A binary image. Components are must be set
- /// to 'true'
+ /// \param[in] input A RGB image.
+ /// \param[in] binary_input A binary image. Components must be set to 'True'.
/// and background to 'false'.
/// \param[in] nbh A neighborhood to be used for labeling.
/// \param[in,out] ncomponents Will store the numbers of components found.
diff --git a/scribo/scribo/primitive/extract/lines_h_single.hh b/scribo/scribo/primitive/extract/lines_h_single.hh
index 1bff2f4..7c9e556 100644
--- a/scribo/scribo/primitive/extract/lines_h_single.hh
+++ b/scribo/scribo/primitive/extract/lines_h_single.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Fast Extraction of single horizontal thick lines.
+/// \brief Fast Extraction of single horizontal thick lines.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -50,16 +50,16 @@ namespace scribo
using namespace mln;
- /// Fast Extraction of single horizontal thick lines.
- /*!
+ /*! \brief Fast Extraction of single horizontal thick lines.
+ *
* Only single non discontinued lines are correctly extracted
* with this routine.
*
- * \param[in] input_ A binary image.
- * \param[in] nbh_ The neighborhood used for labeling image
+ * \param[in] input A binary image.
+ * \param[in] nbh The neighborhood used for labeling image
* components.
* \param[in,out] nlines Type used for labeling.
- * \param[in] line_length The minimum line length.
+ * \param[in] min_line_length The minimum line length.
* \param[in] w_h_ratio The minimum ratio width/height object
* bounding boxes to consider an
* object as a single line.
@@ -75,13 +75,13 @@ namespace scribo
float w_h_ratio);
- /// Fast Extraction of single horizontal thick lines.
- /*!
+ /*! \brief Fast Extraction of single horizontal thick lines.
+ *
* Only single non discontinued lines are correctly extracted
* with this routine.
*
* \param[in] components A labeled image.
- * \param[in] line_length The minimum line length.
+ * \param[in] min_line_length The minimum line length.
* \param[in] w_h_ratio The minimum ratio width/height object
* bounding boxes to consider an
* object as a single line.
diff --git a/scribo/scribo/primitive/extract/lines_h_thick_and_single.hh b/scribo/scribo/primitive/extract/lines_h_thick_and_single.hh
index dd495a4..d0db15d 100644
--- a/scribo/scribo/primitive/extract/lines_h_thick_and_single.hh
+++ b/scribo/scribo/primitive/extract/lines_h_thick_and_single.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Extract horizontal thick lines in a binary image.
+/// \brief Extract horizontal thick lines in a binary image.
# include <mln/core/concept/image.hh>
@@ -56,18 +56,19 @@ namespace scribo
using namespace mln;
- /// Extract horizontal thick lines in a binary image.
- /*!
+
+ /*! \brief Extract horizontal thick lines in a binary image.
+ *
* Only non discontinued lines are correctly extracted with this
* routine. Only lines matching the given criterions are kept
* in the result.
*
- * \param[in] input_ A binary image.
- * \param[in] nbh_ The neighborhood used for labeling image
+ * \param[in] input A binary image.
+ * \param[in] nbh The neighborhood used for labeling image
* components.
* \param[in,out] nlines Type used for labeling.
- * \param[in] line_length The minimum line length.
- * \param[in] w_h_ratio The minimum ratio width/height object
+ * \param[in] min_line_length The minimum line length.
+ * \param[in] h_w_ratio The minimum ratio width/height object
* bounding boxes to consider an
* object as a single line.
*
@@ -75,8 +76,8 @@ namespace scribo
*/
template <typename I, typename N, typename V>
component_set<mln_ch_value(I,V)>
- lines_h_thick_and_single(const Image<I>& input_,
- const Neighborhood<N>& nbh_,
+ lines_h_thick_and_single(const Image<I>& input,
+ const Neighborhood<N>& nbh,
V& nlines,
unsigned min_line_length,
float h_w_ratio);
diff --git a/scribo/scribo/primitive/extract/lines_v_pattern.hh b/scribo/scribo/primitive/extract/lines_v_pattern.hh
index 2908c8b..bf0b1c0 100644
--- a/scribo/scribo/primitive/extract/lines_v_pattern.hh
+++ b/scribo/scribo/primitive/extract/lines_v_pattern.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Extract vertical lines matching a specific pattern.
+/// \brief Extract vertical lines matching a specific pattern.
# include <cmath>
@@ -56,10 +56,12 @@ namespace scribo
using namespace mln;
- /// Extract vertical lines matching a specific pattern.
+ /// \brief Extract vertical lines matching a specific pattern.
///
/// \param[in] input A binary image.
/// \param[in] length The minimum line length.
+ /// \param[in] delta space between the first background pixels
+ /// and the line pixels (usually 2 or 3).
///
/// \result An image of vertical lines.
//
diff --git a/scribo/scribo/primitive/extract/lines_v_single.hh b/scribo/scribo/primitive/extract/lines_v_single.hh
index 67b9f9d..317bf22 100644
--- a/scribo/scribo/primitive/extract/lines_v_single.hh
+++ b/scribo/scribo/primitive/extract/lines_v_single.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Fast Extraction of single vertical thick lines.
+/// \brief Fast Extraction of single vertical thick lines.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -49,16 +49,16 @@ namespace scribo
using namespace mln;
- /// Fast Extraction of single vertical thick lines.
- /*!
+ /*! \brief Fast Extraction of single vertical thick lines.
+ *
* Only single non discontinued lines are correctly extracted
* with this routine.
*
- * \param[in] input_ A binary image.
- * \param[in] nbh_ The neighborhood used for labeling image
+ * \param[in] input A binary image.
+ * \param[in] nbh The neighborhood used for labeling image
* components.
* \param[in,out] nlines Type used for labeling.
- * \param[in] line_length The minimum line length.
+ * \param[in] min_line_length The minimum line length.
* \param[in] h_w_ratio The minimum ratio height/width object
* bounding boxes to consider an
* object as a single line.
@@ -74,13 +74,13 @@ namespace scribo
float h_w_ratio);
- /// Fast Extraction of single vertical thick lines.
- /*!
+ /*! \brief Fast Extraction of single vertical thick lines.
+ *
* Only single non discontinued lines are correctly extracted
* with this routine.
*
* \param[in] components A labeled image.
- * \param[in] line_length The minimum line length.
+ * \param[in] min_line_length The minimum line length.
* \param[in] h_w_ratio The minimum ratio height/width object
* bounding boxes to consider an
* object as a single line.
diff --git a/scribo/scribo/primitive/extract/lines_v_thick_and_single.hh b/scribo/scribo/primitive/extract/lines_v_thick_and_single.hh
index 44f7a9e..c3719cf 100644
--- a/scribo/scribo/primitive/extract/lines_v_thick_and_single.hh
+++ b/scribo/scribo/primitive/extract/lines_v_thick_and_single.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Extract vertical thick lines in a binary image.
+/// \brief Extract vertical thick lines in a binary image.
# include <mln/core/concept/image.hh>
@@ -55,17 +55,17 @@ namespace scribo
using namespace mln;
- /// Extract vertical thick lines in a binary image.
- /*!
+ /*! \brief Extract vertical thick lines in a binary image.
+ *
* Only non discontinued lines are correctly extracted with this
* routine. Only lines matching the given criterions are kept
* in the result.
*
- * \param[in] input_ A binary image.
- * \param[in] nbh_ The neighborhood used for labeling image
+ * \param[in] input A binary image.
+ * \param[in] nbh The neighborhood used for labeling image
* components.
* \param[in,out] nlines Type used for labeling.
- * \param[in] line_length The minimum line length.
+ * \param[in] min_line_length The minimum line length.
* \param[in] h_w_ratio The minimum ratio height/width object
* bounding boxes to consider an
* object as a single line.
@@ -74,8 +74,8 @@ namespace scribo
*/
template <typename I, typename N, typename V>
component_set<mln_ch_value(I,V)>
- lines_v_thick_and_single(const Image<I>& input_,
- const Neighborhood<N>& nbh_,
+ lines_v_thick_and_single(const Image<I>& input,
+ const Neighborhood<N>& nbh,
V& nlines,
unsigned min_line_length,
float h_w_ratio);
diff --git a/scribo/scribo/primitive/group/from_double_link.hh b/scribo/scribo/primitive/group/from_double_link.hh
index 260bfea..a4d97de 100644
--- a/scribo/scribo/primitive/group/from_double_link.hh
+++ b/scribo/scribo/primitive/group/from_double_link.hh
@@ -62,7 +62,6 @@ namespace scribo
validate These links. A link must exist in both ways to be
validated.
- \param[in] components A component set.
\param[in] left_link The left neighbor of each line of text.
\param[in] right_link The right neighbor of each line of text.
diff --git a/scribo/scribo/primitive/identify.hh b/scribo/scribo/primitive/identify.hh
index 6aa6ad3..f4e115b 100644
--- a/scribo/scribo/primitive/identify.hh
+++ b/scribo/scribo/primitive/identify.hh
@@ -28,7 +28,7 @@
/*! \brief try to determine the type of a component.
-\fixme Write it and add support for component type (graphic, images,
+\todo Write it and add support for component type (graphic, images,
...)
*/
diff --git a/scribo/scribo/primitive/link/compute.hh b/scribo/scribo/primitive/link/compute.hh
index 114971f..8c64171 100644
--- a/scribo/scribo/primitive/link/compute.hh
+++ b/scribo/scribo/primitive/link/compute.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -50,6 +50,8 @@ namespace scribo
/*! \brief Compute links between objects according a given functor.
\param[in,out] functor Linking policy.
+ \param[in] anchor Lookup anchor. Starts looking for
+ neighbors from there.
\return Object links.
diff --git a/scribo/scribo/primitive/link/with_single_left_link.hh b/scribo/scribo/primitive/link/with_single_left_link.hh
index cc94f93..c4b080f 100644
--- a/scribo/scribo/primitive/link/with_single_left_link.hh
+++ b/scribo/scribo/primitive/link/with_single_left_link.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Link text components with their left neighbor.
+/// \brief Link text components with their left neighbor.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -65,7 +65,8 @@ namespace scribo
/// \brief Link components with their left neighbor if exists.
///
/// \param[in] components A component set.
- /// \param[in] The maximum distance allowed to seach a neighbor object.
+ /// \param[in] neighb_max_distance The maximum distance allowed
+ /// to seach a neighbor object.
///
/// \return Object links data.
//
@@ -78,6 +79,7 @@ namespace scribo
/// \overload
/// Max distance is set to mln_max(unsigned).
+ //
template <typename L>
inline
object_links<L>
diff --git a/scribo/scribo/primitive/link/with_single_left_link_dmax_ratio_aligned.hh b/scribo/scribo/primitive/link/with_single_left_link_dmax_ratio_aligned.hh
index 146ae17..9fdae67 100644
--- a/scribo/scribo/primitive/link/with_single_left_link_dmax_ratio_aligned.hh
+++ b/scribo/scribo/primitive/link/with_single_left_link_dmax_ratio_aligned.hh
@@ -29,8 +29,8 @@
/// \file
///
-/// Link text objects with their left neighbor according to a maximum
-/// distance.
+/// \brief Link text objects with their left neighbor according to a
+/// maximum distance.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -67,6 +67,10 @@ namespace scribo
\param[in] components A component set.
\param[in] dmax_ratio Size ratio defining the maximum lookup
distance.
+ \param[in] min_angle Minimum difference allowed for
+ alignement angle.
+ \param[in] max_angle Maximum difference allowed for
+ alignement angle.
\param[in] anchor Starting point for the neighbor lookup.
\return Object links data.
@@ -90,6 +94,7 @@ namespace scribo
/// \overload
/// anchor is set to MassCenter.
+ //
template <typename L>
inline
object_links<L>
@@ -102,6 +107,7 @@ namespace scribo
/// \overload
/// dmax_ratio is set to 3.
/// anchor is set to MassCenter.
+ //
template <typename L>
inline
object_links<L>
diff --git a/scribo/scribo/primitive/link/with_single_right_link.hh b/scribo/scribo/primitive/link/with_single_right_link.hh
index bc7942b..60bb4d1 100644
--- a/scribo/scribo/primitive/link/with_single_right_link.hh
+++ b/scribo/scribo/primitive/link/with_single_right_link.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Link components with their right neighbor.
+/// \brief Link components with their right neighbor.
# include <mln/core/concept/image.hh>
@@ -64,7 +64,10 @@ namespace scribo
/// Lookup startup point is the object mass center.
///
/// \param[in] components An object image.
- /// \param[in] The maximum distance allowed to seach a neighbor object.
+ /// \param[in] neighb_max_distance The maximum distance allowed
+ /// to seach a neighbor object.
+ /// \param[in] anchor Anchor from where the neighbor lookup is
+ /// performed.
///
/// \return Object links data.
//
@@ -78,6 +81,7 @@ namespace scribo
/// \overload
/// Max distance is set to mln_max(unsigned).
+ //
template <typename L>
inline
object_links<L>
diff --git a/scribo/scribo/primitive/link/with_single_right_link_bottom.hh b/scribo/scribo/primitive/link/with_single_right_link_bottom.hh
index 85f6b50..9b76919 100644
--- a/scribo/scribo/primitive/link/with_single_right_link_bottom.hh
+++ b/scribo/scribo/primitive/link/with_single_right_link_bottom.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Link text components with their right neighbor.
+/// \brief Link text components with their right neighbor.
# include <mln/core/concept/image.hh>
@@ -63,7 +63,8 @@ namespace scribo
/// Lookup startup point is the object bottom center.
///
/// \param[in] components A component set.
- /// \param[in] The maximum distance allowed to seach a neighbor object.
+ /// \param[in] neighb_max_distance The maximum distance allowed
+ /// to seach a neighbor object.
///
/// \return Object links data.
//
@@ -76,6 +77,7 @@ namespace scribo
/// \overload
/// Max distance is set to mln_max(unsigned).
+ //
template <typename L>
inline
object_links<L>
diff --git a/scribo/scribo/primitive/link/with_single_right_link_dmax_ratio_aligned.hh b/scribo/scribo/primitive/link/with_single_right_link_dmax_ratio_aligned.hh
index dcf804e..f7ed41e 100644
--- a/scribo/scribo/primitive/link/with_single_right_link_dmax_ratio_aligned.hh
+++ b/scribo/scribo/primitive/link/with_single_right_link_dmax_ratio_aligned.hh
@@ -29,8 +29,8 @@
/// \file
///
-/// Link text objects with their right neighbor according to a maximum
-/// distance.
+/// \brief Link text objects with their right neighbor according to a
+/// maximum distance.
# include <mln/core/concept/image.hh>
# include <mln/core/concept/neighborhood.hh>
@@ -67,6 +67,10 @@ namespace scribo
\param[in] components A component set.
\param[in] dmax_ratio Size ratio defining the maximum lookup
distance.
+ \param[in] min_angle Minimum difference allowed for
+ alignement angle.
+ \param[in] max_angle Maximum difference allowed for
+ alignement angle.
\param[in] anchor Starting point for the neighbor lookup.
\return Object links data.
diff --git a/scribo/scribo/primitive/link/with_single_right_link_top.hh b/scribo/scribo/primitive/link/with_single_right_link_top.hh
index ada6650..c8008fc 100644
--- a/scribo/scribo/primitive/link/with_single_right_link_top.hh
+++ b/scribo/scribo/primitive/link/with_single_right_link_top.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Link components with their right neighbor.
+/// \brief Link components with their right neighbor.
# include <mln/core/concept/image.hh>
@@ -64,7 +64,8 @@ namespace scribo
/// Lookup startup point is the object top center.
///
/// \param[in] components A component set.
- /// \param[in] The maximum distance allowed to seach a neighbor object.
+ /// \param[in] neighb_max_distance The maximum distance allowed
+ /// to seach a neighbor object.
///
/// \return Object links data.
//
@@ -77,6 +78,7 @@ namespace scribo
/// \overload
/// Max distance is set to mln_max(unsigned).
+ //
template <typename L>
inline
object_links<L>
diff --git a/scribo/scribo/table/align_lines_horizontaly.hh b/scribo/scribo/table/align_lines_horizontaly.hh
index 25e4036..38d14ec 100644
--- a/scribo/scribo/table/align_lines_horizontaly.hh
+++ b/scribo/scribo/table/align_lines_horizontaly.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -55,13 +55,16 @@ namespace scribo
///
/// \param[in] input Image from which the line bboxes are
/// extracted from.
- /// \param[in, out] line_bboxes horizontal lines bounding boxes.
+ /// \param[in] lines Component set corresponding to table lines.
+ /// \param[in,out] aligned_lines Component set where aligned table
+ /// lines are stored.
/// \param[in] max_alignment_diff max space between two lines to
/// consider they are potentialy on the
/// same line.
///
/// \return A list of the resulting aligned rows. Each integer is actually
/// a row number.
+ //
template <typename I, typename L>
mln::util::array<int>
align_lines_horizontaly(const Image<I>& input,
diff --git a/scribo/scribo/table/rebuild.hh b/scribo/scribo/table/rebuild.hh
index f593153..e433390 100644
--- a/scribo/scribo/table/rebuild.hh
+++ b/scribo/scribo/table/rebuild.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -29,7 +29,7 @@
/// \file
///
-/// Rebuild a table from its line bounding boxes.
+/// \brief Rebuild a table from its line bounding boxes.
# include <mln/core/concept/image.hh>
# include <mln/core/alias/neighb2d.hh>
@@ -57,11 +57,13 @@ namespace scribo
{
- /// Rebuild a table from its line bounding boxes.
- /*!
- ** \param[in] input_ A binary image.
- ** \param[in] linebboxes_ A couple of vertical and horizontal
- ** line bounding boxes.
+ /*! \brief Rebuild a table from its line bounding boxes.
+ **
+ ** \param[in] input A binary image.
+ ** \param[in] vlines Component set corresponding to vertical
+ ** lines.
+ ** \param[in] hlines Component set corresponding to horizontal
+ ** lines.
** \param[in] max_dist_lines The maximum distance allowed between
** vertical and horizontal lines to connect
** them eachother.
@@ -76,7 +78,7 @@ namespace scribo
mln::util::couple<L,
mln::util::couple<component_set<L>,
component_set<L> > >
- rebuild(const Image<I>& input_,
+ rebuild(const Image<I>& input,
const component_set<L>& vlines,
const component_set<L>& hlines,
unsigned max_dist_lines,
diff --git a/scribo/scribo/text/extract_lines.hh b/scribo/scribo/text/extract_lines.hh
index 3e9254a..685079c 100644
--- a/scribo/scribo/text/extract_lines.hh
+++ b/scribo/scribo/text/extract_lines.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Extract line of text bounding boxes.
+/// \brief Extract line of text bounding boxes.
# include <mln/core/concept/image.hh>
@@ -62,17 +62,17 @@ namespace scribo
using namespace mln;
- /// Extract lines of text in a binary image.
- /*!
+ /*! \brief Extract lines of text in a binary image.
+ **
** \param[in] input A binary image.
** \param[in] nbh A neighborhood used for labeling.
- ** \param[in] seps A binary image with separator information.
+ ** \param[in] separators A binary image with separator information.
**
** \return A set of lines.
*/
template <typename I, typename N>
line_set<mln_ch_value(I,scribo::def::lbl_type)>
- extract_lines(const Image<I>& input_, const Neighborhood<N>& nbh_,
+ extract_lines(const Image<I>& input, const Neighborhood<N>& nbh,
const mln_ch_value(I,bool)& separators);
/// \overload
diff --git a/scribo/scribo/text/extract_lines_wo_merge.hh b/scribo/scribo/text/extract_lines_wo_merge.hh
index 0d90600..ea9e6cf 100644
--- a/scribo/scribo/text/extract_lines_wo_merge.hh
+++ b/scribo/scribo/text/extract_lines_wo_merge.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Extract line of text bounding boxes.
+/// \brief Extract line of text bounding boxes.
# include <mln/core/concept/image.hh>
@@ -62,17 +62,17 @@ namespace scribo
using namespace mln;
- /// Extract lines of text in a binary image.
+ /// \brief Extract lines of text in a binary image.
/*!
** \param[in] input A binary image.
** \param[in] nbh A neighborhood used for labeling.
- ** \param[in] seps A binary image with separator information.
+ ** \param[in] separators A binary image with separator information.
**
** \return A set of lines.
*/
template <typename I, typename N>
line_set<mln_ch_value(I,scribo::def::lbl_type)>
- extract_lines_wo_merge(const Image<I>& input_, const Neighborhood<N>& nbh_,
+ extract_lines_wo_merge(const Image<I>& input, const Neighborhood<N>& nbh,
const mln_ch_value(I,bool)& separators);
/// \overload
diff --git a/scribo/scribo/text/recognition.hh b/scribo/scribo/text/recognition.hh
index 2b9d5a1..7bdcb58 100644
--- a/scribo/scribo/text/recognition.hh
+++ b/scribo/scribo/text/recognition.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Passes the text bounding boxes to an OCR (Tesseract).
+/// \brief Recognize text using an OCR (Tesseract).
///
/// \todo For each text bbox, we create a new image. We may like to avoid that.
/// \todo Do not store the result in an image?
@@ -81,20 +81,25 @@ namespace scribo
using namespace mln;
- /// Passes the text bboxes to Tesseract (OCR).
+ /// \brief Passes the text bboxes to Tesseract (OCR).
///
/// \param[in] lines The lines of text.
/// \param[in] language The language which should be recognized by
/// Tesseract. (fra, en, ...)
- /// \param[in] output_file If set, store the recognized text in
- /// this file.
//
template <typename L>
void
recognition(line_set<L>& lines, const char *language);
- /// Recognize text from an image.
+ /// \brief Recognize text from an image.
+ ///
+ /// \param[in] line Image of text line.
+ /// \param[in] language The language which should be recognized by
+ /// Tesseract. (fra, en, ...)
+ /// \param[in] output_file If set, store the recognized text in
+ /// this file.
+ //
template <typename I>
void
recognition(const Image<I>& line,
diff --git a/scribo/scribo/toolchain/content_in_doc.hh b/scribo/scribo/toolchain/content_in_doc.hh
index 0d3430e..5ef6aa8 100644
--- a/scribo/scribo/toolchain/content_in_doc.hh
+++ b/scribo/scribo/toolchain/content_in_doc.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Analyse a document.
+/// \brief Analyse and extract document image content.
# include <scribo/toolchain/internal/content_in_doc_functor.hh>
@@ -41,6 +41,27 @@ namespace scribo
using namespace mln;
+ /*! \brief Analyse and extract document image content.
+
+ \input[in] input A RGB image.
+ \input[in] input_preproc A Binary image.
+ \param[in] denoise Remove too small components (<= 2 pixels).
+ \param[in] find_line_seps Enable/Disable lookup for
+ vertical/horizontal separators.
+ \param[in] find_whitespace_seps Enable/Disable lookup for
+ vertical alignments and
+ whitespaces. Helps finding text
+ layout.
+ \param[in] enable_ocr Enable/Disable text recognition.
+ \param[in] language Main language used in the input
+ document. (OCR settings)
+ \param[in] verbose Enable/Disable debug information printed on
+ std::cout.
+
+ \return A document structure.
+
+ \ingroup grptoolchaindocproc
+ */
template <typename I, typename J>
document<mln_ch_value(I, def::lbl_type)>
content_in_doc(const Image<I>& input, const Image<J>& input_preproc,
diff --git a/scribo/scribo/toolchain/content_in_hdoc.hh b/scribo/scribo/toolchain/content_in_hdoc.hh
index 97233d5..1ffff09 100644
--- a/scribo/scribo/toolchain/content_in_hdoc.hh
+++ b/scribo/scribo/toolchain/content_in_hdoc.hh
@@ -28,7 +28,7 @@
/// \file
///
-/// Analyse a document.
+/// \brief Analyse a historical/degraded document.
# include <scribo/toolchain/internal/content_in_hdoc_functor.hh>
@@ -40,7 +40,26 @@ namespace scribo
using namespace mln;
-
+ /*! \brief Analyse and extract content in a historical/degraded
+ document.
+
+ \input[in] input A RGB image.
+ \input[in] input_preproc A Binary image.
+ \param[in] denoise Remove too small components (<= 2 pixels).
+ \param[in] find_line_seps Enable/Disable lookup for
+ vertical/horizontal separators.
+ \param[in] find_whitespace_seps Enable/Disable lookup for
+ vertical alignments and
+ whitespaces. Helps finding text
+ layout.
+ \param[in] enable_ocr Enable/Disable text recognition.
+ \param[in] language Main language used in the input
+ document. (OCR settings)
+
+ \return A document structure.
+
+ \ingroup grptoolchaindocproc
+ */
template <typename I, typename J>
document<mln_ch_value(I, def::lbl_type)>
content_in_hdoc(const Image<I>& input, const Image<J>& input_preproc,
diff --git a/scribo/scribo/toolchain/internal/content_in_doc_functor.hh b/scribo/scribo/toolchain/internal/content_in_doc_functor.hh
index 448ec24..066168a 100644
--- a/scribo/scribo/toolchain/internal/content_in_doc_functor.hh
+++ b/scribo/scribo/toolchain/internal/content_in_doc_functor.hh
@@ -27,6 +27,10 @@
#ifndef SCRIBO_TOOLCHAIN_INTERNAL_CONTENT_IN_DOC_FUNCTOR_HH
# define SCRIBO_TOOLCHAIN_INTERNAL_CONTENT_IN_DOC_FUNCTOR_HH
+/// \file
+///
+/// \brief Functor analysing and extracting document image content.
+
# ifndef SCRIBO_NDEBUG
# include <mln/util/timer.hh>
# endif // ! SCRIBO_NDEBUG
@@ -92,7 +96,9 @@ namespace scribo
namespace internal
{
+ /*! \brief Functor analysing and extracting document image content.
+ */
template <typename I>
struct content_in_doc_functor
: public Toolchain_Functor
diff --git a/scribo/scribo/toolchain/internal/content_in_hdoc_functor.hh b/scribo/scribo/toolchain/internal/content_in_hdoc_functor.hh
index bda31b3..4ffc9a7 100644
--- a/scribo/scribo/toolchain/internal/content_in_hdoc_functor.hh
+++ b/scribo/scribo/toolchain/internal/content_in_hdoc_functor.hh
@@ -27,6 +27,12 @@
#ifndef SCRIBO_TOOLCHAIN_INTERNAL_CONTENT_IN_HDOC_FUNCTOR_HH
# define SCRIBO_TOOLCHAIN_INTERNAL_CONTENT_IN_HDOC_FUNCTOR_HH
+/// \file
+///
+/// \brief Functor analysing and extracting content in
+/// degraded/historical documents.
+
+
# ifndef SCRIBO_NDEBUG
# include <mln/util/timer.hh>
# endif // ! SCRIBO_NDEBUG
@@ -103,6 +109,9 @@ namespace scribo
{
+ /*! \brief Functor analysing and extracting content in
+ degraded/historical documents.
+ */
template <typename I>
struct content_in_hdoc_functor
: public Toolchain_Functor
diff --git a/scribo/scribo/toolchain/internal/text_in_doc_functor.hh b/scribo/scribo/toolchain/internal/text_in_doc_functor.hh
index edb170d..eeac1ea 100644
--- a/scribo/scribo/toolchain/internal/text_in_doc_functor.hh
+++ b/scribo/scribo/toolchain/internal/text_in_doc_functor.hh
@@ -27,6 +27,10 @@
#ifndef SCRIBO_TOOLCHAIN_INTERNAL_TEXT_IN_DOC_FUNCTOR_HH
# define SCRIBO_TOOLCHAIN_INTERNAL_TEXT_IN_DOC_FUNCTOR_HH
+/// \file
+///
+/// \brief Functor extracting text lines from a document image.
+
# ifndef SCRIBO_NDEBUG
# include <mln/util/timer.hh>
# endif // ! SCRIBO_NDEBUG
@@ -76,7 +80,9 @@ namespace scribo
namespace internal
{
+ /*! \brief Functor extracting text lines from a document image.
+ */
template <typename I>
struct text_in_doc_functor
: public Toolchain_Functor
diff --git a/scribo/scribo/toolchain/internal/text_in_doc_preprocess_functor.hh b/scribo/scribo/toolchain/internal/text_in_doc_preprocess_functor.hh
index d58b4b2..12dc834 100644
--- a/scribo/scribo/toolchain/internal/text_in_doc_preprocess_functor.hh
+++ b/scribo/scribo/toolchain/internal/text_in_doc_preprocess_functor.hh
@@ -27,6 +27,11 @@
#ifndef SCRIBO_TOOLCHAIN_INTERNAL_TEXT_IN_DOC_PREPROCESS_FUNCTOR_HH
# define SCRIBO_TOOLCHAIN_INTERNAL_TEXT_IN_DOC_PREPROCESS_FUNCTOR_HH
+/// \file
+///
+/// \brief Functor performing custom preprocessing algorithms on
+/// documents.
+
#include <mln/core/concept/image.hh>
#include <mln/data/transform.hh>
#include <mln/data/convert.hh>
diff --git a/scribo/scribo/toolchain/internal/text_in_picture_functor.hh b/scribo/scribo/toolchain/internal/text_in_picture_functor.hh
index 5178a82..8d12612 100644
--- a/scribo/scribo/toolchain/internal/text_in_picture_functor.hh
+++ b/scribo/scribo/toolchain/internal/text_in_picture_functor.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Localize text in a picture.
+/// \brief Localize text in a picture.
# include <libgen.h>
@@ -107,7 +107,7 @@ namespace scribo
using namespace mln;
-
+ /// \brief Localize text in a picture.
template <typename I>
struct text_in_picture_functor
: public Toolchain_Functor
diff --git a/scribo/scribo/toolchain/internal/toolchain_functor.hh b/scribo/scribo/toolchain/internal/toolchain_functor.hh
index fbf1702..835cbbd 100644
--- a/scribo/scribo/toolchain/internal/toolchain_functor.hh
+++ b/scribo/scribo/toolchain/internal/toolchain_functor.hh
@@ -27,6 +27,10 @@
#ifndef SCRIBO_TOOLCHAIN_INTERNAL_TOOLCHAIN_FUNCTOR_HH
# define SCRIBO_TOOLCHAIN_INTERNAL_TOOLCHAIN_FUNCTOR_HH
+/// \file
+///
+/// \brief Base class for toolchain functors.
+
# include <unistd.h>
# include <iostream>
@@ -39,6 +43,7 @@ namespace scribo
namespace internal
{
+ /// \brief Base class for toolchain functors.
class Toolchain_Functor
{
public:
diff --git a/scribo/scribo/toolchain/nepomuk/text_extraction.hh b/scribo/scribo/toolchain/nepomuk/text_extraction.hh
index 2534ce8..705787a 100644
--- a/scribo/scribo/toolchain/nepomuk/text_extraction.hh
+++ b/scribo/scribo/toolchain/nepomuk/text_extraction.hh
@@ -65,7 +65,10 @@ namespace scribo
This is a convenient routine to be used in Nepomuk.
- \param[in] ima A document image. The
+ \param[in] input A document image.
+ \param[in] language The main language used in the input
+ document image. Improve text recognition
+ quality if accurate.
\return A set of recognized words.
@@ -76,6 +79,7 @@ namespace scribo
Depending on your version of Tesseract (OCR) you may define
HAVE_TESSERACT_2 or HAVE_TESSERACT_3 .
+ \ingroup grptoolchain
*/
QSet<QString>
text_extraction(const QImage& input, const QString& language);
diff --git a/scribo/scribo/toolchain/text_in_doc.hh b/scribo/scribo/toolchain/text_in_doc.hh
index 960a22a..b1fb511 100644
--- a/scribo/scribo/toolchain/text_in_doc.hh
+++ b/scribo/scribo/toolchain/text_in_doc.hh
@@ -29,7 +29,7 @@
/// \file
///
-/// Extract text from a document.
+/// \brief Extract text from a document.
# include <scribo/toolchain/internal/text_in_doc_functor.hh>
@@ -42,6 +42,26 @@ namespace scribo
using namespace mln;
+ /*! \brief Extract text lines from a document image.
+
+ \param[in] input A RGB image.
+ \param[in] denoise Remove too small components (<= 2 pixels).
+ \param[in] language Main language used in the input
+ document. (OCR settings)
+ \param[in] find_line_seps Enable/Disable lookup for
+ vertical/horizontal separators.
+ \param[in] find_whitespace_seps Enable/Disable lookup for
+ vertical alignments and
+ whitespaces. Helps finding text
+ layout.
+ \param[in] verbose Enable/Disable debug information printed on
+ std::cout.
+
+ \return A line set including text information and recognized
+ text.
+
+ \ingroup grptoolchaindocproc
+ */
template <typename I>
line_set<mln_ch_value(I, def::lbl_type)>
text_in_doc(const Image<I>& input, bool denoise,
diff --git a/scribo/scribo/toolchain/text_in_doc_preprocess.hh b/scribo/scribo/toolchain/text_in_doc_preprocess.hh
index 0a99487..e94af6f 100644
--- a/scribo/scribo/toolchain/text_in_doc_preprocess.hh
+++ b/scribo/scribo/toolchain/text_in_doc_preprocess.hh
@@ -55,12 +55,16 @@ namespace scribo
\param[in] input An image.
\param[in] enable_fg_bg Enable/Disable background removal.
+ \param[in] lambda Parameter to the background removal. Maximum
+ area of foreground objects.
\param[in] K Binarization threshold parameter. (Default 0.34)
+ \param[in] verbose Enable/Disable debug output on std::cout.
If \p enable_fg_bg is set to 'True' then a background removal is
performed. Its parameter lambda is automatically set according
to the input image size.
+ \ingroup grptoolchaindocpreproc
*/
template <typename I>
mln_ch_value(I,bool)
@@ -70,6 +74,8 @@ namespace scribo
/*! \overload
K is set to 0.34.
+
+ \ingroup grptoolchaindocpreproc
*/
template <typename I>
mln_ch_value(I,bool)
@@ -81,11 +87,14 @@ namespace scribo
\param[in] input An image.
\param[in] lambda Parameter to the background removal.
+ \param[in] verbose Enable/Disable debug information printed on
+ std::cout.
If lambda is set to '0' no background removal is
performed. Otherwise, a background removal is performed with the
given \p lambda value.
+ \ingroup grptoolchaindocpreproc
*/
template <typename I>
mln_ch_value(I,bool)
@@ -97,12 +106,17 @@ namespace scribo
\param[in] input An image.
\param[in] lambda Parameter to the background removal.
\param[in] K Binarization threshold parameter. (Default 0.34)
+ \param[in] enable_fg_bg If set to True tries to identify
+ background components and remove them.
\param[in,out] fg The foreground layer of \p input.
+ \param[in] verbose Enable/Disable debug information printed on
+ std::cout.
If lambda is set to '0' no background removal is
performed. Otherwise, a background removal is performed with the
given \p lambda value.
+ \ingroup grptoolchaindocpreproc
*/
template <typename I>
mln_ch_value(I,bool)
diff --git a/scribo/scribo/toolchain/text_in_picture.hh b/scribo/scribo/toolchain/text_in_picture.hh
index 8d05158..0580ec6 100644
--- a/scribo/scribo/toolchain/text_in_picture.hh
+++ b/scribo/scribo/toolchain/text_in_picture.hh
@@ -30,7 +30,7 @@
/// \file
///
-/// Localize text in a picture.
+/// \brief Localize text in a picture.
# include <scribo/toolchain/internal/text_in_picture_functor.hh>
@@ -42,7 +42,26 @@ namespace scribo
using namespace mln;
-
+ /*! \brief Localize text in pictures.
+
+ \param[in] input_rgb_orig A RGB image.
+ \param[in] bg_removal If set to True tries to identify
+ background components and remove them.
+ \param[in] multi_scale_bin Sauvola's binarization is used. Tells
+ whether to use a multi-scale version
+ or not.
+ \param[in] max_dim_size The maximum height or width allowed for
+ the input image. If larger, it is
+ resized.
+ \param[in] lambda Maximum area of components considered as
+ foreground objects. If set to '0' (default),
+ it is fixed automatically.
+ \param[in] verbose Enable/Disable debug output on std::cout.
+
+ \return A component set corresponding to the text components.
+
+ \ingroup grptoolchainpicproc
+ */
template <typename I>
component_set<mln_ch_value(I, def::lbl_type)>
text_in_picture(const Image<I>& input_rgb_orig,
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-66-gfa49a34 Add new groups in Scribo documentation.
by Guillaume Lazzara
* doc/Doxyfile.in: Refine settings.
* doc/groups.dox: Add new groups.
* scribo/core/component_set.hh,
* scribo/core/document.hh,
* scribo/core/line_links.hh,
* scribo/core/line_set.hh,
* scribo/core/object_groups.hh,
* scribo/core/object_links.hh,
* scribo/core/paragraph_set.hh: Set to a documentation group.
---
ChangeLog | 16 ++++++++++++++++
scribo/doc/Doxyfile.in | 13 ++++++++-----
scribo/doc/groups.dox | 19 +++++++++++++++++++
scribo/scribo/core/component_set.hh | 4 ++++
scribo/scribo/core/document.hh | 7 +++++--
scribo/scribo/core/line_links.hh | 4 +++-
scribo/scribo/core/line_set.hh | 7 +++++--
scribo/scribo/core/object_groups.hh | 2 ++
scribo/scribo/core/object_links.hh | 4 +++-
scribo/scribo/core/paragraph_set.hh | 1 +
10 files changed, 66 insertions(+), 11 deletions(-)
create mode 100644 scribo/doc/groups.dox
diff --git a/ChangeLog b/ChangeLog
index bb6ddc8..d796c0b 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,21 @@
2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
+ Add new groups in Scribo documentation.
+
+ * doc/Doxyfile.in: Refine settings.
+
+ * doc/groups.dox: Add new groups.
+
+ * scribo/core/component_set.hh,
+ * scribo/core/document.hh,
+ * scribo/core/line_links.hh,
+ * scribo/core/line_set.hh,
+ * scribo/core/object_groups.hh,
+ * scribo/core/object_links.hh,
+ * scribo/core/paragraph_set.hh: Set to a documentation group.
+
+2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
+
Improve common documentation.
* AUTHORS: Use RST syntax.
diff --git a/scribo/doc/Doxyfile.in b/scribo/doc/Doxyfile.in
index d279d76..7ad639f 100644
--- a/scribo/doc/Doxyfile.in
+++ b/scribo/doc/Doxyfile.in
@@ -26,7 +26,7 @@ PROJECT_LOGO = @top_srcdir@/doc/logo.jpg
OUTPUT_DIRECTORY = @builddir@/user-refman.tmp
CREATE_SUBDIRS = YES
OUTPUT_LANGUAGE = English
-BRIEF_MEMBER_DESC = YES
+BRIEF_MEMBER_DESC = NO
REPEAT_BRIEF = YES
ABBREVIATE_BRIEF =
ALWAYS_DETAILED_SEC = YES
@@ -47,6 +47,8 @@ BUILTIN_STL_SUPPORT = YES
IDL_PROPERTY_SUPPORT = NO
DISTRIBUTE_GROUP_DOC = NO
SUBGROUPING = YES
+INLINE_GROUPED_CLASSES = NO
+SORT_GROUP_NAMES = YES
#---------------------------------------------------------------------------
# Build related configuration options
#---------------------------------------------------------------------------
@@ -95,7 +97,8 @@ INPUT = @top_srcdir@/scribo
FILE_PATTERNS = *.cc \
*.hh \
*.hxx \
- *.hcc
+ *.hcc \
+ *.dox
RECURSIVE = YES
EXCLUDE = @top_srcdir@/scribo/demo \
@top_srcdir@/scribo/sandbox \
@@ -126,8 +129,8 @@ INPUT_ENCODING = ISO-8859-1
SOURCE_BROWSER = YES
INLINE_SOURCES = NO
STRIP_CODE_COMMENTS = YES
-REFERENCED_BY_RELATION = YES
-REFERENCES_RELATION = YES
+REFERENCED_BY_RELATION = NO
+REFERENCES_RELATION = NO
REFERENCES_LINK_SOURCE = NO
USE_HTAGS = NO
VERBATIM_HEADERS = YES
@@ -329,7 +332,7 @@ SKIP_FUNCTION_MACROS = YES
TAGFILES = @builddir@/../../milena/doc/milena.tag=../../../../milena/doc/user-refman/html
GENERATE_TAGFILE = scribo.tag
ALLEXTERNALS = NO
-EXTERNAL_GROUPS = YES
+EXTERNAL_GROUPS = NO
PERL_PATH = /usr/bin/perl
#---------------------------------------------------------------------------
# Configuration options related to the dot tool
diff --git a/scribo/doc/groups.dox b/scribo/doc/groups.dox
new file mode 100644
index 0000000..73d842e
--- /dev/null
+++ b/scribo/doc/groups.dox
@@ -0,0 +1,19 @@
+/*! \defgroup grpstruct Data structures
+ *
+ * \brief Scribo high level data structures.
+ *
+ */
+
+
+/*! \defgroup grproutine Routines
+ *
+ * \brief All routines/algorithms provided in Scribo.
+ *
+ */
+
+
+/*! \defgroup grpalgobin Binarization
+ *
+ * \brief Binarization algorithms.
+ *
+ */
diff --git a/scribo/scribo/core/component_set.hh b/scribo/scribo/core/component_set.hh
index ded64ae..8c3e83f 100644
--- a/scribo/scribo/core/component_set.hh
+++ b/scribo/scribo/core/component_set.hh
@@ -119,6 +119,10 @@ namespace scribo
} // end of namespace scribo::internal
+ /// \brief Represents all the components in a document image.
+ ///
+ /// \ingroup grpstruct
+ //
template <typename L>
class component_set : public Serializable<component_set<L> >
{
diff --git a/scribo/scribo/core/document.hh b/scribo/scribo/core/document.hh
index cc885e4..5bbd226 100644
--- a/scribo/scribo/core/document.hh
+++ b/scribo/scribo/core/document.hh
@@ -79,9 +79,12 @@ namespace scribo
} // end of namespace scribo::internal
-
+ /// \brief Represent document data and structure.
+ ///
+ /// \ingroup grpstruct
+ //
template <typename L>
- struct document : public Serializable<document<L> >
+ class document : public Serializable<document<L> >
{
typedef internal::document_data<L> data_t;
diff --git a/scribo/scribo/core/line_links.hh b/scribo/scribo/core/line_links.hh
index b7b438c..4f807e4 100644
--- a/scribo/scribo/core/line_links.hh
+++ b/scribo/scribo/core/line_links.hh
@@ -67,7 +67,9 @@ namespace scribo
- /// \brief Line group representation.
+ /// \brief Line links representation.
+ ///
+ /// \ingroup grpstruct
//
template <typename L>
class line_links : public Serializable<line_links<L> >
diff --git a/scribo/scribo/core/line_set.hh b/scribo/scribo/core/line_set.hh
index db987e6..4880a77 100644
--- a/scribo/scribo/core/line_set.hh
+++ b/scribo/scribo/core/line_set.hh
@@ -1,5 +1,5 @@
-// Copyright (C) 2009, 2010 EPITA Research and Development Laboratory
-// (LRDE)
+// Copyright (C) 2009, 2010, 2011 EPITA Research and Development
+// Laboratory (LRDE)
//
// This file is part of Olena.
//
@@ -86,6 +86,9 @@ namespace scribo
Line ids start from 1.
+
+ \ingroup grpstruct
+
*/
template <typename L>
class line_set
diff --git a/scribo/scribo/core/object_groups.hh b/scribo/scribo/core/object_groups.hh
index 46447d2..53ba7f9 100644
--- a/scribo/scribo/core/object_groups.hh
+++ b/scribo/scribo/core/object_groups.hh
@@ -88,6 +88,8 @@ namespace scribo
/// \brief Object group representation.
+ ///
+ /// \ingroup grpstruct
//
template <typename L>
class object_groups : public Serializable<object_groups<L> >
diff --git a/scribo/scribo/core/object_links.hh b/scribo/scribo/core/object_links.hh
index 37fd299..a2543a7 100644
--- a/scribo/scribo/core/object_links.hh
+++ b/scribo/scribo/core/object_links.hh
@@ -69,7 +69,9 @@ namespace scribo
- /// \brief Object group representation.
+ /// \brief Object links representation.
+ ///
+ /// \ingroup grpstruct
//
template <typename L>
class object_links : public Serializable<object_links<L> >
diff --git a/scribo/scribo/core/paragraph_set.hh b/scribo/scribo/core/paragraph_set.hh
index c21359d..69194be 100644
--- a/scribo/scribo/core/paragraph_set.hh
+++ b/scribo/scribo/core/paragraph_set.hh
@@ -65,6 +65,7 @@ namespace scribo
Paragraph ids start from 1.
+ \ingroup grpstruct
*/
template <typename L>
class paragraph_set : public Serializable<paragraph_set<L> >
--
1.7.2.5
12 years, 2 months
- 1
- 0
olena-2.0-65-g16fdee1 Improve common documentation.
by Guillaume Lazzara
* AUTHORS: Use RST syntax.
* doc/Doxyfile.in: Use .dox files for documentation.
* doc/DoxygenLayout.xml: Improve layout.
* doc/Makefile.am: Add dependencies.
* doc/lrde.bib,
* doc/contributors.dox: New.
* doc/bibliography.hh,
* doc/mainpage.hh,
* doc/modules_list.hh: Rename to...
* doc/bibliography.dox,
* doc/mainpage.dox,
* doc/modules_list.dox: ...this.
* doc/olena.qhcp: Remove.
---
AUTHORS | 201 ++-
ChangeLog | 25 +
doc/Doxyfile.in | 10 +-
doc/DoxygenLayout.xml | 2 +-
doc/Makefile.am | 46 +-
doc/bibliography.dox | 37 +
doc/bibliography.hh | 19 -
doc/contributors.dox | 5 +
doc/lrde.bib | 3856 +++++++++++++++++++++++++++++
doc/{mainpage.hh => mainpage.dox} | 6 +-
doc/{modules_list.hh => modules_list.dox} | 0
doc/olena.qhcp | 12 -
12 files changed, 4116 insertions(+), 103 deletions(-)
create mode 100644 doc/bibliography.dox
delete mode 100644 doc/bibliography.hh
create mode 100644 doc/contributors.dox
create mode 100644 doc/lrde.bib
rename doc/{mainpage.hh => mainpage.dox} (80%)
rename doc/{modules_list.hh => modules_list.dox} (100%)
delete mode 100644 doc/olena.qhcp
diff --git a/AUTHORS b/AUTHORS
index f12b60e..9bea227 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -5,7 +5,8 @@ This package was written by and with the assistance of
Active LRDE Staff.
==================
-* Thierry Géraud theo(a)lrde.epita.fr
+* **Thierry Géraud** <theo(a)lrde.epita.fr>
+
- Management of the project in the first place.
- Type system.
- Revamp of Metalic (Olena pre-1.0).
@@ -14,11 +15,15 @@ Active LRDE Staff.
- Olena 1.0 core.
- Numerous hours spent thinking about Olena.
-* Guillaume Lazzara z(a)lrde.epita.fr
+
+* **Guillaume Lazzara** <z(a)lrde.epita.fr>
+
- SCRIBO module main author and maintainer.
- Library improvements.
-* Roland Levillain roland(a)lrde.epita.fr
+
+* **Roland Levillain** <roland(a)lrde.epita.fr>
+
- Project manager.
- Theoretical mentor.
- Renovation of Swilena (Olena 0.10a).
@@ -32,30 +37,41 @@ Active LRDE Staff.
Past contributors of LRDE Staff.
================================
-* Jérôme Darbon jerome(a)lrde.epita.fr
+* **Jérôme Darbon** <jerome(a)lrde.epita.fr>
+
- Image morphology and Olena morpho.
-* Réda Dehak reda(a)lrde.epita.fr
+
+* **Réda Dehak** <reda(a)lrde.epita.fr>
+
- Management of the project.
- Color conversions.
- Cleanup the sources.
-* Akim Demaille akim(a)lrde.epita.fr
+
+* **Akim Demaille** <akim(a)lrde.epita.fr>
+
- Configuration system.
- Maintenance of code's clearness.
-* Alexandre Duret-Lutz adl(a)lrde.epita.fr
+
+* **Alexandre Duret-Lutz** <adl(a)lrde.epita.fr>
+
- Maintenance of the source tree for several years.
- Type system.
- Test system.
- First documentation system.
- Numerous hours spent on Olena to make it better.
-* Fabien Freling fabien.freling(a)lrde.epita.fr
+
+* **Fabien Freling** <fabien.freling(a)lrde.epita.fr>
+
- Work on Melimage with IGR.
- Library improvements.
-* Yann Jacquelet jacquelet(a)lrde.epita.fr
+
+* **Yann Jacquelet** <jacquelet(a)lrde.epita.fr>
+
- SCRIBO-related work.
@@ -63,20 +79,25 @@ Past contributors of LRDE Staff.
Interns.
========
-* Arthur Crépin-Leblond
+* **Arthur Crépin-Leblond**
+
- SCRIBO structured outputs (XML, HTML, PDF).
-* Raphaël Boissel
+
+* **Raphaël Boissel**
- Graph implementation of a DIA toolchain in SCRIBO.
=========================
Students from Class 2012.
=========================
-* Frédéric Bour frederic.bour(a)lrde.epita.fr
+* **Frédéric Bour** <frederic.bour(a)lrde.epita.fr>
+
- Prototype of the upcoming functions system.
-* Julien Marquegnies julien.marquegnies(a)lrde.epita.fr
+
+* **Julien Marquegnies** <julien.marquegnies(a)lrde.epita.fr>
+
- Fast implementation of the Hough transform.
- Many improvements in SCRIBO layout analysis.
@@ -84,7 +105,8 @@ Students from Class 2012.
Students from Class 2011.
=========================
-* Edwin Carlinet edwin.carlinet(a)lrde.epita.fr
+* **Edwin Carlinet** <edwin.carlinet(a)lrde.epita.fr>
+
- Accumulators.
- Morphological connected filters.
@@ -92,32 +114,44 @@ Students from Class 2011.
Students from Class 2010.
=========================
-* Florent D'Halluin
+* **Florent D'Halluin**
+
- Document layout viewer.
-* Etienne Folio
+
+* **Etienne Folio**
+
- Color type traits.
=========================
Students from Class 2009.
=========================
-* Alexandre Abraham
+* **Alexandre Abraham**
+
- R/W functions.
- Topological watershed transform.
-* Nicolas Ballas
+
+* **Nicolas Ballas**
+
- Image properties.
- Data routines.
-* Vivien Delmon
+
+* **Vivien Delmon**
+
- Misc. corrections.
-* Matthieu Garrigues
+
+* **Matthieu Garrigues**
+
- Value types and I/Os.
- Draft of the FLLT.
-* Ugo Jardonnet
+
+* **Ugo Jardonnet**
+
- Geometrical transforms.
- ICP registration algorithm.
@@ -125,24 +159,32 @@ Students from Class 2009.
Students from Class 2008.
=========================
-* Tristan Croiset
+* **Tristan Croiset**
+
- Dynamic (Dyn) (Olena proto-1.0)
-* Quentin Hocquet
+
+* **Quentin Hocquet**
+
- SCOOL
-* Thomas Moulard
+
+* **Thomas Moulard**
+
- SCOOL
=========================
Students from Class 2007.
=========================
-* Christophe Berger
+* **Christophe Berger**
+
- Connected filters (Olena proto-1.0, Olena pre-1.0).
- SCOOP 2 (Olena proto-1.0).
-* Nicolas Widynski
+
+* **Nicolas Widynski**
+
- Connected filters (Olena proto-1.0).
- Inpainting (Olena proto-1.0).
- Segmentation (watershed transform, watersnake, snakes) (Olena proto-1.0)
@@ -151,16 +193,21 @@ Students from Class 2007.
Students from Class 2006.
=========================
-* Simon Odou
+* **Simon Odou**
+
- Contributions to document the code.
- Deconvolution.
- Generic morphers.
- SCOOP 2 (Olena proto-1.0).
-* Nicolas Pouillard
+
+* **Nicolas Pouillard**
+
- Dynamic-static bridge prototype (Dyn) (Olena proto-1.0).
-* Damien Thivolle
+
+* **Damien Thivolle**
+
- Contributions to document the code.
- Generic morphers.
- Configuration system.
@@ -174,14 +221,17 @@ Students from Class 2006.
Students from Class 2005.
=========================
-* Giovanni Palma
+* **Giovanni Palma**
+
- Color conversion.
- Attribute opening and closing algorithms.
- New documentation system.
- Document the code.
- Cleanup the sources.
-* Niels Van Vliet
+
+* **Niels Van Vliet**
+
- Color conversion.
- Attribute opening and closing algorithms.
- Contributions to document the code.
@@ -191,21 +241,28 @@ Students from Class 2005.
Students from Class 2004.
=========================
-* Sylvain Berlemont
+* **Sylvain Berlemont**
+
- Combinatorial maps.
- Cleanup the sources.
-* Nicolas Burrus
+
+* **Nicolas Burrus**
+
- Integre.
- Olena I/O.
- Source tree.
-* David Lesage
+
+* **David Lesage**
+
- Type system.
- New paradigm.
- Cleanup the sources.
-* Astrid Wang
+
+* **Astrid Wang**
+
- New documentation system.
- Static arrays implementation.
@@ -213,25 +270,36 @@ Students from Class 2004.
Students from Class 2003.
=========================
-* Ignacy Gawedzki
+* **Ignacy Gawedzki**
+
- Color system.
-* Jean-Sébastien Mouret
+
+* **Jean-Sébastien Mouret**
+
- Image I/O.
- Source tree and configuration system.
- Fast morphological operators.
-* Quôc Peyrot
+
+* **Quôc Peyrot**
+
- Watershed algorithm.
-* Raphaël Poss
+
+* **Raphaël Poss**
+
- Source tree and configuration system.
- Documentation.
-* Yann Régis-Gianas
+
+* **Yann Régis-Gianas**
+
- Type system.
-* Emmanuel Turquin
+
+* **Emmanuel Turquin**
+
- Implementation of transforms.
- Integre.
@@ -239,22 +307,29 @@ Students from Class 2003.
Students from Class 2002.
=========================
-* Jean Chalard
+* **Jean Chalard**
+
- Colors.
- Vectors and matrices implementation.
- Olena iterators implementation.
- Wavelets.
-* Rémi Coupet
+
+* **Rémi Coupet**
+
- Olena morpho.
- Data types (pre-0.6).
- Olena core.
- Bibliographical research.
-* Ludovic Perrine
+
+* **Ludovic Perrine**
+
- Fuzzy types.
-* Pierre-Yves Strub
+
+* **Pierre-Yves Strub**
+
- Olena morpho.
- Source tree and configuration system.
- Type system.
@@ -263,20 +338,28 @@ Students from Class 2002.
Students from Class 2001.
=========================
-* Alexis Angelidis
+* **Alexis Angelidis**
+
- Olena logo.
-* Vincent Berruchon
-* Renaud François
+* **Vincent Berruchon**
+
+
+* **Renaud François**
+
- Bibliographical research.
-* Anthony Pinagot
+
+* **Anthony Pinagot**
+
- Olena I/O.
- Statistical operators.
- FFT algorithm.
-* Michaël Strauss
+
+* **Michaël Strauss**
+
- Image morphology.
- Watershed algorithms.
- Olena I/O.
@@ -285,7 +368,8 @@ Students from Class 2001.
Students from Class 2000.
=========================
-* Yoann Fabre
+* **Yoann Fabre**
+
- Type system.
@@ -293,17 +377,24 @@ Students from Class 2000.
Other Contributors.
===================
-* Guillaume Duhamel
+* **Guillaume Duhamel**
+
- Low-level routines.
-* Simon Nivault
+
+* **Simon Nivault**
+
- Lemming object.
- Many cleanups.
-* Dimitri Papadopoulos-Orfanos
+
+* **Dimitri Papadopoulos-Orfanos**
+
- Type system.
-* Heru Xue
+
+* **Heru Xue**
+
- Color system.
diff --git a/ChangeLog b/ChangeLog
index a9cc1f7..bb6ddc8 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -2,6 +2,31 @@
Improve common documentation.
+ * AUTHORS: Use RST syntax.
+
+ * doc/Doxyfile.in: Use .dox files for documentation.
+
+ * doc/DoxygenLayout.xml: Improve layout.
+
+ * doc/Makefile.am: Add dependencies.
+
+ * doc/lrde.bib,
+ * doc/contributors.dox: New.
+
+ * doc/bibliography.hh,
+ * doc/mainpage.hh,
+ * doc/modules_list.hh: Rename to...
+
+ * doc/bibliography.dox,
+ * doc/mainpage.dox,
+ * doc/modules_list.dox: ...this.
+
+ * doc/olena.qhcp: Remove.
+
+2012-02-01 Guillaume Lazzara <z(a)lrde.epita.fr>
+
+ Improve common documentation.
+
* Makefile.am: Add doc to SUBDIRS.
* doc/Doxyfile.in: Include bib file.
diff --git a/doc/Doxyfile.in b/doc/Doxyfile.in
index a0fcf00..86377ef 100644
--- a/doc/Doxyfile.in
+++ b/doc/Doxyfile.in
@@ -40,7 +40,7 @@ MULTILINE_CPP_IS_BRIEF = NO
INHERIT_DOCS = YES
SEPARATE_MEMBER_PAGES = NO
TAB_SIZE = 8
-ALIASES =
+ALIASES = " "
OPTIMIZE_OUTPUT_FOR_C = NO
OPTIMIZE_OUTPUT_JAVA = NO
BUILTIN_STL_SUPPORT = YES
@@ -92,17 +92,15 @@ WARN_LOGFILE =
# configuration options related to the input files
#---------------------------------------------------------------------------
INPUT = @top_srcdir@/doc
-FILE_PATTERNS = *.cc \
- *.hh \
- *.hxx \
- *.hcc
+FILE_PATTERNS = *.dox
RECURSIVE = YES
EXCLUDE = @top_srcdir@/milena/ \
@top_srcdir@/scribo/
EXCLUDE_SYMLINKS = YES
EXCLUDE_PATTERNS = *spe.hh
EXCLUDE_SYMBOLS = *::internal* *_ mln::trait::*
-EXAMPLE_PATH =
+EXAMPLE_PATH = @top_srcdir@/doc/ \
+ @builddir@
EXAMPLE_PATTERNS = *.cc \
*.cc.raw \
*.txt
diff --git a/doc/DoxygenLayout.xml b/doc/DoxygenLayout.xml
index a279910..0dbf0d7 100644
--- a/doc/DoxygenLayout.xml
+++ b/doc/DoxygenLayout.xml
@@ -2,7 +2,7 @@
<!-- Navigation index tabs for HTML output -->
<navindex>
<tab type="mainpage" visible="yes" title="Home"/>
- <tab type="pages" visible="no" title=""/>
+ <tab type="pages" visible="yes" title=""/>
<tab type="modules" visible="no" title=""/>
<tab type="namespaces" visible="no" title="">
<tab type="namespaces" visible="no" title=""/>
diff --git a/doc/Makefile.am b/doc/Makefile.am
index e733ab9..1bf95d5 100644
--- a/doc/Makefile.am
+++ b/doc/Makefile.am
@@ -53,13 +53,17 @@ regen-doc:
DOXYFILE_USER = Doxyfile_user
-REFMAN_deps = $(srcdir)/footer.html \
- $(srcdir)/header.html \
- $(srcdir)/doxygen.css \
- $(srcdir)/mainpage.hh \
- $(srcdir)/DoxygenLayout.xml \
- $(srcdir)/modules_list.hh \
- $(srcdir)/olena.bib
+REFMAN_deps = $(srcdir)/footer.html \
+ $(srcdir)/header.html \
+ $(srcdir)/doxygen.css \
+ $(srcdir)/mainpage.dox \
+ $(srcdir)/DoxygenLayout.xml \
+ $(srcdir)/modules_list.dox \
+ $(srcdir)/olena.bib \
+ $(srcdir)/contributors.dox \
+ $(srcdir)/bibliography.dox \
+ contributors.html \
+ lrde_olena.html
# Sed is used to generate Doxyfile from Doxyfile.in instead of
# configure, because the former is way faster than the latter.
@@ -113,8 +117,36 @@ $(srcdir)/$(USER_REFMAN): $(srcdir)/user-refman.stamp
EXTRA_DIST += $(USER_REFMAN)
+## -------------- ##
+## Contributors. ##
+## -------------- ##
+
+contributors.html: $(top_srcdir)/AUTHORS
+ rst2html --stylesheet-path=$(srcdir)/doxygen.css --link-stylesheet \
+ --input-encoding=latin1 --output-encoding=latin1 \
+ $(top_srcdir)/AUTHORS > contributors.html
+
+
+## -------------- ##
+## Bibliography. ##
+## -------------- ##
+
+# Warning: also generate lrde_olena.cite
+lrde_olena.bib:
+ bib2bib -s plain -oc lrde_olena.cite -ob lrde_olena.bib -c \
+ 'project = "Image" or project = "Olena"' \
+ -s year -r $(srcdir)/lrde.bib
+
+# Warning: also generate lrde_olena_bib.html
+lrde_olena.html: lrde_olena.bib
+ bibtex2html -s plain -nokeywords -citefile lrde_olena.cite \
+ -noabstract -nofooter -labelname -nodoc -linebreak -nf urllrde PDF lrde_olena.bib
+MAINTAINERCLEANFILES += lrde_olena.cite \
+ lrde_olena.bib \
+ lrde_olena_bib.html \
+ lrde_olena.html
## ---------- ##
## Cleaning. ##
diff --git a/doc/bibliography.dox b/doc/bibliography.dox
new file mode 100644
index 0000000..59518cb
--- /dev/null
+++ b/doc/bibliography.dox
@@ -0,0 +1,37 @@
+/** \page bibliography Bibliography
+
+ In this page, all publications used to implement Olena or related to Olena itself are referenced.
+
+ Summary:
+ \li \ref olnpublis
+ \li \ref olnpublisjournals
+ \li \ref olnpublisconfs
+ \li \ref olnpublisposters
+ \li \ref olnpublistechreps
+
+ \li \ref{burrus.03.mpool}
+ \li \ref{darbon.02.ismm}
+ \li \ref{darbon.04.ecoopphd}
+ \li \cite duret.00.gcse
+ \li \cite geraud.00.icpr
+ \li \cite geraud.01.ai
+ \li \cite geraud.08.mpool
+ \li \cite geraud.99.gretsi
+ \li \cite lazzara.11.icdar
+ \li \cite levillain.09.ismm
+ \li \cite levillain.10.icip
+ \li \cite levillain.10.wadgmm
+ \li \cite levillain.11.gretsi
+
+ \section olnpublis Olena Publications
+
+ \subsection olnpublisjournals In Journals
+
+ \subsection olnpublisconfs In Conferences
+ \htmlinclude lrde_olena.html
+
+ \subsection olnpublisposters Posters
+
+ \subsection olnpublistechreps Student Technical Reports
+
+ */
diff --git a/doc/bibliography.hh b/doc/bibliography.hh
deleted file mode 100644
index 841372b..0000000
--- a/doc/bibliography.hh
+++ /dev/null
@@ -1,19 +0,0 @@
-/** \page bibliography Bibliography
-
- Here you can find all the articles related to the Olena platform.
-
- \li \cite burrus.03.mpool
- \li \cite darbon.02.ismm
- \li \cite darbon.04.ecoopphd
- \li \cite duret.00.gcse
- \li \cite geraud.00.icpr
- \li \cite geraud.01.ai
- \li \cite geraud.08.mpool
- \li \cite geraud.99.gretsi
- \li \cite lazzara.11.icdar
- \li \cite levillain.09.ismm
- \li \cite levillain.10.icip
- \li \cite levillain.10.wadgmm
- \li \cite levillain.11.gretsi
-
- */
diff --git a/doc/contributors.dox b/doc/contributors.dox
new file mode 100644
index 0000000..82b35c6
--- /dev/null
+++ b/doc/contributors.dox
@@ -0,0 +1,5 @@
+/** \page Contributors List of contributors
+
+ \htmlinclude contributors.html
+
+ */
diff --git a/doc/lrde.bib b/doc/lrde.bib
new file mode 100644
index 0000000..c1d6a03
--- /dev/null
+++ b/doc/lrde.bib
@@ -0,0 +1,3856 @@
+ %% ----------------------- %%
+ %% The LRDE Bibliography. %%
+ %% ----------------------- %%
+
+%% This document lists all the publications by members of the LRDE.
+%% $Id: lrde.bib 4195 2011-10-11 10:08:20Z levill_r $
+%%
+%% See
+%%
+%% - http://www.lrde.epita.fr
+%% for more information about the LRDE
+%%
+%% - http://publis.lrde.epita.fr
+%% for more information about these publications
+%%
+%% - http://www.lrde.epita.fr/dload/papers/lrde.bib
+%% for the most recent version of this BibTeX file
+
+@InProceedings{ angelidis.01.wscg,
+ author = {Alexis Angelidis and Geoffroy Fouquier},
+ title = {Visualization issues in virtual environments: from
+ computer graphics techniques to intentional visualization},
+ booktitle = {Proceedings of the 9th International Conference in Central
+ Europe on Computer Graphics, Visualization and Computer
+ Vision (WSCG)},
+ year = 2001,
+ editor = {V. Skala},
+ volume = 3,
+ pages = {90--98},
+ address = {Plzen, Czech Republic},
+ month = feb,
+ isbn = {80-7082-713-0},
+ project = {Urbi},
+ urllrde = {200102-Wscg},
+ abstract = {Rendering efficiently large virtual environment scenes
+ composed of many elements, dynamic objects, and a highly
+ moving viewpoint is a major issue. This paper focuses on
+ the first of the two viewing stage operations: required
+ elements determination, the second being shading/filtering.
+ We propose a classification, extending the existing
+ computer graphic techniques toward display scalability
+ requirements, that distinguishes two key points: keeping
+ only required elements (culling), and keeping only required
+ details (which includes traditional LODs). The mechanisms
+ needed for display scalability are presented.}
+}
+
+@InProceedings{ baillard.05.adass,
+ author = {Anthony Baillard and Emmanuel Bertin and Yannic Mellier
+ and Henry Joy {McCracken} and {\relax Th}ierry G\'eraud and
+ Roser Pell\'o and Jean-Fran{\,c}ois {LeBorgne} and Pascal
+ Fouqu\'e},
+ title = {Project {EFIGI}: Automatic classification of galaxies},
+ year = 2005,
+ booktitle = {Astronomical Data Analysis Software and Systems XV},
+ volume = 351,
+ pages = {236--239},
+ publisher = {Astronomical Society of the Pacific},
+ series = {Conference},
+ url = {http://www.aspbooks.org/custom/publications/paper/index.phtml?paper_id=3398}
+ ,
+ editor = {Carlos Gabriel and {\relax Ch}ristophe Arviset and Daniel
+ Ponz and Enrique Solano},
+ isbn = {1-58381-219-9},
+ project = {Image},
+ urllrde = {200512-ADASS},
+ abstract = {We propose an automatic system to classify images of
+ galaxies with varying resolution. Morphologically typing
+ galaxies is a difficult task in particular for distant
+ galaxies convolved by a point-spread function and suffering
+ from a poor signal-to-noise ratio. In the context of the
+ first phase of the project EFIGI (extraction of the
+ idealized shapes of galaxies in imagery), we present the
+ three steps of our software: cleaning, dimensionality
+ reduction and supervised learning. We present preliminary
+ results derived from a subset of 774 galaxies from the
+ Principal Galaxies Catalog and compare them to human
+ classifications made by astronomers. We use g-band images
+ from the Sloan Digital Sky Survey. Finally, we discuss
+ future improvements which we intend to implement before
+ releasing our tool to the community.}
+}
+
+@InProceedings{ baillard.07.gretsi,
+ author = {Anthony Baillard and {\relax Ch}ristophe Berger and
+ Emmanuel Bertin and {\relax Th}ierry G\'eraud and Roland
+ Levillain and Nicolas Widynski},
+ title = {Algorithme de calcul de l'arbre des composantes avec
+ applications \`a la reconnaissance des formes en imagerie
+ satellitaire},
+ booktitle = {Proceedings of the 21st Symposium on Signal and Image
+ Processing (GRETSI)},
+ category = {national},
+ year = 2007,
+ address = {Troyes, France},
+ month = sep,
+ project = {Image},
+ urllrde = {200705-GRETSI},
+ abstract = {In this paper a new algorithm to compute the component
+ tree is presented. As compared to the state-of-the-art,
+ this algorithm does not use excessive memory and is able to
+ work efficiently on images whose values are highly
+ quantized or even with images having floating values. We
+ also describe how it can be applied to astronomical data to
+ identify relevant objects.}
+}
+
+@InProceedings{ bensalem.11.sumo,
+ author = {Ala Eddine Ben{ S}alem and Alexandre Duret-Lutz and
+ Fabrice Kordon},
+ title = {Generalized {B\"u}chi Automata versus Testing Automata for
+ Model Checking},
+ booktitle = {Proceedings of the second International Workshop on
+ Scalable and Usable Model Checking for Petri Net and other
+ models of Concurrency (SUMO'11)},
+ address = {Newcastle, UK},
+ series = {Workshop Proceedings},
+ year = 2011,
+ month = jun,
+ volume = 626,
+ publisher = {CEUR},
+ urllrde = {201106-SUMO},
+ url = {http://ftp.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-726/}
+
+}
+
+@InProceedings{ berger.07.icip,
+ author = {{\relax Ch}ristophe Berger and {\relax Th}ierry G\'eraud
+ and Roland Levillain and Nicolas Widynski and Anthony
+ Baillard and Emmanuel Bertin},
+ title = {Effective Component Tree Computation with Application to
+ Pattern Recognition in Astronomical Imaging},
+ booktitle = {Proceedings of the IEEE International Conference on Image
+ Processing (ICIP)},
+ pages = {IV-41--IV-44},
+ volume = 4,
+ year = 2007,
+ address = {San Antonio, TX, USA},
+ month = sep,
+ project = {Image},
+ urllrde = {200705-ICIP},
+ abstract = {In this paper a new algorithm to compute the component
+ tree is presented. As compared to the state of the art,
+ this algorithm does not use excessive memory and is able to
+ work efficiently on images whose values are highly
+ quantized or even with images having floating values. We
+ also describe how it can be applied to astronomical data to
+ identify relevant objects.}
+}
+
+@Article{ bloch.03.ai,
+ author = {Isabelle Bloch and {\relax Th}ierry G\'eraud and Henri
+ Ma\^itre},
+ title = {Representation and fusion of heterogeneous fuzzy
+ information in the {3D} space for model-based structural
+ recognition---application to {3D} brain imaging},
+ journal = {Artificial Intelligence},
+ month = aug,
+ year = 2003,
+ volume = 148,
+ number = {1-2},
+ pages = {141--175},
+ project = {Image},
+ urllrde = {200308-AI},
+ abstract = {We present a novel approach of model-based pattern
+ recognition where structural information and spatial
+ relationships have a most important role. It is illustrated
+ in the domain of 3D brain structure recognition using an
+ anatomical atlas. Our approach performs simultaneously
+ segmentation and recognition of the scene and the solution
+ of the recognition task is progressive, processing
+ successively different objects, using different of
+ knowledge about the object and about relationships between
+ objects. Therefore the core of the approach is the
+ representation part, and constitutes the main contribution
+ of this paper. We make use of a spatial representation of
+ each piece of information, as a spatial set representing a
+ constraint to be satisfied by the searched object, thanks
+ in particular to fuzzy mathematical operations. Fusion of
+ these constraints allows to, segment and recognize the
+ desired object.}
+}
+
+@Article{ bloch.05.prl,
+ author = {Isabelle Bloch and Olivier Colliot and Oscar Camara and
+ {\relax Th}ierry G\'eraud},
+ title = {Fusion of spatial relationships for guiding recognition,
+ example of brain structure recognition in {3D} {MRI}},
+ journal = {Pattern Recognition Letters},
+ year = 2005,
+ volume = 26,
+ number = 4,
+ month = mar,
+ pages = {449--457},
+ project = {Image},
+ urllrde = {200407-PRL},
+ abstract = {Spatial relations play an important role in recognition of
+ structures embedded in a complex environment and for
+ reasoning under imprecision. Several types of relationships
+ can be modeled in a unified way using fuzzy mathematical
+ morphology. Their combination benefits from the powerful
+ framework of fuzzy set theory for fusion tasks and decision
+ making. This paper presents several methods of fusion of
+ information about spatial relationships and illustrates
+ them on the example of model-based recognition of brain
+ structures in 3D magnetic resonance imaging.}
+}
+
+@Article{ borghi.06.crossroads,
+ author = {Alexandre Borghi and Valentin David and Akim Demaille},
+ title = {{C}-{T}ransformers --- {A} Framework to Write {C} Program
+ Transformations},
+ journal = {ACM Crossroads},
+ year = 2006,
+ volume = 12,
+ number = 3,
+ month = {Spring},
+ project = {Transformers},
+ note = {\url{http://www.acm.org/crossroads/xrds12-3/contractc.html}}
+ ,
+ urllrde = {200510-Crossroads},
+ abstract = {Program transformation techniques have reached a maturity
+ level that allows processing high-level language sources in
+ new ways. Not only do they revolutionize the implementation
+ of compilers and interpreters, but with modularity as a
+ design philosophy, they also permit the seamless extension
+ of the syntax and semantics of existing programming
+ languages. The C-Transformers project provides a
+ transformation environment for C, a language that proves to
+ be hard to transform. We demonstrate the effectiveness of
+ C-Transformers by extending C's instructions and control
+ flow to support Design by Contract. C-Transformers is
+ developed by members of the LRDE: EPITA undergraduate
+ students.}
+}
+
+@InProceedings{ burrus.03.mpool,
+ author = {Nicolas Burrus and Alexandre Duret-Lutz and {\relax
+ Th}ierry G\'eraud and David Lesage and Rapha\"el Poss},
+ title = {A static {C++} object-oriented programming ({SCOOP})
+ paradigm mixing benefits of traditional {OOP} and generic
+ programming},
+ booktitle = {Proceedings of the Workshop on Multiple Paradigm with
+ Object-Oriented Languages (MPOOL)},
+ year = 2003,
+ address = {Anaheim, CA, USA},
+ month = oct,
+ project = {Olena},
+ urllrde = {200310-MPOOL},
+ abstract = {Object-oriented and generic programming are both supported
+ in C++. OOP provides high expressiveness whereas GP leads
+ to more efficient programs by avoiding dynamic typing. This
+ paper presents SCOOP, a new paradigm which enables both
+ classical OO design and high performance in C++ by mixing
+ OOP and GP. We show how classical and advanced OO features
+ such as virtual methods, multiple inheritance, argument
+ covariance, virtual types and multimethods can be
+ implemented in a fully statically typed model, hence
+ without run-time overhead.}
+}
+
+@InProceedings{ cadilhac.06.avocs,
+ author = {Micha\"el Cadilhac and {\relax Th}omas H\'erault and
+ Richard Lassaigne and Sylvain Peyronnet and Sebastien
+ Tixeuil},
+ title = {Evaluating complex {MAC} protocols for sensor networks
+ with {APMC}},
+ booktitle = {Proceedings of the 6th International Workshop on Automated
+ Verification of Critical Systems (AVoCS)},
+ year = 2006,
+ series = {Electronic Notes in Theoretical Computer Science Series},
+ pages = {33--46},
+ volume = 185,
+ project = {APMC},
+ urllrde = {200606-Avocs},
+ abstract = {In this paper we present an analysis of a MAC (Medium
+ Access Control) protocol for wireless sensor networks. The
+ purpose of this protocol is to manage wireless media access
+ by constructing a Time Division Media Access (TDMA)
+ schedule. APMC (Approximate Probabilistic Model Checker) is
+ a tool that uses approximation-based verification
+ techniques in order to analyse the behavior of complex
+ probabilistic systems. Using APMC, we approximately
+ computed the probabilities of several properties of the MAC
+ protocol being studied, thus giving some insights about its
+ performance.}
+}
+
+@InProceedings{ carlier.02.itrs,
+ author = {S\'ebastien Carlier},
+ title = {Polar type inference with intersection types and $\omega$},
+ booktitle = {Proceedings of the 2nd Workshop on Intersection Types and
+ Related Systems (ITRS), published in: Electronic Notes in
+ Theoretical Computer Science},
+ volume = 70,
+ issue = 1,
+ publisher = {Elsevier},
+ year = 2002,
+ address = {Copenhagen, Denmark},
+ month = jul,
+ project = {Software},
+ urllrde = {200207-ITRS},
+ abstract = {We present a type system featuring intersection types and
+ omega, a type constant which is assigned to unused terms.
+ We exploit and extend the technology of expansion variables
+ from the recently developed System I, with which we believe
+ our system shares many interesting properties, such as
+ strong normalization, principal typings, and compositional
+ analysis. Our presentation emphasizes a polarity discipline
+ and shows its benefits. We syntactically distinguish
+ positive and negative types, and give them different
+ interpretations. We take the point of view that the
+ interpretation of a type is intrinsic to it, and should not
+ change implicitly when it appears at the opposite polarity.
+ Our system is the result of a process which started with an
+ extension of Trevor Jim's Polar Type System.}
+}
+
+@InProceedings{ chekroun.06.iciar,
+ author = {Mickael Chekroun and J\'er\^ome Darbon and Igor Ciril},
+ title = {On a Polynomial Vector Field Model for Shape
+ Representation},
+ booktitle = {Proceedings of the International Conference on Image
+ Analysis and Recognition (ICIAR)},
+ publisher = {Springer-Verlag},
+ year = 2006,
+ address = {Povoa de Varzim, Portugal},
+ month = sep,
+ project = {Image},
+ urllrde = {200609-ICIAR},
+ abstract = {In this paper we propose an efficient algorithm to perform
+ a polynomial approximation of the vector field derived from
+ the usual distance mapping method. The main ingredients
+ consist of minimizing a quadratic functional and
+ transforming this problem in an appropriate setting for
+ implementation. With this approach, we reduce the problem
+ of obtaining an approximating polynomial vector field to
+ the resolution of a not expansive linear algebraic system.
+ By this procedure, we obtain an analytical shape
+ representation that relies only on some coefficients.
+ Fidelity and numerical efficiency of our approach are
+ presented on illustrative examples.}
+}
+
+@InProceedings{ claveirole.05.ciaa,
+ author = {{\relax Th}omas Claveirole and Sylvain Lombardy and Sarah
+ O'Connor and Louis-No\"el Pouchet and Jacques Sakarovitch},
+ title = {Inside {V}aucanson},
+ booktitle = {Proceedings of Implementation and Application of Automata,
+ 10th International Conference (CIAA)},
+ year = 2005,
+ pages = {117--128},
+ editor = {Springer-Verlag},
+ volume = 3845,
+ series = {Lecture Notes in Computer Science Series},
+ address = {Sophia Antipolis, France},
+ month = jun,
+ urllrde = {200506-Ciaa},
+ project = {Vaucanson},
+ abstract = {This paper presents some features of the Vaucanson
+ platform. We describe some original algorithms on weighted
+ automata and transducers (computation of the quotient,
+ conversion of a regular expression into a weighted
+ automaton, and composition). We explain how complex
+ declarations due to the generic programming are masked from
+ the user and finally we present a proposal for an XML
+ format that allows implicit descriptions for simple types
+ of automata.}
+}
+
+@TechReport{ clouard.99.tr,
+ author = {R\'egis Clouard and Abderrahim Elmoataz and Fran\c{c}ois
+ Angot and Olivier Lezoray and Alexandre Duret-Lutz},
+ title = {Une biblioth\`eque et un environnement de programmation
+ d'op\'erateurs de traitement d'images},
+ institution = {GREYC-ISMRA},
+ year = 1999,
+ number = 99008,
+ address = {Caen, France},
+ month = nov,
+ url = {http://www.greyc.ismra.fr/~regis/Pandore/},
+ project = {Olena},
+ urllrde = {199911-TR}
+}
+
+@InProceedings{ darbon.01.ei,
+ author = {J\'er\^ome Darbon and Bulent Sankur and Henri Ma\^{\i}tre},
+ title = {Error correcting code performance for watermark
+ protection},
+ booktitle = {Proceedings of the 13th Symposium SPIE on Electronic
+ Imaging----Security and Watermarking of Multimedia Contents
+ III (EI27)},
+ year = 2001,
+ address = {San Jose, CA, USA},
+ month = jan,
+ volume = 4314,
+ editors = {P.W. Wong and E.J. Delp III},
+ pages = {663--672},
+ project = {Image},
+ urllrde = {200101-Ei},
+ abstract = {The watermark signals are weakly inserted in images due to
+ imperceptibility constraints which makes them prone to
+ errors in the extraction stage. Although the error
+ correcting codes can potentially improve their performance
+ one must pay attention to the fact that the watermarking
+ channel is in general very noisy. We have considered the
+ trade-off of the BCH codes and repetition codes in various
+ concatenation modes. At the higher rates that can be
+ encountered in watermarking channels such as due to
+ low-quality JPEG compression, codes like the BCH codes
+ cease being useful. Repetition coding seems to be the last
+ resort at these error rates of 25\% and beyond. It has been
+ observed that there is a zone of bit error rate where their
+ concatenation turns out to be more useful. In fact the
+ concatenation of repetition and BCH codes judiciously
+ dimensioned, given the available number of insertion sites
+ and the payload size, achieves a higher reliability level.}
+}
+
+@InProceedings{ darbon.02.ismm,
+ author = {J\'er\^ome Darbon and {\relax Th}ierry G\'eraud and
+ Alexandre Duret-Lutz},
+ title = {Generic implementation of morphological image operators},
+ booktitle = {Mathematical Morphology, Proceedings of the 6th
+ International Symposium (ISMM)},
+ pages = {175--184},
+ year = 2002,
+ address = {Sydney, Australia},
+ month = apr,
+ publisher = {CSIRO Publishing},
+ project = {Olena},
+ urllrde = {200204-Ismm},
+ abstract = {Several libraries dedicated to mathematical morphology
+ exist. But they lack genericity, that is to say, the
+ ability for operators to accept input of different natures
+ ---2D binary images, graphs enclosing floating values, etc.
+ We describe solutions which are integrated in Olena, a
+ library providing morphological operators. We demonstrate
+ with some examples that translating mathematical formulas
+ and algorithms into source code is made easy and safe with
+ Olena. Moreover, experimental results show that no extra
+ costs at run-time are induced.}
+}
+
+@InProceedings{ darbon.04.ecoopphd,
+ author = {J\'er\^ome Darbon and {\relax Th}ierry G\'eraud and
+ Patrick Bellot},
+ title = {Generic algorithmic blocks dedicated to image processing},
+ booktitle = {Proceedings of the ECOOP Workshop for PhD Students},
+ year = 2004,
+ address = {Oslo, Norway},
+ month = jun,
+ project = {Olena},
+ urllrde = {200406-ECOOPPHD},
+ abstract = {This paper deals with the implementation of algorithms in
+ the specific domain of image processing. Although many
+ image processing libraries are available, they generally
+ lack genericity and flexibility. Many image processing
+ algorithms can be expressed as compositions of elementary
+ algorithmic operations referred to as blocks. Implementing
+ these compositions is achieved using generic programming.
+ Our solution is compared to previous ones and we
+ demonstrate it on a class image processing algorithms.}
+}
+
+@InProceedings{ darbon.04.iwcia,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {Exact optimization of discrete constrained total variation
+ minimization problems},
+ booktitle = {Proceedings of the 10th International Workshop on
+ Combinatorial Image Analysis (IWCIA)},
+ year = 2004,
+ address = {Auckland, New Zealand},
+ month = dec,
+ pages = {548--557},
+ editors = {R. Klette and J. Zunic},
+ series = {Lecture Notes in Computer Science Series},
+ publisher = {Springer-Verlag},
+ volume = 3322,
+ project = {Image},
+ urllrde = {200412-IWCIA},
+ abstract = {This paper deals with the total variation minimization
+ problem when the fidelity is either the $L^2$-norm or the
+ $L^1$-norm. We propose an algorithm which computes the
+ exact solution of these two problems after discretization.
+ Our method relies on the decomposition of an image into its
+ level sets. It maps the original problems into independent
+ binary Markov Random Field optimization problems associated
+ with each level set. Exact solutions of these binary
+ problems are found thanks to minimum-cut techniques. We
+ prove that these binary solutions are increasing and thus
+ allow to reconstruct the solution of the original
+ problems.}
+}
+
+@TechReport{ darbon.04.tr,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {Exact optimization of discrete constrained total variation
+ minimization problems},
+ institution = {ENST},
+ year = 2004,
+ number = {2004C004},
+ address = {Paris, France},
+ month = oct,
+ project = {Image},
+ annote = {This technical report corresponds to the publication
+ darbon.04.iwcia. ; 200412-IWCIA},
+ urllrde = {200410-TR},
+ abstract = {This paper deals with the total variation minimization
+ problem when the fidelity is either the $L^2$-norm or the
+ $L^1$-norm. We propose an algorithm which computes the
+ exact solution of these two problems after discretization.
+ Our method relies on the decomposition of an image into its
+ level sets. It maps the original problems into independent
+ binary Markov Random Field optimization problems associated
+ with each level set. Exact solutions of these binary
+ problems are found thanks to minimum-cut techniques. We
+ prove that these binary solutions are increasing and thus
+ allow to reconstruct the solution of the original
+ problems.}
+}
+
+@InProceedings{ darbon.05.eusipco,
+ author = {J\'er\^ome Darbon and Ceyhun Burak Akg\"ul},
+ title = {An efficient algorithm for attribute openings and
+ closings},
+ booktitle = {Proceedings of the 13th European Signal Processing
+ Conference (EUSIPCO)},
+ year = 2005,
+ address = {Antalya, Turkey},
+ month = sep,
+ project = {Image},
+ urllrde = {200509-Eusipco},
+ abstract = {In this paper, we present fast algorithms for area opening
+ and closing on grayscale images. Salembier's max-tree based
+ algorithm is one of the well known methods to perform area
+ opening. It makes use of a special representation where
+ each node in the tree stands for a flat region and the tree
+ itself is oriented towards the maxima of the grayscale
+ image. Pruning the tree with respect to some attribute,
+ e.g., the area, boils down to attribute opening. Following
+ the same approach, we propose an algorithm for area opening
+ (closing) without building the max-tree (min-tree). Our
+ algorithm exhibit considerable performance compared to the
+ state-of-the art in this domain.}
+}
+
+@InProceedings{ darbon.05.ibpria,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {A Fast and Exact Algorithm for Total Variation
+ Minimization},
+ booktitle = {Proceedings of the 2nd Iberian Conference on Pattern
+ Recognition and Image Analysis (IbPRIA)},
+ publisher = {Springer-Verlag},
+ volume = 3522,
+ pages = {351--359},
+ year = 2005,
+ address = {Estoril, Portugal},
+ month = jun,
+ project = {Image},
+ urllrde = {200506-IbPria},
+ abstract = {This paper deals with the minimization of the total
+ variation under a convex data fidelity term. We propose an
+ algorithm which computes an exact minimizer of this
+ problem. The method relies on the decomposition of an image
+ into its level sets. Using these level sets, we map the
+ problem into optimizations of independent binary Markov
+ Random Fields. Binary solutions are found thanks to
+ graph-cut techniques and we show how to derive a fast
+ algorithm. We also study the special case when the fidelity
+ term is the $L^1$-norm. Finally we provide some
+ experiments.}
+}
+
+@InProceedings{ darbon.05.ispa,
+ author = {J\'er\^ome Darbon},
+ title = {Total Variation Minimization with $L^1$ Data Fidelity as a
+ Contrast Invariant Filter},
+ booktitle = {Proceedings of the 4th International Symposium on Image
+ and Signal Processing and Analysis (ISPA 2005)},
+ year = 2005,
+ address = {Zagreb, Croatia},
+ month = sep,
+ pages = {221--226},
+ project = {Image},
+ urllrde = {200509-Ispa},
+ abstract = {This paper sheds new light on minimization of the total
+ variation under the $L^1$-norm as data fidelity term
+ ($L^1+TV$) and its link with mathematical morphology. It is
+ well known that morphological filters enjoy the property of
+ being invariant with respect to any change of contrast.
+ First, we show that minimization of $L^1+TV$ yields a
+ self-dual and contrast invariant filter. Then, we further
+ constrain the minimization process by only optimizing the
+ grey levels of level sets of the image while keeping their
+ boundaries fixed. This new constraint is maintained thanks
+ to the Fast Level Set Transform which yields a complete
+ representation of the image as a tree. We show that this
+ filter can be expressed as a Markov Random Field on this
+ tree. Finally, we present some results which demonstrate
+ that these new filters can be particularly useful as a
+ preprocessing stage before segmentation.}
+}
+
+@InProceedings{ darbon.05.isvc,
+ author = {J\'er\^ome Darbon and Sylvain Peyronnet},
+ title = {A Vectorial Self-Dual Morphological Filter based on Total
+ Variation Minimization},
+ booktitle = {Proceedings of the First International Conference on
+ Visual Computing},
+ year = 2005,
+ address = {Lake Tahoe, Nevada, USA},
+ month = dec,
+ project = {Image},
+ pages = {388--395},
+ series = {Lecture Notes in Computer Science Series},
+ publisher = {Springer-Verlag},
+ volume = 3804,
+ urllrde = {200512-Isvc},
+ abstract = {We present a vectorial self dual morphological filter.
+ Contrary to many methods, our approach does not require the
+ use of an ordering on vectors. It relies on the
+ minimization of the total variation with $L^1$ norm as data
+ fidelity on each channel. We further constraint this
+ minimization in order not to create new values. It is shown
+ that this minimization yields a self-dual and contrast
+ invariant filter. Although the above minimization is not a
+ convex problem, we propose an algorithm which computes a
+ global minimizer. This algorithm relies on minimum cost
+ cut-based optimizations.}
+}
+
+@PhDThesis{ darbon.05.phd,
+ author = {J\'er\^ome Darbon},
+ title = {Composants logiciels et algorithmes de minimisation exacte
+ d'\'energies d\'edid\'ees au traitement d'images},
+ school = {\'Ecole Nationale Sup\'erieure des T\'el\'ecommunications
+ de Paris (ENST)},
+ address = {Paris, France},
+ year = 2005,
+ month = oct,
+ number = {XXX},
+ note = {In French},
+ urllrde = {200510-PhD},
+ abstract = {Dans cette th\`ese nous \'etudions la minimisation
+ d'\'energies markoviennes rencontr\'ees dans les domaines
+ du traitement des images et de la vision par ordinateur.
+ Nous proposons des algorithmes de minimisation exacte pour
+ diff\'erents types d'\'energies. Ces algorithmes ont
+ l'int\'er\^et de fournir un minimum global quand bien
+ m\^eme l'\'energie n'est pas convexe. Enfin, nous mettons
+ en \'evidence quelques liens entre les champs de Markov
+ binaires et la morphologie math\'ematique. La version
+ finale de ce manuscrit suit les recommandations des rapporteurs.}
+}
+
+@TechReport{ darbon.05.tr,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {A fast and exact algorithm for total variation
+ minimization},
+ institution = {ENST},
+ year = 2005,
+ number = {2005D002},
+ address = {Paris, France},
+ month = jan,
+ project = {Image},
+ annote = {This technical report corresponds to the publication
+ darbon.05.ibpria.},
+ urllrde = {200501-TR},
+ abstract = {This paper deals with the minimization of the total
+ variation under a convex data fidelity term. We propose an
+ algorithm which computes an exact minimizer of this
+ problem. The method relies on the decomposition of an image
+ into its level sets. Using these level sets, we map the
+ problem into optimizations of independent binary Markov
+ Random Fields. Binary solutions are found thanks to
+ graph-cut techniques and we show how to derive a fast
+ algorithm. We also study the special case when the fidelity
+ term is the $L^1$-norm. Finally we provide some
+ experiments.}
+}
+
+@InProceedings{ darbon.06.iccp,
+ author = {J\'er\^ome Darbon and Richard Lassaigne and Sylvain
+ Peyronnet},
+ title = {Approximate Probabilistic Model Checking for Programs},
+ booktitle = {Proceedings of the {IEEE} 2nd International Conference on
+ Intelligent Computer Communication and Processing
+ ({ICCP'06)}},
+ year = 2006,
+ address = {Technical University of Cluj-Napoca, Romania},
+ month = sep,
+ urllrde = {200606-Iccp2},
+ abstract = {In this paper we deal with the problem of applying model
+ checking to real programs. We verify a program without
+ constructing the whole transition system using a technique
+ based on Monte-Carlo sampling, also called ``approximate
+ model checking''. This technique combines model checking
+ and randomized approximation. Thus, it avoids the so called
+ state space explosion phenomenon. We propose a prototype
+ implementation that works directly on C source code. It
+ means that, contrary to others approaches, we do not need
+ to use a specific language nor specific data structures in
+ order to describe the system we wish to verify. Finally, we
+ present experimental results that show the effectiveness of
+ the approach applied to finding bugs in real programs.}
+}
+
+@Article{ darbon.06.jmiv,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {Image restoration with discrete constrained {T}otal
+ {Variation}---Part~{I}: Fast and exact optimization},
+ journal = {Journal of Mathematical Imaging and Vision},
+ year = 2006,
+ volume = 26,
+ number = 3,
+ month = dec,
+ pages = {261--276},
+ project = {Image},
+ urllrde = {2006XXX-JMIVa},
+ abstract = {This paper deals with the total variation minimization
+ problem in image restoration for convex data fidelity
+ functionals. We propose a new and fast algorithm which
+ computes an exact solution in the discrete framework. Our
+ method relies on the decomposition of an image into its
+ level sets. It maps the original problems into independent
+ binary Markov Random Field optimization problems at each
+ level. Exact solutions of these binary problems are found
+ thanks to minimum cost cut techniques in graphs. These
+ binary solutions are proved to be monotone increasing with
+ levels and yield thus an exact solution of the discrete
+ original problem. Furthermore we show that minimization of
+ total variation under $L^1$ data fidelity term yields a
+ self-dual contrast invariant filter. Finally we present
+ some results.}
+}
+
+@Article{ darbon.06.jmivb,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {Image restoration with discrete constrained {T}otal
+ {Variation}---Part~{II}: Levelable functions, convex priors
+ and non-convex case},
+ journal = {Journal of Mathematical Imaging and Vision},
+ year = 2006,
+ volume = 26,
+ number = 3,
+ month = dec,
+ pages = {277--291},
+ project = {Image},
+ urllrde = {2006XXX-JMIVb},
+ abstract = {In Part II of this paper we extend the results obtained in
+ Part I for total variation minimization in image
+ restoration towards the following directions: first we
+ investigate the decomposability property of energies on
+ levels, which leads us to introduce the concept of
+ levelable regularization functions (which TV is the
+ paradigm of). We show that convex levelable posterior
+ energies can be minimized exactly using the
+ level-independant cut optimization scheme seen in part I.
+ Next we extend this graph cut scheme optimization scheme to
+ the case of non-convex levelable energies. We present
+ convincing restoration results for images corrupted with
+ impulsive noise. We also provide a minimum-cost based
+ algorithm which computes a global minimizer for Markov
+ Random Field with convex priors. Last we show that
+ non-levelable models with convex local conditional
+ posterior energies such as the class of generalized
+ gaussian models can be exactly minimized with a generalized
+ coupled Simulated Annealing.}
+}
+
+@InProceedings{ darbon.06.siam,
+ author = {J\'er\^ome Darbon and Marc Sigelle},
+ title = {Fast and Exact Discrete Image Restoration Based on Total
+ Variation and on Its Extensions to Levelable Potentials},
+ booktitle = {SIAM Conference on Imaging Sciences},
+ year = 2006,
+ address = {Minneapolis, USA},
+ month = may,
+ project = {Image},
+ urllrde = {200605-SIAM},
+ abstract = {We investigate the decomposition property of posterior
+ restoration energies on level sets in a discrete Markov
+ Random Field framework. This leads us to the concept of
+ 'levelable' potentials (which TV is shown to be the
+ paradigm of). We prove that convex levelable posterior
+ energies can be minimized exactly with level-independant
+ binary graph cuts. We extend this scheme to the case of
+ non-convex levelable energies, and present convincing
+ restoration results for images degraded by impulsive
+ noise.}
+}
+
+@TechReport{ darbon.06.tr,
+ author = {J\'er\^ome Darbon and Marc Sigelle and Florence Tupin},
+ title = {A note on nice-levelable {MRFs} for {SAR} image denoising
+ with contrast preservation},
+ institution = {Signal and Image Processing Group, Ecole Nationale
+ Sup\'erieure des T\'el\'ecommunications},
+ year = 2006,
+ number = {2006D006},
+ address = {Paris, France},
+ month = sep,
+ project = {Image},
+ annote = {On this technical report is based the publication
+ darbon.07.ei ; 200701-SPIE},
+ urllrde = {200701-SPIE}
+}
+
+@InProceedings{ darbon.07.ei,
+ author = {J\'er\^ome Darbon and Marc Sigelle and Florence Tupin},
+ title = {The use of levelable regularization functions for {MRF}
+ restoration of {SAR} images},
+ booktitle = {Proceedings of the 19th Symposium SPIE on Electronic
+ Imaging},
+ year = 2007,
+ address = {San Jose, CA, USA},
+ month = jan,
+ project = {Image},
+ urllrde = {200701-SPIE},
+ abstract = {It is well-known that Total Variation (TV) minimization
+ with L2 data fidelity terms (which corresponds to white
+ Gaussian additive noise) yields a restored image which
+ presents some loss of contrast. The same behavior occurs
+ for TVmodels with non-convex data fidelity terms that
+ represent speckle noise. In this note we propose a new
+ approach to cope with the restoration of Synthetic Aperture
+ Radar images while preserving the contrast.}
+}
+
+@InProceedings{ darbon.07.mirage,
+ author = {J\'er\^ome Darbon},
+ title = {A Note on the Discrete Binary {Mumford-Shah} Model},
+ booktitle = {Proceedings of the international Computer Vision /
+ Computer Graphics Collaboration Techniques and Applications
+ (MIRAGE 2007)},
+ year = 2007,
+ address = {Paris, France},
+ month = mar,
+ project = {Image},
+ urllrde = {200703-MIRAGE},
+ abstract = {This paper is concerned itself with the analysis of the
+ two-phase Mumford-Shah model also known as the active
+ contour without edges model introduced by Chan and Vese. It
+ consists of approximating an observed image by a piecewise
+ constant image which can take only two values. First we
+ show that this model with the $L^1$-norm as data fidelity
+ yields a contrast invariant filter which is a well known
+ property of morphological filters. Then we consider a
+ discrete version of the original problem. We show that an
+ inclusion property holds for the minimizers. The latter is
+ used to design an efficient graph-cut based algorithm which
+ computes an exact minimizer. Some preliminary results are
+ presented.}
+}
+
+@InProceedings{ darbon.08.iwcia,
+ author = {J\'er\^ome Darbon},
+ title = {Global Optimization for First Order {Markov} Random Fields
+ with Submodular Priors},
+ booktitle = {Proceedings of the twelfth International Workshop on
+ Combinatorial Image Analysis (IWCIA'08) },
+ year = 2008,
+ address = {Buffalo, New York, USA},
+ month = apr,
+ project = {Image},
+ urllrde = {200801-IWCIA},
+ abstract = {This paper copes with the optimization of Markov Random
+ Fields with pairwise interactions defined on arbitrary
+ graphs. The set of labels is assumed to be linearly ordered
+ and the priors are supposed to be submodular. Under these
+ assumptions we propose an algorithm which computes an exact
+ minimizer of the Markovian energy. Our approach relies on
+ mapping the original into a combinatorial one which
+ involves only binary variables. The latter is shown to be
+ exactly solvable via computing a maximum flow. The
+ restatement into a binary combinatorial problem is done by
+ considering the level-sets of the labels instead of the
+ label values themselves. The submodularity of the priors is
+ shown to be a necessary and sufficient condition for the
+ applicability of the proposed approach.}
+}
+
+@Misc{ david.05.sud,
+ author = {Valentin David and Akim Demaille and Renaud Durlin and
+ Olivier Gournet},
+ title = {{C}/{C++} Disambiguation Using Attribute Grammars},
+ year = 2005,
+ note = {Communication to Stratego Users Day 2005},
+ address = {Utrecht {U}niversity, {N}etherland},
+ month = may,
+ project = {Transformers},
+ urllrde = {200505-SUD-disamb},
+ abstract = {We propose a novel approach to semantics driven
+ disambiguation based on Attribute Grammars (AGs). AGs share
+ the same modularity model as its host grammar language,
+ here Syntax Definition Formalism (SDF), what makes them
+ particularly attractive for working on unstable grammars,
+ or grammar extensions. The framework we propose is
+ effective, since a full ISO-C99 disambiguation chain
+ already works, and the core of the hardest ambiguities of
+ C++ is solved. This requires specific techniques, and some
+ extensions to the stock AG model.}
+}
+
+@InProceedings{ david.06.iccp,
+ author = {Valentin David and Akim Demaille and Olivier Gournet},
+ title = {Attribute Grammars for Modular Disambiguation},
+ booktitle = {Proceedings of the {IEEE} 2nd International Conference on
+ Intelligent Computer Communication and Processing
+ ({ICCP'06)}},
+ year = 2006,
+ address = {Technical University of Cluj-Napoca, Romania},
+ month = sep,
+ urllrde = {200606-ICCP},
+ abstract = {To face the challenges to tomorrow's software engineering
+ tools, powerful language-generic program-transformation
+ components are needed. We propose the use of attribute
+ grammars (AGs) to generate language specific disambiguation
+ filters. In this paper, a complete implementation of a
+ language-independent AGs system is presented. As a full
+ scale experiment, we present an implementation of a
+ flexible C front-end. Its specifications are concise,
+ modular, and the result is efficient. On top of it,
+ transformations such as software renovation, code metrics,
+ domain specific language embedding can be implemented.},
+ project = {Transformers}
+}
+
+@Article{ dehak.05.pami,
+ author = {R\'eda Dehak and Isabelle Bloch and Henri Ma{\^\i}tre},
+ title = {Spatial reasoning with relative incomplete information on
+ relative positioning},
+ journal = {IEEE Transactions on Pattern Analysis and Machine
+ Intelligence},
+ year = 2005,
+ pages = {1473--1484},
+ volume = 27,
+ month = sep,
+ number = 9,
+ project = {Image},
+ urllrde = {200509-PAMI},
+ abstract = {This paper describes a probabilistic method of inferring
+ the position of a point with respect to a reference point
+ knowing their relative spatial position to a third point.
+ We address this problem in the case of incomplete
+ information where only the angular spatial relationships
+ are known. The use of probabilistic representations allows
+ us to model prior knowledge. We derive exact formulae
+ expressing the conditional probability of the position
+ given the two known angles, in typical cases: uniform or
+ Gaussian random prior distributions within rectangular or
+ circular regions. This result is illustrated with respect
+ to two different simulations: The first is devoted to the
+ localization of a mobile phone using only angular
+ relationships, the second, to geopositioning within a city.
+ This last example uses angular relationships and some
+ additional knowledge about the position.}
+}
+
+@InProceedings{ dehak.06.nist,
+ author = {R\'eda Dehak and Charles-Alban Deledalle and Najim Dehak},
+ title = {{LRDE} System description},
+ booktitle = {NIST SRE'06 Workshop: speaker recognition evaluation
+ campaign},
+ year = 2006,
+ address = {San Juan, Puerto Rico},
+ month = jun,
+ urllrde = {200606-NIST-A}
+}
+
+@InProceedings{ dehak.07.interspeech,
+ author = {R\'eda Dehak and Najim Dehak and Patrick Kenny and Pierre
+ Dumouchel},
+ title = {Linear and Non Linear Kernel {GMM} SuperVector Machines
+ for Speaker Verification},
+ booktitle = {Proceedings of the European Conference on Speech
+ Communication and Technologies (Interspeech'07)},
+ year = 2007,
+ address = {Antwerp, Belgium},
+ month = aug,
+ urllrde = {200708-INTERSPEECH},
+ abstract = {This paper presents a comparison between Support Vector
+ Machines (SVM) speaker verification systems based on linear
+ and non linear kernels defined in GMM supervector space. We
+ describe how these kernel functions are related and we show
+ how the nuisance attribute projection (NAP) technique can
+ be used with both of these kernels to deal with the session
+ variability problem. We demonstrate the importance of GMM
+ model normalization (M-Norm) especially for the non linear
+ kernel. All our experiments were performed on the core
+ condition of NIST 2006 speaker recognition evaluation (all
+ trials). Our best results (an equal error rate of 6.3\%)
+ were obtained using NAP and GMM model normalization with
+ the non linear kernel.}
+}
+
+@InProceedings{ dehak.08.nist,
+ author = {R\'eda Dehak and Najim Dehak and Patrick Kenny},
+ title = {The {LRDE} Systems for the 2008 {NIST} Speaker Recognition
+ Evaluation},
+ booktitle = {NIST-SRE 2008},
+ year = 2008,
+ address = {Montr\'eal, Canada},
+ month = jun
+}
+
+@InProceedings{ dehak.08.odysseya,
+ author = {R\'eda Dehak and Najim Dehak and Patrick Kenny and Pierre
+ Dumouchel},
+ title = {Kernel Combination for {SVM} Speaker Verification},
+ booktitle = {Proceedings of the Speaker and Language Recognition
+ Workshop (IEEE-Odyssey 2008)},
+ year = 2008,
+ address = {Stellenbosch, South Africa},
+ month = jan,
+ urllrde = {200709-ODYSSEY-A},
+ abstract = {We present a new approach for constructing the kernels
+ used to build support vector machines for speaker
+ verification. The idea is to construct new kernels by
+ taking linear combination of many kernels such as the GLDS
+ and GMM supervector kernels. In this new kernel
+ combination, the combination weights are speaker dependent
+ rather than universal weights on score level fusion and
+ there is no need for extra-data to estimate them. An
+ experiment on the NIST 2006 speaker recognition evaluation
+ dataset (all trial) was done using three different kernel
+ functions (GLDS kernel, linear and Gaussian GMM supervector
+ kernels). We compared our kernel combination to the optimal
+ linear score fusion obtained using logistic regression.
+ This optimal score fusion was trained on the same test
+ data. We had an equal error rate of $\simeq 5,9\%$ using
+ the kernel combination technique which is better than the
+ optimal score fusion system ($\simeq 6,0\%$).}
+}
+
+@InProceedings{ dehak.08.odysseyb,
+ author = {Najim Dehak and R\'eda Dehak and Patrick Kenny and Pierre
+ Dumouchel},
+ title = {Comparison Between Factor Analysis and {GMM} Support
+ Vector Machines for Speaker Verification},
+ booktitle = {Proceedings of the Speaker and Language Recognition
+ Workshop (IEEE-Odyssey 2008)},
+ year = 2008,
+ address = {Stellenbosch, South Africa},
+ month = jan,
+ urllrde = {200709-ODYSSEY-B},
+ abstract = {We present a comparison between speaker verification
+ systems based on factor analysis modeling and support
+ vector machines using GMM supervectors as features. All
+ systems used the same acoustic features and they were
+ trained and tested on the same data sets. We test two types
+ of kernel (one linear, the other non-linear) for the GMM
+ support vector machines. The results show that factor
+ analysis using speaker factors gives the best results on
+ the core condition of the NIST 2006 speaker recognition
+ evaluation. The difference is particularly marked on the
+ English language subset. Fusion of all systems gave an
+ equal error rate of 4.2\% (all trials) and 3.2\% (English
+ trials only).}
+}
+
+@InProceedings{ dehak.09.icassp,
+ author = {Najim Dehak and Patrick Kenny and R\'eda Dehak and Ondrej
+ Glember and Pierre Dumouchel and Lukas Burget and
+ Valiantsina Hubeika and Fabio Castaldo},
+ title = {Support Vector Machines and Joint Factor Analysis for
+ Speaker Verification},
+ booktitle = {IEEE-ICASSP},
+ year = 2009,
+ address = {Taipei - Taiwan},
+ month = apr,
+ urllrde = {200904-ICASSP},
+ abstract = {This article presents several techniques to combine
+ between Support vector machines (SVM) and Joint Factor
+ Analysis (JFA) model for speaker verification. In this
+ combination, the SVMs are applied on different sources of
+ information produced by the JFA. These informations are the
+ Gaussian Mixture Model supervectors and speakers and Common
+ factors. We found that the use of JFA factors gave the best
+ results especially when within class covariance
+ normalization method is applied in the speaker factors
+ space, in order to compensate for the channel effect. The
+ new combination results are comparable to other classical
+ JFA scoring techniques.}
+}
+
+@InProceedings{ dehak.09.interspeech,
+ author = {Najim Dehak and R\'eda Dehak and Patrick Kenny and Niko
+ Brummer and Pierre Ouellet and Pierre Dumouchel},
+ title = {Support Vector Machines versus Fast Scoring in the
+ Low-Dimensional Total Variability Space for Speaker
+ Verification},
+ booktitle = {Interspeech},
+ year = 2009,
+ month = sep,
+ urllrde = {200909-INTERSPEECH-B},
+ abstract = {This paper presents a new speaker verification system
+ architecture based on Joint Factor Analysis (JFA) as
+ feature extractor. In this modeling, the JFA is used to
+ define a new low-dimensional space named the total
+ variability factor space, instead of both channel and
+ speaker variability spaces for the classical JFA. The main
+ contribution in this approach, is the use of the cosine
+ kernel in the new total factor space to design two
+ different systems: the first system is Support Vector
+ Machines based, and the second one uses directly this
+ kernel as a decision score. This last scoring method makes
+ the process faster and less computation complex compared to
+ others classical methods. We tested several intersession
+ compensation methods in total factors, and we found that
+ the combination of Linear Discriminate Analysis and Within
+ Class Covariance Normalization achieved the best
+ performance.}
+}
+
+@InProceedings{ dehak.09.interspeechb,
+ author = {Pierre Dumouchel and Najim Dehak and Yazid Attabi and
+ R\'eda Dehak and Narj\`es Boufaden},
+ title = {Cepstral and Long-Term Features for Emotion Recognition},
+ booktitle = {Interspeech},
+ year = 2009,
+ month = sep,
+ note = {Open Performance Sub-Challenge Prize},
+ urllrde = {200909-INTERSPEECH-A},
+ abstract = {In this paper, we describe systems that were developed for
+ the Open Performance Sub-Challenge of the INTERSPEECH 2009
+ Emotion Challenge. We participate to both two-class and
+ five-class emotion detection. For the two-class problem,
+ the best performance is obtained by logistic regression
+ fusion of three systems. Theses systems use short- and
+ long-term speech features. This fusion achieved an absolute
+ improvement of 2,6\% on the unweighted recall value
+ compared with [6]. For the five-class problem, we submitted
+ two individual systems: cepstral GMM vs. long-term GMM-UBM.
+ The best result comes from a cepstral GMM and produced an
+ absolute improvement of 3,5\% compared to [6].}
+}
+
+@InProceedings{ dehak.10.nist,
+ author = {R. Dehak and N.Dehak},
+ title = {LRDE {S}peaker {R}ecognition {S}ystem for {NIST-SRE}
+ 2010},
+ booktitle = {NIST 2010 Speaker Recognition Evaluation},
+ year = 2010,
+ address = {Brno, CZ}
+}
+
+@InProceedings{ dehak.10.odyssey,
+ author = {N. Dehak and R. Dehak and J. Glass and D. Reynolds and P.
+ Kenny},
+ title = {Cosine {S}imilarity {S}coring without {S}core
+ {N}ormalization {T}echniques},
+ booktitle = {Odyssey The Speaker and Language Recognition},
+ year = 2010,
+ address = {Brno, Czech Republic}
+}
+
+@InProceedings{ dehak.10.odyssey2,
+ author = {S. Shum and N. Dehak and R. Dehak and J. Glass},
+ title = {Unsupervised {S}peaker {A}daptation based on the {C}osine
+ {S}imilarity for {T}ext-{I}ndependent {S}peaker
+ {V}erification},
+ booktitle = {Odyssey The Speaker and Language Recognition},
+ year = 2010,
+ address = {Brno, Czech Republic}
+}
+
+@InProceedings{ dehak.11.icassp,
+ author = {N. Dehak and Z. Karam and D. Reynolds and R. Dehak and W.
+ Campbell and J. Glass},
+ title = {A {C}hannel-{B}lind {S}ystem for {S}peaker
+ {V}erification},
+ booktitle = {International Conference on Acoustics, Speech and Signal
+ Processing (ICASSP)},
+ pages = {4536--4539},
+ year = 2011,
+ address = {Prage, Czech Republic},
+ month = may
+}
+
+@InProceedings{ dehak.11.interspeech,
+ author = {Najim Dehak and Pedro A. Torres-Carrasquillo and Douglas
+ Reynolds and Reda Dehak},
+ title = {Language {R}ecognition via {I}-{V}ectors and
+ {D}imensionality {R}eduction},
+ booktitle = {INTERSPEECH 2011},
+ pages = {857--860},
+ year = 2011,
+ address = {Florence, Italy},
+ month = aug
+}
+
+@Article{ dehak.11.taslp,
+ author = {N. Dehak and P. Kenny and R. Dehak and P. Dumouchel and P.
+ Ouellet},
+ title = {Front-{E}nd {F}actor {A}nalysis {F}or {S}peaker
+ {V}erification},
+ year = 2011,
+ journal = {IEEE Transactions on Audio, Speech, and Language
+ Processing},
+ volume = 13,
+ number = 4,
+ pages = {788--798},
+ month = may
+}
+
+@InProceedings{ demaille.05.iticse,
+ author = {Akim Demaille},
+ title = {Making Compiler Construction Projects Relevant to Core
+ Curriculums},
+ booktitle = {Proceedings of the Tenth Annual Conference on Innovation
+ and Technology in Computer Science Education
+ ({ITICSE'05})},
+ year = 2005,
+ address = {Universidade Nova de {L}isboa, {M}onte da {P}acarita,
+ {P}ortugal},
+ month = jun,
+ project = {Tiger},
+ urllrde = {200506-ITICSE},
+ isbn = {1-59593-024-8},
+ pages = {266--270},
+ abstract = {Having 300 students a year implement a compiler is a
+ debatable enterprise, since the industry will certainly
+ \emph{not} recruit them for this competence. Yet we made
+ that decision five years ago, for reasons not related to
+ compiler construction. We detail these motivations, the
+ resulting compiler design, and how we manage the
+ assignment. The project meets its goals, since the majority
+ of former students invariably refer to it as \emph{the}
+ project that taught them the most.}
+}
+
+@InProceedings{ demaille.06.isola,
+ author = {Akim Demaille and Sylvain Peyronnet and Beno\^it Sigoure},
+ title = {Modeling of Sensor Networks Using {XRM}},
+ booktitle = {Proceedings of the 2nd International Symposium on
+ Leveraging Applications of Formal Methods, Verification and
+ Validation ({ISoLA'06})},
+ year = 2006,
+ address = {Coral Beach Resort, {P}aphos, {C}yprus},
+ month = nov,
+ project = {Transformers},
+ abstract = {Sensor networks are composed of small electronic devices
+ that embed processors, sensors, batteries, memory and
+ communication capabilities. One of the main goal in the
+ design of such systems is the handling of the inherent
+ complexity of the nodes, strengthened by the huge number of
+ nodes in the network. For these reasons, it becomes very
+ difficult to model and verify such systems. In this paper,
+ we investigate the main characteristics of sensor nodes,
+ discuss about the use of a language derived from Reactive
+ Modules for their modeling and propose a language (and a
+ tool set) that ease the modeling of this kind of systems.},
+ urllrde = {200609-ISOLA}
+}
+
+@InProceedings{ demaille.06.rivf,
+ author = {Akim Demaille and Sylvain Peyronnet and {\relax Th}omas
+ H\'erault},
+ title = {Probabilistic Verification of Sensor Networks},
+ booktitle = {Proceedings of the Fourth International Conference on
+ Computer Sciences, Research, Innovation and Vision for the
+ Future (RIVF'06)},
+ year = 2006,
+ address = {Ho Chi Minh City, Vietnam},
+ isbn = {1-4244-0316-2},
+ month = feb,
+ project = {APMC},
+ urllrde = {200602-rivf},
+ abstract = {Sensor networks are networks consisting of miniature and
+ low-cost systems with limited computation power and energy.
+ Thanks to the low cost of the devices, one can spread a
+ huge number of sensors into a given area to monitor, for
+ example, physical change of the environment. Typical
+ applications are in defense, environment, and design of
+ ad-hoc networks areas. In this paper, we address the
+ problem of verifying the correctness of such networks
+ through a case study. We modelize a simple sensor network
+ whose aim is to detect the apparition of an event in a
+ bounded area (such as a fire in a forest). The behaviour of
+ the network is probabilistic, so we use APMC, a tool that
+ allows to approximately check the correctness of extremely
+ large probabilistic systems, to verify it.}
+}
+
+@InProceedings{ demaille.08.fsmnlp,
+ author = {Akim Demaille and Alexandre Duret-Lutz and Florian Lesaint
+ and Sylvain Lombardy and Jacques Sakarovitch and Florent
+ Terrones},
+ title = {An {XML} format proposal for the description of weighted
+ automata, transducers, and regular expressions},
+ booktitle = {Proceedings of the seventh international workshop on
+ Finite-State Methods and Natural Language Processing
+ (FSMNLP'08)},
+ year = 2008,
+ address = {Ispra, Italia},
+ month = sep,
+ abstract = {We present an XML format that allows to describe a large
+ class of finite weighted automata and transducers. Our
+ design choices stem from our policy of making the
+ implementation as simple as possible. This format has been
+ tested for the communication between the modules of our
+ automata manipulation platform Vaucanson, but this document
+ is less an experiment report than a position paper intended
+ to open the discussion among the community of automata
+ software writers.},
+ urllrde = {200809-FSMNLP}
+}
+
+@InProceedings{ demaille.08.iticse,
+ author = {Akim Demaille and Roland Levillain and Beno\^it Perrot},
+ title = {A Set of Tools to Teach Compiler Construction},
+ booktitle = {Proceedings of the Thirteenth Annual Conference on
+ Innovation and Technology in Computer Science Education
+ ({ITICSE'08})},
+ pages = {68--72},
+ year = 2008,
+ address = {Universidad Polit\'ecnica de Madrid, Spain},
+ month = jun,
+ project = {Tiger},
+ urllrde = {200806-ITICSE},
+ abstract = {Compiler construction is a widely used software
+ engineering exercise, but because most students will not be
+ compiler writers, care must be taken to make it relevant in
+ a core curriculum. Auxiliary tools, such as generators and
+ interpreters, often hinder the learning: students have to
+ fight tool idiosyncrasies, mysterious errors, and other
+ poorly educative issues. We introduce a set of tools
+ especially designed or improved for compiler construction
+ educative projects in \Cxx. We also provide suggestions
+ about new approaches to compiler construction. We draw
+ guidelines from our experience to make tools suitable for
+ education purposes.}
+}
+
+@Misc{ demaille.08.kex,
+ author = {Akim Demaille and Roland Levillain},
+ title = {Compiler Construction as an Effective Application to Teach
+ Object-Oriented Programming},
+ howpublished = {The seventh ``Killer Examples'' workshop, Worked Examples
+ for Sound OO Pedagogy, at OOPSLA'08.},
+ address = {Nashville, USA},
+ month = oct,
+ year = 2008,
+ note = {Oral presentation},
+ project = {Tiger},
+ urllrde = {200810-KEX},
+ abstract = {Compiler construction, a course feared by most students,
+ and a competence seldom needed in the industry. Yet we
+ claim that compiler construction is wonderful topic that
+ benefits from virtually all the computer-science topics. In
+ this paper we show in particular why compiler construction
+ is a killer example for Object-Oriented Programming,
+ providing a unique opportunity for students to understand
+ what it is, what it can be used for, and how it works.}
+}
+
+@InProceedings{ demaille.08.ldta,
+ oldkeys = {durlin.08.seminar},
+ author = {Akim Demaille and Renaud Durlin and Nicolas Pierron and
+ Beno\^it Sigoure},
+ title = {{Semantics driven disambiguation: A comparison of
+ different approaches}},
+ booktitle = {Proceedings of the 8th workshop on Language Descriptions,
+ Tools and Applications (LDTA'08)},
+ year = 2008,
+ urllrde = {200802-LDTA},
+ abstract = {Context-sensitive languages such as \C or \Cxx can be
+ parsed using a context-free but ambiguous grammar, which
+ requires another stage, disambiguation, in order to select
+ the single parse tree that complies with the language's
+ semantical rules. Naturally, large and complex languages
+ induce large and complex disambiguation stages. If, in
+ addition, the parser should be extensible, for instance to
+ enable the embedding of domain specific languages, the
+ disambiguation techniques should feature traditional
+ software-engineering qualities: modularity, extensibility,
+ scalability and expressiveness. \\ We evaluate three
+ approaches to write disambiguation filters for \acs{sdf}
+ grammars: algebraic equations with \acs{asf}, rewrite-rules
+ with programmable traversals for \stratego, and attribute
+ grammars with \acr{tag}, our system. To this end we
+ introduce \phenix, a highly ambiguous language. Its
+ ``standard'' grammar exhibits ambiguities inspired by those
+ found in the \C and \Cxx standard grammars. To evaluate
+ modularity, the grammar is layered: it starts with a small
+ core language, and several layers add new features, new
+ production rules, and new ambiguities.},
+ project = {Transformers},
+ keywords = {Transformers, context-free grammar, attribute grammar,
+ Stratego, ASF, SDF, disambiguation, parsing, program
+ transformation, term rewriting}
+}
+
+@InProceedings{ demaille.09.sac,
+ author = {Akim Demaille and Roland Levillain and Beno\^it Sigoure},
+ title = {{TWEAST}: A Simple and Effective Technique to Implement
+ Concrete-Syntax {AST} Rewriting Using Partial Parsing},
+ booktitle = {Proceedings of the 24th Annual ACM Symposium on Applied
+ Computing (SAC'09)},
+ pages = {1924--1929},
+ year = 2009,
+ address = {Waikiki Beach, Honolulu, Hawaii, USA},
+ month = mar,
+ project = {Tiger},
+ urllrde = {200903-SAC},
+ abstract = {ASTs are commonly used to represent an input/output
+ program in compilers and language processing tools. Many of
+ the tasks of these tools consist in generating and
+ rewriting ASTs. Such an approach can become tedious and
+ hard to maintain for complex operations, namely program
+ transformation, optimization, instrumentation, etc. On the
+ other hand, \emph{concrete syntax} provides a natural and
+ simpler representation of programs, but it is not usually
+ available as a direct feature of the aforementioned tools.
+ We propose a simple technique to implement AST generation
+ and rewriting in general purpose languages using concrete
+ syntax. Our approach relies on extensions made in the
+ scanner and the parser and the use of objects supporting
+ partial parsing called Text With Embedded Abstract Syntax
+ Trees (TWEASTS). A compiler for a simple language (Tiger)
+ written in \Cxx serves as an example, featuring
+ transformations in concrete syntax: syntactic desugaring,
+ optimization, code instrumentation such as bounds-checking,
+ etc. Extensions of this technique to provide a full-fledged
+ concrete-syntax rewriting framework are presented as well.}
+}
+
+@InProceedings{ denise.06.rt,
+ author = {Alain Denise and Marie-Claude Gaudel and
+ Sandrine-Dominique Gouraud and Richard Lassaigne and
+ Sylvain Peyronnet},
+ title = {Uniform Random Sampling of Traces in Very Large Models},
+ booktitle = {Proceedings of the 1st international workshop on Random
+ Testing 2006 (RT06)},
+ year = 2006,
+ series = {ACM digital library},
+ pages = {10--19},
+ project = {APMC},
+ urllrde = {200606-RT},
+ abstract = {This paper presents some first results on how to perform
+ uniform random walks (where every trace has the same
+ probability to occur) in very large models. The models
+ considered here are described in a succinct way as a set of
+ communicating reactive modules. The method relies upon
+ techniques for counting and drawing uniformly at random
+ words in regular languages. Each module is considered as an
+ automaton defining such a language. It is shown how it is
+ possible to combine local uniform drawings of traces, and
+ to obtain some global uniform random sampling, without
+ construction of the global model.}
+}
+
+@Article{ dolstra.10.jfp,
+ author = {Eelco Dolstra and Andres L\"oh and Nicolas Pierron},
+ title = {{NixOS}: A purely functional {Linux} distribution},
+ journal = {Journal of Functional Programming},
+ year = 2010,
+ note = {Published online by Cambridge University Press 15 Oct
+ 2010},
+ urllrde = {201010-JFP},
+ doi = {10.1017/S0956796810000195}
+}
+
+@InProceedings{ duflot.04.avocs,
+ author = {Marie Duflot and Laurent Fribourg and {\relax Th}omas
+ Herault and Richard Lassaigne and Fr\'ed\'eric Magniette
+ and Stephane Messika and Sylvain Peyronnet and Claudine
+ Picaronny},
+ title = {Probabilistic model checking of the {CSMA/CD}, protocol
+ using {PRISM} and {APMC}},
+ booktitle = {Proceedings of the 4th International Workshop on Automated
+ Verification of Critical Systems (AVoCS)},
+ year = 2004,
+ series = {Electronic Notes in Theoretical Computer Science Series},
+ volume = 128,
+ issue = 6,
+ pages = {195--214},
+ project = {APMC},
+ urllrde = {200409-AVOCS},
+ abstract = {Carrier Sense Multiple Access/Collision Detection
+ (CSMA/CD) is the protocol for carrier transmission access
+ in Ethernet networks (international standard IEEE 802.3).
+ On Ethernet, any Network Interface Card (NIC) can try to
+ send a packet in a channel at any time. If another NIC
+ tries to send a packet at the same time, a collision is
+ said to occur and the packets are discarded. The CSMA/CD
+ protocol was designed to avoid this problem, more precisely
+ to allow a NIC to send its packet without collision. This
+ is done by way of a randomized exponential backoff process.
+ In this paper, we analyse the correctness of the CSMA/CD
+ protocol, using techniques from probabilistic model
+ checking and approximate probabilistic model checking. The
+ tools that we use are PRISM and APMC. Moreover, we provide
+ a quantitative analysis of some CSMA/CD properties.}
+}
+
+@InCollection{ duflot.06.book,
+ author = {Marie Duflot and Marta Kwiatkowska and Gethin Norman and
+ Dave Parker and Sylvain Peyronnet and Claudine Picaronny
+ and Jeremy Sproston},
+ title = {Practical Application of Probabilistic Model Checking to
+ Communication Protocols},
+ booktitle = {FMICS Handbook on Industrial Critical Systems},
+ year = 2006,
+ editor = {Stefania Gnesi and Tiziana Margaria},
+ chapter = 7,
+ note = {To appear},
+ project = {APMC},
+ urllrde = {200600-BOOK}
+}
+
+@InProceedings{ duret.00.gcse,
+ author = {Alexandre Duret-Lutz},
+ title = {Olena: a component-based platform for image processing,
+ mixing generic, generative and {OO} programming},
+ booktitle = {Proceedings of the 2nd International Symposium on
+ Generative and Component-Based Software Engineering
+ (GCSE)---Young Researchers Workshop; published in
+ ``Net.ObjectDays2000''},
+ pages = {653--659},
+ year = 2000,
+ address = {Erfurt, Germany},
+ month = oct,
+ isbn = {3-89683-932-2},
+ project = {Olena},
+ urllrde = {200010-NetObjectDays},
+ abstract = {This paper presents Olena, a toolkit for programming and
+ designing image processing chains in which each processing
+ is a component. But since there exist many image types
+ (different structures such as 2D images, 3D images or
+ graphs, as well as different value types) the platform has
+ been designed with genericity and reusability in mind: each
+ component is written as a generic C++ procedure, \`a la
+ STL. Other libraries, such as Khoros [Kon94] have a
+ different approach where a processing component contains an
+ implementation for each type supported by the library. This
+ makes code maintenance hard and prevents easy addition of
+ new image types. Still, Olena is not only a generic
+ component library [Jaz95], it shall contain additional
+ tools such as a visual programming environment (VPE). Those
+ tools may be programmed in a classical object-oriented
+ fashion (using operation and inclusion polymorphism) which
+ may seems antagonist with the generic programming paradigm
+ used in the library. Section 2 outlines the architecture of
+ Olena and elaborates more on the design problems resulting
+ from the use of generic components. Section 3 presents the
+ solution chosen to address these problems.}
+}
+
+@InProceedings{ duret.01.ae,
+ author = {Alexandre Duret-Lutz},
+ title = {Expression templates in {A}da~95},
+ booktitle = {Proceedings of the 6th International Conference on
+ Reliable Software Technologies (Ada-Europe)},
+ year = 2001,
+ series = {Lecture Notes in Computer Science Series},
+ volume = 2043,
+ address = {Leuven, Belgium},
+ month = may,
+ publisher = {Springer-Verlag},
+ pages = {191--202},
+ note = {Best Paper Award.},
+ project = {Software},
+ urllrde = {200105-Ae},
+ abstract = {High-order matrix or vector expressions tend to be
+ penalized by the use of huge temporary variables.
+ Expression templates is a C++ technique which can be used
+ to avoid these temporaries, in a way that is transparent to
+ the user. We present an Ada adaptation of this technique
+ which - while not transparent - addresses the same
+ efficiency issue as the original. We make intensive use of
+ the signature idiom to combine packages together, and
+ discuss its importance in generic programming. Finally, we
+ express some concerns about generic programming in Ada.}
+}
+
+@InProceedings{ duret.01.coots,
+ author = {Alexandre Duret-Lutz and {\relax Th}ierry G\'eraud and
+ Akim Demaille},
+ title = {Generic design patterns in {C++}},
+ booktitle = {Proceedings of the 6th USENIX Conference on
+ Object-Oriented Technologies and Systems (COOTS)},
+ year = 2001,
+ address = {San Antonio, TX, USA},
+ pages = {189--202},
+ month = {January-February},
+ publisher = {USENIX Association},
+ project = {Software},
+ urllrde = {200102-Coots},
+ abstract = {Generic programming is a paradigm whose wide adoption by
+ the C++ community is quite recent. In this approach most
+ classes and procedures are parameterized, leading to the
+ construction of general and efficient software components.
+ In this paper, we show how some design patterns from Gamma
+ et al. can be adapted to this paradigm. Although these
+ patterns rely highly on dynamic binding. We show that, by
+ making intensive use of parametric polymorphism in the
+ context of generic programming, the method calls in these
+ patterns can be resolved at compile-time. The speed-up
+ achieved using these patterns is significant.}
+}
+
+@InProceedings{ duret.09.atva,
+ author = {Alexandre Duret-Lutz and Denis Poitrenaud and Jean-Michel
+ Couvreur},
+ title = {On-the-fly Emptiness Check of Transition-based {S}treett
+ Automata},
+ booktitle = {Proceedings of the 7th International Symposium on
+ Automated Technology for Verification and Analysis
+ (ATVA'09)},
+ year = 2009,
+ editor = {Zhiming Liu and Anders P. Ravn},
+ series = {Lecture Notes in Computer Science},
+ publisher = {Springer-Verlag},
+ pages = {213--227},
+ volume = 5799,
+ abstract = {In the automata theoretic approach to model checking,
+ checking a state-space $S$ against a linear-time property
+ $\varphi$ can be done in $\RO(|S|\times
+ 2^{\RO(|\varphi|)})$ time. When model checking under $n$
+ strong fairness hypotheses expressed as a Generalized
+ B\"uchi automaton, this complexity becomes $\RO(|S|\times
+ 2^{\RO(|\varphi|+n)})$.\par Here we describe an algorithm
+ to check the emptiness of Streett automata, which allows
+ model checking under $n$ strong fairness hypotheses in
+ $\RO(|S|\times 2^{\RO(|\varphi|)}\times n)$. We focus on
+ transition-based Streett automata, because it allows us to
+ express strong fairness hypotheses by injecting Streett
+ acceptance conditions into the state-space without any blowup.},
+ urllrde = {200910-ATVA}
+}
+
+@TechReport{ duret.11.arxiv,
+ author = {Alexandre Duret-Lutz and Kais Klai and Denis Poitrenaud
+ and Yann Thierry-Mieg},
+ title = {Combining Explicit and Symbolic Approaches for Better
+ On-the-Fly {LTL} Model Checking},
+ institution = {arXiv},
+ year = 2011,
+ number = {1106.5700},
+ month = jun,
+ note = {Extended version of our ATVA'11 paper, presenting two new
+ techniques instead of one.},
+ url = {http://arxiv.org/abs/1106.5700},
+ urllrde = {201106-ARXIV},
+ abstract = {We present two new hybrid techniques that replace the
+ synchronized product used in the automata-theoretic
+ approach for LTL model checking. The proposed products are
+ explicit graphs of aggregates (symbolic sets of states)
+ that can be interpreted as B\"uchi automata. These hybrid
+ approaches allow on the one hand to use classical
+ emptiness-check algorithms and build the graph on-the-fly,
+ and on the other hand, to have a compact encoding of the
+ state space thanks to the symbolic representation of the
+ aggregates. The \emph{Symbolic Observation Product} assumes
+ a globally stuttering property (e.g., LTL$\setminus \X$) to
+ aggregate states. The \emph{Self-Loop Aggregation Product}
+ does not require the property to be globally stuttering
+ (i.e., it can tackle full LTL), but dynamically detects and
+ exploits a form of stuttering where possible. Our
+ experiments show that these two variants, while
+ incomparable with each other, can outperform other existing
+ approaches.}
+}
+
+@InProceedings{ duret.11.atva,
+ author = {Alexandre Duret-Lutz and Kais Klai and Denis Poitrenaud
+ and Yann Thierry-Mieg},
+ title = {Self-Loop Aggregation Product --- A New Hybrid Approach to
+ On-the-Fly {LTL} Model Checking},
+ booktitle = {Proceedings of the 9th International Symposium on
+ Automated Technology for Verification and Analysis
+ (ATVA'11)},
+ year = 2011,
+ series = {Lecture Notes in Computer Science},
+ volume = {6996},
+ pages = {336--350},
+ address = {Taipei, Taiwan},
+ month = oct,
+ publisher = {Springer},
+ urllrde = {201110-ATVA},
+ abstract = {We present the \emph{Self-Loop Aggregation Product}
+ (SLAP), a new hybrid technique that replaces the
+ synchronized product used in the automata-theoretic
+ approach for LTL model checking. The proposed product is an
+ explicit graph of aggregates (symbolic sets of states) that
+ can be interpreted as a B\"uchi automata. The criterion
+ used by SLAP to aggregate states from the Kripke structure
+ is based on the analysis of self-loops that occur in the
+ B\"uchi automaton expressing the property to verify. Our
+ hybrid approach allows on the one hand to use classical
+ emptiness-check algorithms and build the graph on-the-fly,
+ and on the other hand, to have a compact encoding of the
+ state space thanks to the symbolic representation of the
+ aggregates. Our experiments show that this technique often
+ outperforms other existing (hybrid or fully symbolic)
+ approaches.}
+}
+
+@Misc{ duret.11.sumo,
+ author = {Alexandre Duret-Lutz},
+ title = {Building {LTL} Model Checkers using {T}ransition-based
+ {G}eneralized {B\"u}chi {A}utomata},
+ howpublished = {Invited talk to SUMo'11},
+ month = jun,
+ year = 2011
+}
+
+@InProceedings{ duret.11.vecos,
+ author = {Alexandre Duret-Lutz},
+ title = {{LTL} Translation Improvements in {Spot}},
+ booktitle = {Proceedings of the 5th International Workshop on
+ Verification and Evaluation of Computer and Communication
+ Systems (VECoS'11)},
+ year = {2011},
+ series = {Electronic Workshops in Computing},
+ address = {Tunis, Tunisia},
+ month = sep,
+ publisher = {British Computer Society},
+ abstract = {Spot is a library of model-checking algorithms. This paper
+ focuses on the module translating LTL formul{\ae} into
+ automata. We discuss improvements that have been
+ implemented in the last four years, we show how Spot's
+ translation competes on various benchmarks, and we give
+ some insight into its implementation.},
+ url = {http://ewic.bcs.org/category/15853},
+ urllrde = {201109-VECOS}
+}
+
+@InProceedings{ fabre.00.egve,
+ author = {Yoann Fabre and Guillaume Pitel and Laurent Soubrevilla
+ and Emmanuel Marchand and {\relax Th}ierry G\'eraud and
+ Akim Demaille},
+ title = {An asynchronous architecture to manage communication,
+ display, and user interaction in distributed virtual
+ environments},
+ booktitle = {Virtual Environments 2000, Proceedings of the 6th
+ Eurographics Workshop on Virtual Environments (EGVE)},
+ year = 2000,
+ address = {Amsterdam, The Netherlands},
+ month = jun,
+ pages = {105--113},
+ series = {Computer Science / Eurographics Series},
+ editor = {J.D. Mulder and R. van Liere},
+ publisher = {Springer-Verlag WienNewYork},
+ project = {Urbi},
+ urllrde = {200006-Egve},
+ abstract = {In Distributed Virtual Environments, each machine runs the
+ same software, which is in charge of handling the
+ communications over the network, providing the user with a
+ view of the world, and processing his requests. A major
+ issue in the design of such a software is to ensure that
+ network communication does not degrade the interactivity
+ between the machine and the user. In this paper, we present
+ a software designed to achieve this goal, based on tools
+ rarely used in this area.}
+}
+
+@InProceedings{ fabre.00.vsmm,
+ author = {Yoann Fabre and Guillaume Pitel and Didier Verna},
+ title = {Urbi et {O}rbi: unusual design and implementation choices
+ for distributed virtual environments},
+ booktitle = {Proceedings of the 6th International Conference on Virtual
+ Systems and MultiMedia (VSMM)---Intelligent Environments
+ Workshop},
+ pages = {714--724},
+ year = 2000,
+ address = {Gifu, Japan},
+ month = oct,
+ publisher = {IOS Press, USA},
+ isbn = {1-58603-108-2},
+ project = {Urbi},
+ urllrde = {200010-Vsmm},
+ abstract = {This paper describes Urbi et Orbi, a distributed virtual
+ environment (DVE) project that is being conducted in the
+ Research and Development Laboratory at EPITA. Our ultimate
+ goal is to provide support for large scale multi-user
+ virtual worlds on end-user machines. The incremental
+ development of this project led us to take unusual design
+ and implementation decisions that we propose to relate in
+ this paper. Firstly, a general overview of the project is
+ given, along with the initial requirements we wanted to
+ meet. Then, we go on with a description of the system's
+ architecture. Lastly, we describe and justify the unusual
+ choices we have made in the project's internals.}
+}
+
+@InProceedings{ fabre.00.vw,
+ author = {Yoann Fabre and Guillaume Pitel and Laurent Soubrevilla
+ and Emmanuel Marchand and {\relax Th}ierry G\'eraud and
+ Akim Demaille},
+ title = {A framework to dynamically manage distributed virtual
+ environments},
+ booktitle = {Proceedings of the 2nd International Conference on Virtual
+ Worlds (VW)},
+ year = 2000,
+ address = {Paris, France},
+ month = jul,
+ pages = {54--64},
+ editor = {J.-C. Heudin},
+ publisher = {Springer Verlag},
+ series = {Lecture Notes in Computer Science Series},
+ volume = {LNAI 1834},
+ project = {Urbi},
+ urllrde = {200007-VW},
+ abstract = {In this paper, we present the project urbi, a framework to
+ dynamically manage distributed virtual environments (DVEs).
+ This framework relies on a dedicated scripting language,
+ goal, which is typed, object-oriented and dynamically
+ bound. goal is interpreted by the application hosted by
+ each machine and is designed to handle efficiently both
+ network communications and interactivity. Finally, we have
+ made an unusual design decision: our project is based on a
+ functional programming language, ocaml.}
+}
+
+@InProceedings{ fouquier.07.gbr,
+ author = {Geoffroy Fouquier and Jamal Atif and Isabelle Bloch},
+ title = {Local reasoning in fuzzy attribute graphs for optimizing
+ sequential segmentation},
+ booktitle = {Proceedings of the 6th IAPR TC-15 Workshop on Graph-based
+ Representations in Pattern Recognition (GBR)},
+ year = 2007,
+ month = jun,
+ address = {Alicante, Spain},
+ volume = {LNCS 4538},
+ editor = {F. Escolano and M. Vento},
+ publisher = {Springer Verlag},
+ isbn = {978-3-540-72902-0},
+ pages = {138--147},
+ urllrde = {200706-GBR},
+ abstract = {Spatial relations play a crucial role in model-based image
+ recognition and interpretation due to their stability
+ compared to many other image appearance characteristics.
+ Graphs are well adapted to represent such information.
+ Sequential methods for knowledge-based recognition of
+ structures require to define in which order the structures
+ have to be recognized. We propose to address this problem
+ of order definition by developing algorithms that
+ automatically deduce sequential segmentation paths from
+ fuzzy spatial attribute graphs. As an illustration, these
+ algorithms are applied on brain image understanding.}
+}
+
+@InProceedings{ fouquier.07.icassp,
+ author = {Geoffroy Fouquier and Laurence Likforman and J\'er\^ome
+ Darbon and Bulent Sankur},
+ title = {The Biosecure Geometry-based System for Hand Modality},
+ booktitle = {Proceedings of the 32nd IEEE International Conference on
+ Acoustics, Speech, and Signal Processing (ICASSP)},
+ year = 2007,
+ month = apr,
+ address = {Honolulu, Hawaii, USA},
+ volume = {I},
+ pages = {801--804},
+ isbn = {1-4244-0728-1},
+ urllrde = {200704-ICASSP},
+ abstract = {We present an identification and authentification system
+ based on hand modality which is part of a reference system
+ for all modalities developed within the Biosecure
+ consortium. It relies on simple geometric features
+ extracted from hand boundary. The different steps of this
+ system are detailed, namely: pre-processing, feature
+ extraction and hand matching. This system has been tested
+ on the Biosecure hand database which consists of 4500 hand
+ images of 750 individuals. Results are detailed with
+ respect to different enrolment conditions such as
+ population size, enrolment size, and image resolution.}
+}
+
+@InProceedings{ geraud.00.europlop,
+ author = {{\relax Th}ierry G\'eraud and Alexandre Duret-Lutz},
+ title = {Generic programming redesign of patterns},
+ booktitle = {Proceedings of the 5th European Conference on Pattern
+ Languages of Programs (EuroPLoP)},
+ year = 2000,
+ month = jul,
+ address = {Irsee, Germany},
+ pages = {283--294},
+ editors = {M. Devos and A. R\"uping},
+ publisher = {UVK, Univ. Verlag, Konstanz},
+ project = {Software},
+ urllrde = {200007-EuroPlop},
+ abstract = {This pattern faces the problem of improving the
+ performances of design patterns when they are involved in
+ intensive algorithms. Generic programming is a paradigm in
+ which most classes and procedures are parameterized, thus
+ leading to the construction of general and efficient
+ software components. We demonstrate that some design
+ patterns from Gamma et al. can be translated into this
+ paradigm while handling operation polymorphism by
+ parametric polymorphism. We thus preserve their modularity
+ and reusability properties but we avoid the performance
+ penalty due to their dynamic behavior, which is a critical
+ issue in numerical computing.}
+}
+
+@InProceedings{ geraud.00.icpr,
+ author = {{\relax Th}ierry G\'eraud and Yoann Fabre and Alexandre
+ Duret-Lutz and Dimitri Papadopoulos-Orfanos and
+ Jean-Fran\c{c}ois Mangin},
+ title = {Obtaining genericity for image processing and pattern
+ recognition algorithms},
+ booktitle = {Proceedings of the 15th International Conference on
+ Pattern Recognition (ICPR)},
+ year = 2000,
+ month = sep,
+ address = {Barcelona, Spain},
+ volume = 4,
+ pages = {816--819},
+ publisher = {IEEE Computer Society},
+ project = {Olena},
+ urllrde = {200009-Icpr},
+ abstract = {Algorithm libraries dedicated to image processing and
+ pattern recognition are not reusable; to run an algorithm
+ on particular data, one usually has either to rewrite the
+ algorithm or to manually ``copy, paste, and modify''. This
+ is due to the lack of genericity of the programming
+ paradigm used to implement the libraries. In this paper, we
+ present a recent paradigm that allows algorithms to be
+ written once and for all and to accept input of various
+ types. Moreover, this total reusability can be obtained
+ with a very comprehensive writing and without significant
+ cost at execution, compared to a dedicated algorithm. This
+ new paradigm is called ``generic programming'' and is fully
+ supported by the C++ language. We show how this paradigm
+ can be applied to image processing and pattern recognition
+ routines. The perspective of our work is the creation of a
+ generic library.}
+}
+
+@InProceedings{ geraud.00.rfia,
+ author = {{\relax Th}ierry G\'eraud and Isabelle Bloch and Henri
+ Ma{\^\i}tre},
+ title = {Reconnaissance de structures c\'er\'ebrales \`a l'aide
+ d'un atlas et par fusion d'informations structurelles floues},
+ booktitle = {Actes du 12\`eme Congr\`es Francophone AFRIF-AFIA de
+ Reconnaissance des Formes et Intelligence Artificielle
+ (RFIA)},
+ year = 2000,
+ address = {Paris, France},
+ month = feb,
+ volume = 1,
+ pages = {287--295},
+ note = {EPITA as current address.},
+ category = {national},
+ project = {Image},
+ urllrde = {200002-RFIA},
+ abstract = {Nous proposons une proc\'edure automatique de
+ reconnaissance progressive des structures internes du
+ cerveau guid\'ee par un atlas anatomique. L'originalit\'e
+ de notre proc\'edure est multiple. D'une part, elle prend
+ en compte des informations structurelles sous la forme de
+ contraintes spatiales flexibles, en utilisant les
+ formalismes de la th\'eorie des ensembles flous et de la
+ fusion d'informations. D'autre part, le calcul de la
+ correspondance entre volume IRM et atlas que nous proposons
+ permet d'inf\'erer un champ de d\'eformations discret,
+ respectant des contraintes sur la surface des objets.
+ Enfin, le caract\`ere s\'equentiel de la proc\'edure permet
+ de s'appuyer sur la connaissance des objets d\'ej\`a
+ segment\'es pour acc\'eder \`a des objets dont l'obtention
+ est a priori de plus en plus difficile.}
+}
+
+@InProceedings{ geraud.01.ai,
+ author = {{\relax Th}ierry G\'eraud and Yoann Fabre and Alexandre
+ Duret-Lutz},
+ title = {Applying generic programming to image processing},
+ booktitle = {Proceedings of the IASTED International Conference on
+ Applied Informatics (AI)---Symposium on Advances in
+ Computer Applications},
+ year = 2001,
+ publisher = {ACTA Press},
+ editor = {M.H.~Hamsa},
+ address = {Innsbruck, Austria},
+ pages = {577--581},
+ month = feb,
+ project = {Olena},
+ urllrde = {200102-Ai},
+ abstract = {This paper presents the evolution of algorithms
+ implementation in image processing libraries and discusses
+ the limits of these implementations in terms of
+ reusability. In particular, we show that in C++, an
+ algorithm can have a general implementation; said
+ differently, an implementation can be generic, i.e.,
+ independent of both the input aggregate type and the type
+ of the data contained in the input aggregate. A total
+ reusability of algorithms can therefore be obtained;
+ moreover, a generic implementation is more natural and does
+ not introduce a meaningful additional cost in execution
+ time as compared to an implementation dedicated to a
+ particular input type.}
+}
+
+@InProceedings{ geraud.01.icip,
+ author = {{\relax Th}ierry G\'eraud and Pierre-Yves Strub and
+ J\'er\^ome Darbon},
+ title = {Color image segmentation based on automatic morphological
+ clustering},
+ booktitle = {Proceedings of the IEEE International Conference on Image
+ Processing (ICIP)},
+ year = 2001,
+ volume = 3,
+ pages = {70--73},
+ address = {Thessaloniki, Greece},
+ month = oct,
+ project = {Image},
+ urllrde = {200110-Icip},
+ abstract = {We present an original method to segment color images
+ using a classification in the 3-D color space. In the case
+ of ordinary images, clusters that appear in 3-D histograms
+ usually do not fit a well-known statistical model. For that
+ reason, we propose a classifier that relies on mathematical
+ morphology, and more precisely on the watershed algorithm.
+ We show on various images that the expected color clusters
+ are correctly identified by our method. Last, to segment
+ color images into coherent regions, we perform a Markovian
+ labeling that takes advantage of the morphological
+ classification results.}
+}
+
+@InProceedings{ geraud.01.icisp,
+ author = {{\relax Th}ierry G\'eraud and Pierre-Yves Strub and
+ J\'er\^ome Darbon},
+ title = {Segmentation d'images en couleur par classification
+ morphologique non supervis\'ee},
+ booktitle = {Proceedings of the International Conference on Image and
+ Signal Processing (ICISP)},
+ year = 2001,
+ pages = {387--394},
+ address = {Agadir, Morocco},
+ month = may,
+ publisher = {Faculty of Sciences at Ibn Zohr University, Morocco},
+ note = {In French},
+ project = {Image},
+ urllrde = {200105-Icisp},
+ abstract = {In this paper, we present an original method to segment
+ color images using a classification of the image histogram
+ in the 3D color space. As color modes in natural images
+ usually do not fit a well-known statistical model, we
+ propose a classifier that rely on mathematical morphology
+ and, more particularly, on the watershed algorithm. We show
+ on various images that the expected color modes are
+ correctly identified and, in order to obtain coherent
+ region, we extend the method to make the segmentation
+ contextual.}
+}
+
+@InProceedings{ geraud.03.grec,
+ author = {{\relax Th}ierry G\'eraud and Geoffroy Fouquier and Quoc
+ Peyrot and Nicolas Lucas and Franck Signorile},
+ title = {Document type recognition using evidence theory},
+ booktitle = {Proceedings of the 5th IAPR International Workshop on
+ Graphics Recognition (GREC)},
+ year = 2003,
+ pages = {212--221},
+ editors = {Josep Llad\`os},
+ address = {Computer Vision Center, UAB, Barcelona, Spain},
+ month = jul,
+ project = {Image},
+ urllrde = {200307-Grec},
+ abstract = {This paper presents a method to recognize the type of a
+ document when a database of models (document types) is
+ given. For instance, when every documents are forms and
+ when we know every different types of forms, we want to be
+ able to assign to an input document its type of form. To
+ that aim, we define each model by a set of characteristics
+ whose nature can vary from one to another. For instance, a
+ characteristic can be having a flower-shaped logo on
+ top-left as well as having about 12pt fonts. This paper
+ does not intent to explain how to extract such knowledge
+ from documents but it describes how to use such information
+ to decide what the type of a given document is when
+ different document types are described by
+ characteristics.}
+}
+
+@InProceedings{ geraud.03.ibpria,
+ author = {{\relax Th}ierry G\'eraud},
+ title = {Segmentation of curvilinear objects using a
+ watershed-based curve adjacency graph},
+ booktitle = {Proceedings of the 1st Iberian Conference on Pattern
+ Recognition and Image Analysis (IbPRIA)},
+ pages = {279--286},
+ year = 2003,
+ editor = {Springer-Verlag},
+ volume = 2652,
+ series = {Lecture Notes in Computer Science Series},
+ address = {Mallorca, Spain},
+ month = jun,
+ publisher = {Springer-Verlag},
+ project = {Image},
+ urllrde = {200306-Ibpria},
+ abstract = {This paper presents a general framework to segment
+ curvilinear objects in 2D images. A pre-processing step
+ relies on mathematical morphology to obtain a connected
+ line which encloses curvilinear objects. Then, a graph is
+ constructed from this line and a Markovian Random Field is
+ defined to perform objects segmentation. Applications of
+ our framework are numerous: they go from simple surve
+ segmentation to complex road network extraction in
+ satellite images.}
+}
+
+@InProceedings{ geraud.03.icisp,
+ author = {{\relax Th}ierry G\'eraud},
+ title = {Segmentation d'objets curvilignes \`a l'aide des champs de
+ Markov sur un graphe d'adjacence de courbes issu de
+ l'algorithme de la ligne de partage des eaux},
+ booktitle = {Proceedings of the International Conference on Image and
+ Signal Processing (ICISP)},
+ year = 2003,
+ volume = 2,
+ pages = {404--411},
+ address = {Agadir, Morocco},
+ month = jun,
+ publisher = {Faculty of Sciences at Ibn Zohr University, Morocco},
+ note = {In French},
+ project = {Image},
+ urllrde = {200306-Icisp},
+ abstract = {This paper presents a general framework to segment
+ curvilinear objects in 2D images. A pre-processing step
+ relies on mathematical morphology to obtain a connected
+ line which encloses curvilinear objects. Then, a graph is
+ constructed from this line and a Markovian Random Field is
+ defined to perform objects segmentation. Applications of
+ our framework are numerous: they go from simple surve
+ segmentation to complex road network extraction in
+ satellite images.}
+}
+
+@InProceedings{ geraud.03.nsip,
+ author = {{\relax Th}ierry G\'eraud},
+ title = {Fast Road Network Extraction in Satellite Images using
+ Mathematical Morphology and {MRF}},
+ booktitle = {Proceedings of the EURASIP Workshop on Nonlinear Signal
+ and Image Processing (NSIP)},
+ year = 2003,
+ address = {Trieste, Italy},
+ month = jun,
+ project = {Image},
+ urllrde = {200306-Nsip},
+ abstract = {This paper presents a fast method to extract road network
+ in satellite images. A pre-processing stage relies on
+ mathematical morphology to obtain a connected line which
+ encloses road network. Then, a graph is constructed from
+ this line and a Markovian Random Field is defined to
+ perform road extraction.}
+}
+
+@InProceedings{ geraud.04.iccvg,
+ author = {{\relax Th}ierry G\'eraud and Giovanni Palma and Niels
+ {Van Vliet}},
+ title = {Fast color image segmentation based on levellings in
+ feature Space},
+ booktitle = {Computer Vision and Graphics---International Conference on
+ Computer Vision and Graphics (ICCVG), Warsaw, Poland,
+ September 2004},
+ year = 2004,
+ series = {Computational Imaging and Vision},
+ volume = 32,
+ editor = {Kluwer Academic Publishers},
+ pages = {800--807},
+ note = {On CD.},
+ project = {Image},
+ urllrde = {200408-ICCVG},
+ abstract = {This paper presents a morphological classifier with
+ application to color image segmentation. The basic idea of
+ a morphological classifier is to consider that a color
+ histogram is a 3D gray-level image and that morphological
+ operators can be applied to modify this image. The final
+ objective is to extract clusters in color space, that is,
+ identify regions in the 3D image. In this paper, we
+ particularly focus on a powerful class of morphology-based
+ filters called levellings to transform the 3D
+ histogram-image to identify clusters. We also show that our
+ method gives better results than the ones of
+ state-of-the-art methods.}
+}
+
+@Article{ geraud.04.jasp,
+ author = {{\relax Th}ierry G\'eraud and Jean-Baptiste Mouret},
+ title = {Fast road network extraction in satellite images using
+ mathematical morphology and {M}arkov random fields},
+ journal = {EURASIP Journal on Applied Signal Processing},
+ year = 2004,
+ number = 16,
+ volume = 2004,
+ pages = {2503--2514},
+ month = nov,
+ note = {Special issue on Nonlinear Signal and Image Processing -
+ Part II},
+ project = {Image},
+ doi = {http://doi.acm.org/10.1155/S1110865704409093},
+ urllrde = {200409-JASP},
+ abstract = {This paper presents a fast method for road network
+ extraction in satellite images. It can be seen as a
+ transposition of the segmentation scheme "watershed
+ transform + region adjacency graph + Markov random fields"
+ to the extraction of curvilinear objects. Many road
+ extractors can be found in the literature which are
+ composed of two stages. The first one acts like a filter
+ that can decide from a local analysis, at every image
+ point, if there is a road or not. The second stage aims at
+ obtaining the road network structure. In the method we
+ propose, we rely on a "potential" image, that is,
+ unstructured image data that can be derived from any road
+ extractor filter. In such a potential image, the value
+ assigned to a point is a measure of its likelihood to be
+ located in the middle of a road. A filtering step applied
+ on the potential image relies on the area closing operator
+ followed by the watershed transform to obtain a connected
+ line which encloses the road network. Then a graph
+ describing adjacency relationships between watershed lines
+ is built. Defining Markov random fields upon this graph,
+ associated with an energetic model of road networks, leads
+ to the expression of road network extraction as a global
+ energy minimization problem. This method can easily be
+ adapted to other image processing fields where the
+ recognition of curvilinear structures is involved.}
+}
+
+@InProceedings{ geraud.05.ismm,
+ author = {{\relax Th}ierry G\'eraud},
+ title = {Ruminations on {T}arjan's {U}nion-{F}ind algorithm and
+ connected operators},
+ booktitle = ismm05,
+ year = 2005,
+ address = {Paris, France},
+ month = apr,
+ pages = {105--116},
+ publisher = {Springer},
+ series = {Computational Imaging and Vision},
+ volume = 30,
+ project = {Image},
+ urllrde = {200504-ISMM},
+ abstract = {This papers presents a comprehensive and general form of
+ the Tarjan's union-find algorithm dedicated to connected
+ operators. An interesting feature of this form is to
+ introduce the notion of separated domains. The properties
+ of this form and its flexibility are discussed and
+ highlighted with examples. In particular, we give clues to
+ handle correctly the constraint of domain-disjointness
+ preservation and, as a consequence, we show how we can rely
+ on ``union-find'' to obtain algorithms for self-dual
+ filters approaches and levelings with a marker function.}
+}
+
+@InProceedings{ geraud.08.mpool,
+ author = {{\relax Th}ierry G\'eraud and Roland Levillain},
+ title = {Semantics-Driven Genericity: A Sequel to the Static {C++}
+ Object-Oriented Programming Paradigm ({SCOOP 2})},
+ booktitle = {Proceedings of the 6th International Workshop on
+ Multiparadigm Programming with Object-Oriented Languages
+ (MPOOL)},
+ year = 2008,
+ address = {Paphos, Cyprus},
+ month = jul,
+ project = {Olena},
+ urllrde = {200807-MPOOL},
+ abstract = {Classical (unbounded) genericity in \Cxx{}03 defines the
+ interactions between generic data types and algorithms in
+ terms of concepts. Concepts define the requirements over a
+ type (or a parameter) by expressing constraints on its
+ methods and dependent types (typedefs). The upcoming
+ \Cxx{}0x standard will promote concepts from abstract
+ entities (not directly enforced by the tools) to language
+ constructs, enabling compilers and tools to perform
+ additional checks on generic constructs as well as enabling
+ new features (e.g., concept-based overloading). Most modern
+ languages support this notion of signature on generic
+ types. However, generic types built on other types and
+ relying on concepts to both ensure type conformance and
+ drive code specialization, restrain the interface and the
+ implementation of the newly created type: specific methods
+ and associated types not mentioned in the concept will not
+ be part of the new type. The paradigm of concept-based
+ genericity lacks the required semantics to transform types
+ while retaining or adapting their intrinsic capabilities.
+ We present a new form of semantically-enriched genericity
+ allowing static generic type transformations through a
+ simple form of type introspection based on type metadata
+ called properties. This approach relies on a new Static
+ \Cxx Object-Oriented Programming (SCOOP) paradigm, and is
+ adapted to the creation of generic and efficient libraries,
+ especially in the field of scientific computing. Our
+ proposal uses a metaprogramming facility built into a \Cxx
+ library called Static, and doesn't require any language
+ extension nor additional processing (preprocessor,
+ transformation tool).}
+}
+
+@InCollection{ geraud.10.book,
+ author = {{\relax Th}ierry G\'eraud and Hugues Talbot and Marc Van
+ Droogenbroeck},
+ title = {Algorithms for Mathematical Morphology},
+ booktitle = {Mathematical Morphology---From Theory to Applications},
+ pages = {323--353},
+ publisher = {Wiley-ISTE},
+ year = 2010,
+ editor = {Laurent Najman and Hugues Talbot},
+ isbn = {978-1-84821-215-2},
+ month = jul,
+ url = {http://eu.wiley.com/WileyCDA/WileyTitle/productCd-1848212151.html}
+ ,
+ urllrde = {201007-BOOK}
+}
+
+@InCollection{ geraud.10.livre,
+ author = {{\relax Th}ierry G\'eraud and Hugues Talbot and Marc Van
+ Droogenbroeck},
+ title = {Morphologie et algorithmes},
+ booktitle = {Morphologie math\'ematique 2~: estimation, choix et mise
+ en {\oe}uvre},
+ pages = {151--180},
+ publisher = {Herm\`es Science Publications},
+ year = 2010,
+ series = {IC2 signal et image},
+ chapter = 6,
+ editor = {Laurent Najman and Hugues Talbot},
+ month = sep,
+ urllrde = {201009-LIVRE}
+}
+
+@InProceedings{ geraud.99.cimaf,
+ author = {{\relax Th}ierry G\'eraud and Isabelle Bloch and Henri
+ Ma\^tre},
+ title = {Atlas-guided recognition of cerebral structures in {MRI}
+ using fusion of fuzzy structural information},
+ booktitle = {Proceeding of CIMAF Symposium on Artificial Intelligence},
+ pages = {99--106},
+ year = 1999,
+ address = {La Havana, Cuba},
+ note = {EPITA as current address.},
+ project = {Image},
+ urllrde = {199900-CIMAF}
+}
+
+@InProceedings{ geraud.99.gretsi,
+ author = {{\relax Th}ierry G\'eraud and Yoann Fabre and Dimitri
+ Papadopoulos-Orfanos and Jean-Fran\c{c}ois Mangin},
+ title = {Vers une r\'eutilisabilit\'e totale des algorithmes de
+ traitement d'images},
+ booktitle = {Proceedings of the 17th Symposium on Signal and Image
+ Processing (GRETSI)},
+ category = {national},
+ pages = {331--334},
+ volume = 2,
+ year = 1999,
+ address = {Vannes, France},
+ month = sep,
+ note = {In French},
+ project = {Olena},
+ urllrde = {199909-Gretsi},
+ abstract = {Cet article pr\'esente l'\'evolution des techniques de
+ programmation d'algorithmes de traitement d'images et
+ discute des limites de la r\'eutilisabilit\'e de ces
+ algorithmes. En particulier, nous montrons qu'en C++ un
+ algorithme peut s'\'ecrire sous une forme g\'en\'erale,
+ ind\'ependante aussi bien du type des donn\'ees que du type
+ des structures de donn\'ees sur lesquelles il peut
+ s'appliquer. Une r\'eutilisabilit\'e totale des algorithmes
+ peut donc \^etre obtenue ; mieux, leur \'ecriture est plus
+ naturelle et elle n'introduit pas de surco\^ut significatif
+ en temps d'ex\'ecution.}
+}
+
+@Misc{ gournet.05.sud,
+ author = {Alexandre Borghi and Valentin David and Akim Demaille and
+ Olivier Gournet},
+ title = {Implementing attributes in {SDF}},
+ year = 2005,
+ address = {Utrecht {U}niversity, {N}etherland},
+ note = {Communication to Stratego Users Day 2005},
+ month = may,
+ project = {Transformers},
+ urllrde = {200505-SUD-ag},
+ abstract = {Attribute Grammars (AGs) provide a very convenient means
+ to bind semantics to syntax. They enjoy an extensive
+ bibliography and are used in several types of applications.
+ Yet, to our knowledge, their use to disambiguate is novel.
+ We present our implementation of an evaluator of attributes
+ for ambiguous AGs, tailored to ambiguous parse trees
+ disambiguation. This paper focuses on its implementation
+ that heavily relies on Stratego/XT, which is also used as
+ language to express the attribute rules. A companion paper
+ presents the disambiguation process in details
+ 200505-SUD-disamb.}
+}
+
+@InProceedings{ grosicki.04.icc,
+ author = {Emmanuel Grosicki and Karim Abed-Meraim and R\'eda Dehak},
+ title = {A novel method to fight the non line of sight error in
+ {AOA} measurements for mobile location},
+ booktitle = {Proceedings of the IEEE International Conference on
+ Communications (ICC)},
+ year = 2004,
+ volume = 5,
+ pages = {2794--2798},
+ address = {Paris, France},
+ month = jun,
+ project = {Image},
+ urllrde = {200406-ICC},
+ abstract = {In this contribution, a mobile location method is provided
+ using measurements from two different Base-Stations.
+ Although computationally from two different Base-Stations.
+ Although based on a simple trilateration and takes into
+ account error measurements caused by Non-Line-Of-Sight
+ (NLOS) and near-far effect. The new method attributes an
+ index of confidence for each measure, in order to allow the
+ mobile to select the two most reliable measures and not to
+ use all measures, equally.}
+}
+
+@InProceedings{ guirado.05.pdmc,
+ author = {Guillaume Guirado and {\relax Th}omas Herault and Richard
+ Lassaigne and Sylvain Peyronnet},
+ title = {Distribution, approximation and probabilistic model
+ checking},
+ booktitle = {Proceedings of the 4th international workshop on Parallel
+ and Distributed Model Checking (PDMC)},
+ year = 2005,
+ project = {APMC},
+ urllrde = {200507-Pdmc},
+ abstract = {APMC is a model checker dedicated to the quantitative
+ verification of fully probabilistic systems against LTL
+ formulas. Using a Monte-Carlo method in order to
+ efficiently approximate the verification of probabilistic
+ specifications, it could be used naturally in a distributed
+ framework. We present here the tool and his distribution
+ scheme, together with extensive performance evaluation,
+ showing the scalability of the method, even on clusters
+ containing 500+ heterogeneous workstations.}
+}
+
+@InProceedings{ hamez.07.pohll,
+ author = {Alexandre Hamez and Fabrice Kordon and Yann Thierry-Mieg},
+ title = {{libDMC}: a library to Operate Efficient Distributed Model
+ Checking},
+ booktitle = {Workshop on Performance Optimization for High-Level
+ Languages and Libraries --- associated to IPDPS'2007},
+ year = 2007,
+ project = {Verification},
+ urllrde = {200703-POHLL},
+ abstract = {Model checking is a formal verification technique that
+ allows to automatically prove that a system's behavior is
+ correct. However it is often prohibitively expensive in
+ time and memory complexity, due to the so-called state
+ space explosion problem. We present a generic
+ multi-threaded and distributed infrastructure library
+ designed to allow distribution of the model checking
+ procedure over a cluster of machines. This library is
+ generic, and is designed to allow encapsulation of any
+ model checker in order to make it distributed. Performance
+ evaluations are reported and clearly show the advantages of
+ multi-threading to occupy processors while waiting for the
+ network, with linear speedup over the number of
+ processors.}
+}
+
+@InProceedings{ hamez.08.atpn,
+ author = {Alexandre Hamez and Yann Thierry-Mieg and Fabrice Kordon},
+ title = {Hierarchical Set Decision Diagrams and Automatic
+ Saturation},
+ booktitle = {Petri Nets and Other Models of Concurrency --ICATPN 2008},
+ year = 2008,
+ project = {Verification},
+ urllrde = {200806-ATPN},
+ abstract = {Shared decision diagram representations of a state-space
+ have been shown to provide efficient solutions for
+ model-checking of large systems. However, decision diagram
+ manipulation is tricky, as the construction procedure is
+ liable to produce intractable intermediate structures
+ (a.k.a peak effect). The definition of the so-called
+ saturation method has empirically been shown to mostly
+ avoid this peak effect, and allows verification of much
+ larger systems. However, applying this algorithm currently
+ requires deep knowledge of the decision diagram
+ data-structures, of the model or formalism manipulated, and
+ a level of interaction that is not offered by the API of
+ public DD packages.
+
+ Hierarchical Set Decision Diagrams (SDD) are decision
+ diagrams in which arcs of the structure are labeled with
+ sets, themselves stored as SDD. This data structure offers
+ an elegant and very efficient way of encoding structured
+ specifications using decision diagram technology. It also
+ offers, through the concept of inductive homomorphisms,
+ unprecedented freedom to the user when defining the
+ transition relation. Finally, with very limited user input,
+ the SDD library is able to optimize evaluation of a
+ transition relation to produce a saturation effect at
+ runtime. We further show that using recursive folding, SDD
+ are able to offer solutions in logarithmic complexity with
+ respect to other DD. We conclude with some performances on
+ well known examples.}
+}
+
+@Article{ hamez.09.fi,
+ abstract = {Shared decision diagram representations of a state-space
+ provide efficient solutions for model-checking of large
+ systems. However, decision diagram manipulation is tricky,
+ as the construction procedure is liable to produce
+ intractable intermediate structures (a.k.a peak effect).
+ The definition of the so-called saturation method has
+ empirically been shown to mostly avoid this peak effect,
+ and allows verification of much larger systems. However,
+ applying this algorithm currently requires deep knowledge
+ of the decision diagram data-structures.
+
+ Hierarchical Set Decision Diagrams (SDD) are decision
+ diagrams in which arcs of the structure are labeled with
+ sets, themselves stored as SDD. This data structure offers
+ an elegant and very efficient way of encoding structured
+ specifications using decision diagram technology. It also
+ offers, through the concept of inductive homomorphisms,
+ flexibility to a user defining a transition relation. We
+ show in this paper how, with very limited user input, the
+ SDD library is able to optimize evaluation of a transition
+ relation to produce a saturation effect at runtime.
+
+ We build as an example an SDD model-checker for a
+ compositional formalism: Instantiable Petri Nets (IPN). IPN
+ define a \emph{type} as an abstract contract. Labeled P/T
+ nets are used as an elementary type. A composite type is
+ defined to hierarchically contain instances (of elementary
+ or composite type). To compose behaviors, IPN use classic
+ label synchronization semantics from process calculi.
+
+ With a particular recursive folding SDD are able to offer
+ solutions for symmetric systems in logarithmic complexity
+ with respect to other DD. Even in less regular cases, the
+ use of hierarchy in the specification is shown to be well
+ supported by SDD. Experimentations and performances are
+ reported on some well known examples. },
+ author = {Alexandre Hamez and Yann Thierry-Mieg and Fabrice Kordon},
+ date-added = {2009-05-06 16:39:07 +0200},
+ date-modified = {2009-05-06 16:48:10 +0200},
+ journal = {Fundamenta Informaticae},
+ title = {Building Efficient Model checkers using Hierarchical Set
+ Decision Diagrams and automatic Saturation},
+ year = 2009,
+ urllrde = {2009-FI}
+}
+
+@InProceedings{ hemon.08.sagt,
+ author = {S\'ebastien H\'emon and Michel de Rougemont and Miklos
+ Santha},
+ title = {Approximate {N}ash Equilibria for Multi-Player Games},
+ titre = {\'Equilibres de Nash approch\'es dans les jeux
+ multi-joueurs},
+ booktitle = {1st International Symposium on Algorithmic Games Theory},
+ year = 2008,
+ address = {Paderborn, Germany},
+ month = apr,
+ urllrde = {2008-01-18-SAGT},
+ resume = {Les \'equilibres de Nash sont des positions-cl\'es de tout
+ jeu admettant une repr\'esentation finie : en effet, quel
+ que soit le nombre de joueurs et de strat\'egies, une telle
+ position existe toujours. Lorsqu'elle est atteinte, elle
+ dissuade tout joueur de vouloir se d\'etourner de sa
+ strat\'egie actuelle, d'o\`u la notion d'\'equilibre. De
+ nombreux probl\`emes y font appel mais calculer de
+ fa\c{c}on effective l'\'equilibre demeure un probl\`eme
+ difficile. En effet, le meilleur algorithme connu pour,
+ dans le cas g\'en\'eral, calculer un \'equilibre est
+ exponentiel en le nombre de strat\'egies.
+
+ Nous pr\'esenterons ici la notion d'\'equilibres
+ approch\'es, et donnerons des r\'esultats concernant leur
+ calcul. Nous montrerons qu'il ne saurait exister
+ d'algorithmes pouvant calculer un \'equilibre, m\^eme
+ approch\'e, sans utiliser au moins, pour un joueur, un
+ nombre logarithmique de strat\'egies. Nous montrerons
+ comment calculer un \'equilibre approch\'e en temps
+ sub-exponentiel $n^{\mathcal{O}(\frac{\ln
+ n}{\varepsilon^2})}$, ce qui demeure actuellement, pour le
+ cas g\'en\'eral, la meilleure complexit\'e en pire cas.
+
+ Enfin, nous pr\'esenterons une approche inductive de
+ transfert d'approximation d'une position d'un jeu \`a deux
+ joueurs en une approximation pour un jeu \`a $r$ joueurs,
+ ce qui conf\`ere des r\'esultats novateurs dans le domaine.}
+}
+
+@InProceedings{ herault.06.qest,
+ author = {Thomas H\'erault and Richard Lassaigne and Sylvain
+ Peyronnet},
+ title = {{APMC 3.0}: Approximate verification of Discrete and
+ Continuous Time Markov Chains},
+ booktitle = {Proceedings of Qest 2006},
+ year = 2006,
+ pages = {129--130},
+ project = {APMC},
+ urllrde = {200606-Qest},
+ abstract = {In this paper, we give a brief overview of APMC
+ (Approximate Probabilistic Model Checker). APMC is a model
+ checker that implements approximate probabilistic
+ verification of probabilistic systems. It is based on
+ Monte-Carlo method and the theory of randomized
+ approximation schemes and allows to verify extremely large
+ models without explicitly representing the global
+ transition system. To avoid the state-space explosion
+ phenomenon, APMC gives an accurate approximation of the
+ satisfaction probability of the property instead of the
+ exact value, but using only a very small amount of memory.
+ The version of APMC we present in this paper can now handle
+ efficiently both discrete and continuous time probabilistic
+ systems.}
+}
+
+@InProceedings{ kenny.08.odyssey,
+ author = {Patrick Kenny and Najim Dehak and R\'eda Dehak and Vishwa
+ Gupta and Pierre Dumouchel},
+ title = {The Role of Speaker Factors in the {NIST} Extended Data
+ Task},
+ booktitle = {Proceedings of the Speaker and Language Recognition
+ Workshop (IEEE-Odyssey 2008)},
+ year = 2008,
+ address = {Stellenbosch, South Africa},
+ month = jan,
+ urllrde = {200709-ODYSSEY-C},
+ abstract = {We tested factor analysis models having various numbers of
+ speaker factors on the core condition and the extended data
+ condition of the 2006 NIST speaker recognition evaluation.
+ In order to ensure strict disjointness between training and
+ test sets, the factor analysis models were trained without
+ using any of the data made available for the 2005
+ evaluation. The factor analysis training set consisted
+ primarily of Switchboard data and so was to some degree
+ mismatched with the 2006 test data (drawn from the Mixer
+ collection). Consequently, our initial results were not as
+ good as those submitted for the 2006 evaluation. However we
+ found that we could compensate for this by a simple
+ modification to our score normalization strategy, namely by
+ using 1000 z-norm utterances in zt-norm. Our purpose in
+ varying the number of speaker factors was to evaluate the
+ eigenvoiceMAP and classicalMAP components of the
+ inter-speaker variability model in factor analysis. We
+ found that on the core condition (i.e. 2--3 minutes of
+ enrollment data), only the eigenvoice MAP component plays a
+ useful role. On the other hand, on the extended data
+ condition (i.e. 15--20 minutes of enrollment data) both the
+ classical MAP component and the eigenvoice component proved
+ to be useful provided that the number of speaker factors
+ was limited. Our best result on the extended data condition
+ (all trials) was an equal error rate of 2.2\% and a
+ detection cost of 0.011. }
+}
+
+@Article{ laplante.07.tocl,
+ author = {Sophie Laplante and Richard Lassaigne and Fr\'ed\'eric
+ Magniez and Sylvain Peyronnet and Michel de Rougemont},
+ title = {Probabilistic abstraction for model checking: an approach
+ based on property testing},
+ journal = {ACM Transactions on Computational Logic},
+ year = 2007,
+ project = {APMC},
+ month = aug,
+ volume = 8,
+ number = 4,
+ urllrde = {2006XX-TOCL},
+ abstract = {The goal of model checking is to verify the correctness of
+ a given program, on all its inputs. The main obstacle, in
+ many cases, is the intractably large size of the program's
+ transition system. Property testing is a randomized method
+ to verify whether some fixed property holds on individual
+ inputs, by looking at a small random part of that input. We
+ join the strengths of both approaches by introducing a new
+ notion of probabilistic abstraction, and by extending the
+ framework of model checking to include the use of these
+ abstractions. Our abstractions map transition systems
+ associated with large graphs to small transition systems
+ associated with small random subgraphs. This reduces the
+ original transition system to a family of small, even
+ constant-size, transition systems. We prove that with high
+ probability, ``sufficiently'' incorrect programs will be
+ rejected ($\eps$-robustness). We also prove that under a
+ certain condition (exactness), correct programs will never
+ be rejected (soundness). Our work applies to programs for
+ graph properties such as bipartiteness, $k$-colorability,
+ or any $\exists\forall$ first order graph properties. Our
+ main contribution is to show how to apply the ideas of
+ property testing to syntactic programs for such properties.
+ We give a concrete example of an abstraction for a program
+ for bipartiteness. Finally, we show that the relaxation of
+ the test alone does not yield transition systems small
+ enough to use the standard model checking method. More
+ specifically, we prove, using methods from communication
+ complexity, that the OBDD size remains exponential for
+ approximate bipartiteness.}
+}
+
+@InProceedings{ lassaigne.05.wollic,
+ author = {Richard Lassaigne and Sylvain Peyronnet},
+ title = {Probabilistic verification and approximation},
+ booktitle = {Proceedings of 12th Workshop on Logic, Language,
+ Information and Computation (Wollic)},
+ year = 2005,
+ series = {Electronic Notes in Theoretical Computer Science},
+ volume = 143,
+ pages = {101--114},
+ project = {APMC},
+ urllrde = {200507-Wollic},
+ abstract = {Model checking is an algorithmic method allowing to
+ automatically verify if a system which is represented as a
+ Kripke model satisfies a given specification.
+ Specifications are usually expressed by formulas of
+ temporal logic. The first objective of this paper is to
+ give an overview of methods able to verify probabilistic
+ systems. Models of such systems are labelled discrete time
+ Markov chains and specifications are expressed in
+ extensions of temporal logic by probabilistic operators.
+ The second objective is to focus on complexity of these
+ methods and to answer the question: can probabilistic
+ verification be efficiently approximated? In general, the
+ answer is negative. However, in many applications, the
+ specification formulas can be expressed in some positive
+ fragment of linear time temporal logic. In this paper, we
+ show how some simple randomized approximation algorithms
+ can improve the efficiency of the verification of such
+ probabilistic specifications.}
+}
+
+@InProceedings{ lazzara.11.icdar,
+ author = {Guillaume Lazzara and Roland Levillain and {\relax
+ Th}ierry G\'eraud and Yann Jacquelet and Julien Marquegnies
+ and Arthur Cr\'epin-Leblond},
+ title = {The {SCRIBO} Module of the {Olena} Platform: a Free
+ Software Framework for Document Image Analysis},
+ booktitle = {Proceedings of the 11th International Conference on
+ Document Analysis and Recognition (ICDAR)},
+ year = 2011,
+ address = {Beijing, China},
+ month = sep,
+ organization = {International Association for Pattern Recognition (IAPR)},
+ project = {Olena},
+ urllrde = {201109-ICDAR},
+ abstract = {Electronic documents are being more and more usable thanks
+ to better and more affordable network, storage and
+ computational facilities. But in order to benefit from
+ computer-aided document management, paper documents must be
+ digitized and analyzed. This task may be challenging at
+ several levels. Data may be of multiple types thus
+ requiring different adapted processing chains. The tools to
+ be developed should also take into account the needs and
+ knowledge of users, ranging from a simple graphical
+ application to a complete programming framework. Finally,
+ the data sets to process may be large. In this paper, we
+ expose a set of features that a Document Image Analysis
+ framework should provide to handle the previous issues. In
+ particular, a good strategy to address both flexibility and
+ efficiency issues is the Generic Programming (GP) paradigm.
+ These ideas are implemented as an open source module,
+ SCRIBO, built on top of Olena, a generic and efficient
+ image processing platform. Our solution features services
+ such as preprocessing filters, text detection, page
+ segmentation and document reconstruction (as XML, PDF or
+ HTML documents). This framework, composed of reusable
+ software components, can be used to create full-fledged
+ graphical applications, small utilities, or processing
+ chains to be integrated into third-party projects.},
+ keywords = {Document Image Analysis, Software Design, Reusability,
+ Free Software}
+}
+
+@InProceedings{ le-quoc.07.ntms,
+ author = {Cuong Le Quoc and Patrick Bellot and Akim Demaille},
+ title = {On the security of quantum networks: a proposal framework
+ and its capacity},
+ booktitle = {Proceedings of the 2007 International Conference on New
+ Technologies, Mobility and Security (NTMS'07)},
+ year = 2007,
+ address = {Paris, France},
+ month = may,
+ urllrde = {200705-NTMS},
+ abstract = {In large Quantum Key Distribution (QKD)-based networks,
+ intermediate nodes are necessary because of the short
+ length of QKD links. They have tendency to be used more
+ than classical networks. A realistic assumption is that
+ there are eavesdropping operations in these nodes without
+ knowledge of legitimate network participants. We develop a
+ QKD-based network framework. We present a percolation-based
+ approach to discuss about conditions of extremely high
+ secret key transmission. We propose also an adaptive
+ stochastic routing algorithm that helps on protecting keys
+ from reasonable eavesdroppers in a dense QKD network. We
+ show that under some assumptions, one could prevent
+ eavesdroppers from sniffing the secrets with an arbitrarily
+ large probability.}
+}
+
+@InProceedings{ le-quoc.07.rivf,
+ author = {Cuong Le Quoc and Patrick Bellot and Akim Demaille},
+ title = {Stochastic routing in large grid-shaped quantum networks},
+ booktitle = {Proceedings of the Fifth International Conference on
+ Computer Sciences, Research, Innovation and Vision for the
+ Future (RIVF'07)},
+ year = 2007,
+ address = {Hanoi, Vietnam},
+ month = mar,
+ isbn = {1-4244-0695-1},
+ urllrde = {200703-RIVF},
+ abstract = {This paper investigates the problem of secret key
+ transmissions for an arbitrary Alice-Bob pair in Quantum
+ Key Distribution-based networks. We develop a realistic
+ QKD-based network framework and we show that the key
+ transmission problem on such a framework can be considered
+ as a variant of the classical percolation problem. We also
+ present an adaptive stochastic routing algorithm protect
+ from inevitable eavesdroppers. Simulations were carried out
+ not only to validate our approach, but also to compute
+ critical parameters ensuring security. These results show
+ that large quantum networks with eavesdroppers do provide
+ security.},
+ keywords = {Quantum Key Distribution, QKD network, percolation theory,
+ stochastic routing}
+}
+
+@InProceedings{ le-quoc.08.ispec,
+ author = {Cuong Le Quoc and Patrick Bellot and Akim Demaille},
+ title = {Towards the World-Wide Quantum Network},
+ booktitle = {Proceedings of the 4th Information Security Practice and
+ Experience Conference (ISPEC'08)},
+ year = 2008,
+ address = {Sydney, Australia},
+ month = april,
+ urllrde = {200804-ISPEC},
+ abstract = {Quantum Key Distribution (QKD) networks are of much
+ interest due to their capacity of providing extremely high
+ security keys to network participants. Most QKD network
+ studies so far focus on trusted models where all the
+ network nodes are assumed to be perfectly secured. This
+ restricts QKD networks to be small. In this paper, we first
+ develop a novel model dedicated to large-scale QKD
+ networks, some of whose nodes could be eavesdropped
+ secretly. Then, we investigate the key transmission problem
+ in the new model by an approach based on percolation theory
+ and stochastic routing. Analyses show that under computable
+ conditions large-scale QKD networks could protect secret
+ keys with an extremely high probability. Simulations
+ validate our results.},
+ keywords = {Quantum Key Distribution, QKD network, percolation theory,
+ stochastic routing}
+}
+
+@TechReport{ lefebvre.04.tr,
+ author = {Sylvain Lefebvre and J\'er\^ome Darbon and Fabrice Neyret},
+ title = {Unified texture management for arbitrary meshes},
+ institution = {INRIA-Rhone-Alpes},
+ year = 2004,
+ number = {RR-5210},
+ address = {France},
+ month = may,
+ project = {Image},
+ urllrde = {200405-RRinria},
+ abstract = {Video games and simulators commonly use very detailed
+ textures, whose cumulative size is often larger than the
+ GPU memory. Textures may be loaded progressively, but
+ dynamically loading and transferring this large amount of
+ data in GPU memory results in loading delays and poor
+ performance. Therefore, managing texture memory has become
+ an important issue. While this problem has been (partly)
+ addressed early for the specific case of terrain rendering,
+ there is no generic texture management system for arbitrary
+ meshes. We propose such a system, implemented on today's
+ GPUs, which unifies classical solutions aimed at reducing
+ memory transfer: progressive loading, texture compression,
+ and caching strategies. For this, we introduce a new
+ algorithm -- running on GPU -- to solve the major
+ difficulty of detecting which parts of the texture are
+ required for rendering. Our system is based on three
+ components manipulating a tile pool which stores texture
+ data in GPU memory. First, the Texture Load Map determines
+ at every frame the appropriate list of texture tiles (i.e.
+ location and MIP-map level) to render from the current
+ viewpoint. Second, the Texture Cache manages the tile pool.
+ Finally, the Texture Producer loads and decodes required
+ texture tiles asynchronously in the tile pool. Decoding of
+ compressed texture data is implemented on GPU to minimize
+ texture transfer. The Texture Producer can also generate
+ procedural textures. Our system is transparent to the user,
+ and the only parameter that must be supplied at runtime is
+ the current viewpoint. No modifications of the mesh are
+ required. We demonstrate our system on large scenes
+ displayed in real time. We show that it achieves
+ interactive frame rates even in low-memory low-bandwidth
+ situations.}
+}
+
+@InProceedings{ lesage.06.isvc,
+ author = {David Lesage and J\'er\^ome Darbon and Ceyhun Burak Akg\"ul},
+ title = {An Efficient Algorithm for Connected Attribute Thinnings
+ and Thickenings},
+ booktitle = {Proceedings of the second International Conference on
+ Visual Computing},
+ year = 2006,
+ address = {Lake Tahoe, Nevada, USA},
+ month = nov,
+ project = {Image},
+ pages = {393--404},
+ volume = 4292,
+ series = {Lecture Notes in Computer Science Series},
+ publisher = {Springer-Verlag},
+ urllrde = {200611-ISVC},
+ abstract = {Connected attribute filters are anti-extensive
+ morphological operators widely used for their ability of
+ simplifying the image without moving its contours. In this
+ paper, we present a fast, versatile and easy-to-implement
+ algorithm for grayscale connected attribute thinnings and
+ thickennings, a subclass of connected filters for the wide
+ range of non-increasing attributes. We show that our
+ algorithm consumes less memory and is computationally more
+ efficient than other available methods on natural images.}
+}
+
+@Misc{ levillain.05.olenaposter,
+ author = {Roland Levillain},
+ title = {{O}lena {P}roject poster},
+ month = oct,
+ year = 2005,
+ urllrde = {200510-OlenaPoster}
+}
+
+@Misc{ levillain.05.tigerposter,
+ author = {Roland Levillain},
+ title = {{T}iger {P}roject poster},
+ month = oct,
+ year = 2005,
+ urllrde = {200510-TigerPoster}
+}
+
+@InProceedings{ levillain.09.ismm,
+ author = {Roland Levillain and {\relax Th}ierry G\'eraud and Laurent
+ Najman},
+ title = {{Milena}: Write Generic Morphological Algorithms Once, Run
+ on Many Kinds of Images},
+ booktitle = {Mathematical Morphology and Its Application to Signal and
+ Image Processing -- Proceedings of the Ninth International
+ Symposium on Mathematical Morphology (ISMM)},
+ pages = {295--306},
+ year = 2009,
+ editor = {Michael H. F. Wilkinson and Jos B. T. M. Roerdink},
+ series = {Lecture Notes in Computer Science},
+ address = {Groningen, The Netherlands},
+ month = aug,
+ publisher = {Springer Berlin / Heidelberg},
+ volume = 5720,
+ project = {Olena},
+ urllrde = {200908-ISMM},
+ abstract = {We present a programming framework for discrete
+ mathematical morphology centered on the concept of
+ genericity. We show that formal definitions of
+ morphological algorithms can be translated into actual
+ code, usable on virtually any kind of compatible images,
+ provided a general definition of the concept of image is
+ given. This work is implemented in Milena, a generic,
+ efficient, and user-friendly image processing library.},
+ keywords = {mathematical morphology, image processing operator,
+ genericity, programming}
+}
+
+@InProceedings{ levillain.10.icip,
+ author = {Roland Levillain and {\relax Th}ierry G\'eraud and Laurent
+ Najman},
+ title = {Why and How to Design a Generic and Efficient Image
+ Processing Framework: The Case of the {Milena} Library},
+ booktitle = {Proceedings of the IEEE International Conference on Image
+ Processing (ICIP)},
+ pages = {1941--1944},
+ year = 2010,
+ address = {Hong Kong},
+ month = sep,
+ project = {Olena},
+ urllrde = {201009-ICIP},
+ abstract = {Most image processing frameworks are not generic enough to
+ provide true reusability of data structures and algorithms.
+ In fact, genericity allows users to write and experiment
+ virtually any method on any compatible input(s). In this
+ paper, we advocate the use of generic programming in the
+ design of image processing software, while preserving
+ performances close to dedicated code. The implementation of
+ our proposal, Milena, a generic and efficient library,
+ illustrates the benefits of our approach.},
+ keywords = {Genericity, Image Processing, Software Design,
+ Reusability, Efficiency}
+}
+
+@InProceedings{ levillain.10.wadgmm,
+ author = {Roland Levillain and {\relax Th}ierry G\'eraud and Laurent
+ Najman},
+ title = {Writing Reusable Digital Geometry Algorithms in a Generic
+ Image Processing Framework},
+ booktitle = {Proceedings of the Workshop on Applications of Digital
+ Geometry and Mathematical Morphology (WADGMM)},
+ pages = {96--100},
+ year = 2010,
+ address = {Istanbul, Turkey},
+ month = aug,
+ url = {http://mdigest.jrc.ec.europa.eu/wadgmm2010/},
+ project = {Olena},
+ urllrde = {201008-WADGMM},
+ abstract = {Digital Geometry software should reflect the generality of
+ the underlying mathematics: mapping the latter to the
+ former requires genericity. By designing generic solutions,
+ one can effectively reuse digital geometry data structures
+ and algorithms. We propose an image processing framework
+ centered on the Generic Programming paradigm in which an
+ algorithm on the paper can be turn into a single code,
+ written once and usable with various input types. This
+ approach enables users to design and implement new methods
+ at a lower cost, try cross-domain experiments and help
+ generalize results.},
+ keywords = {Generic Programming, Interface, Skeleton, Complex}
+}
+
+@InProceedings{ levillain.11.gretsi,
+ author = {Roland Levillain and {\relax Th}ierry G\'eraud and Laurent
+ Najman},
+ title = {Une approche g\'en\'erique du logiciel pour le traitement
+ d'images pr\'eservant les performances},
+ booktitle = {Proceedings of the 23rd Symposium on Signal and Image
+ Processing (GRETSI)},
+ category = {national},
+ year = 2011,
+ address = {Bordeaux, France},
+ month = sep,
+ note = {In French.},
+ project = {Olena},
+ urllrde = {201109-GRETSI},
+ abstract = {De plus en plus d'outils logiciels modernes pour le
+ traitement d'images sont con\,c{}us en prenant en compte le
+ probl\`eme de la g\'en\'ericit\'e du code, c'est-\`a-dire
+ la possibilit\'e d'\'ecrire des algorithmes
+ r\'eutilisables, compatibles avec de nombreux types
+ d'entr\'ees. Cependant, ce choix de conception se fait
+ souvent au d\'etriment des performances du code
+ ex\'ecut\'e. Du fait de la grande vari\'et\'e des types
+ d'images existants et de la n\'ecessit\'e d'avoir des
+ impl\'ementations rapides, g\'en\'ericit\'e et performance
+ apparaissent comme des qualit\'es essentielles du logiciel
+ en traitement d'images. Cet article pr\'esente une approche
+ pr\'eservant les performances dans un framework logiciel
+ g\'en\'erique tirant parti des caract\'eristiques des types
+ de donn\'ees utilis\'es. Gr\^ace \`a celles-ci, il est
+ possible d'\'ecrire des variantes d'algorithmes
+ g\'en\'eriques offrant un compromis entre g\'en\'ericit\'e
+ et performance. Ces alternatives sont capables de
+ pr\'eserver une partie des aspects g\'en\'eriques d'origine
+ tout en apportant des gains substantiels \`a l'ex\'ecution.
+ D'apr\`es nos essais, ces optimisations g\'en\'eriques
+ fournissent des performances supportant la comparaison avec
+ du code d\'edi\'e, allant parfois m\^eme jusqu'\`a surpasser des routines optimis\'ees manuellement.}
+}
+
+@InProceedings{ linard.10.acsd,
+ author = {Alban Linard and Emmanuel Paviot-Adet and Fabrice Kordon
+ and Didier Buchs and Samuel Charron},
+ title = {{polyDD}: Towards a Framework Generalizing Decision
+ Diagrams},
+ booktitle = {Proceedings of the 10th International Conference on
+ Application of Concurrency to System Design (ACSD)},
+ pages = {124--133},
+ year = 2010,
+ address = {Braga, Portugal},
+ month = jun,
+ project = {Verification},
+ urllrde = {201006-ACSD},
+ publisher = {IEEE Computer Society},
+ abstract = {Decision Diagrams are now widely used in model checking as
+ extremely compact representations of state spaces. Many
+ Decision Diagram categories have been developed over the
+ past twenty years based on the same principles. Each one
+ targets a specific domain with its own characteristics.
+ Moreover, each one provides its own definition. It prevents
+ sharing concepts and techniques between these structures.
+ This paper aims to propose a basis for a common Framework
+ for Decision Diagrams. It should help users of this
+ technology to define new Decision Diagram categories thanks
+ to a simple specification mechanism called Controller. This
+ enables the building of efficient Decision Diagrams
+ dedicated to a given problem.}
+}
+
+@InProceedings{ lombardy.03.ciaa,
+ author = {Sylvain Lombardy and Rapha\"el Poss and Yann
+ R\'egis-Gianas and Jacques Sakarovitch},
+ title = {Introducing {V}aucanson},
+ booktitle = {Proceedings of Implementation and Application of Automata,
+ 8th International Conference (CIAA)},
+ pages = {96--107},
+ year = 2003,
+ editor = {Springer-Verlag},
+ volume = 2759,
+ series = {Lecture Notes in Computer Science Series},
+ address = {Santa Barbara, CA, USA},
+ month = jul,
+ project = {Vaucanson},
+ urllrde = {200307-Ciaa},
+ abstract = {This paper reports on a new software platform dedicated to
+ the computation with automata and transducers, called
+ Vaucanson, the main feature of which is the capacity of
+ dealing with automata whose labels may belong to various
+ algebraic structures. The paper successively shows how
+ Vaucanson allows to program algorithms on automata in a way
+ which is very close to the mathematical expression of the
+ algorithm, describes some features of the Vaucanson
+ platform, including the fact that the very rich data
+ structure used to implement automata does not weight too
+ much on the performance and finally explains the main
+ issues of the programming design that allow to achieve both
+ genericity and efficiency.}
+}
+
+@Article{ lombardy.04.tcs,
+ author = {Sylvain Lombardy and Yann R\'egis-{G}ianas and Jacques
+ Sakarovitch},
+ title = {Introducing {V}aucanson},
+ journal = {Theoretical Computer Science},
+ volume = 328,
+ year = 2004,
+ pages = {77--96},
+ month = nov,
+ project = {Vaucanson},
+ urllrde = {200411-TCS},
+ abstract = {This paper reports on a new software platform called
+ VAUCANSON and dedicated to the computation with automata
+ and transducers. Its main feature is the capacity of
+ dealing with automata whose labels may belong to various
+ algebraic structures. The paper successively describes the
+ main features of the VAUCANSON platform, including the fact
+ that the very rich data structure used to implement
+ automata does not weigh too much on the performance, shows
+ how VAUCANSON allows to program algorithms on automata in a
+ way which is very close to the mathematical expression of
+ the algorithm and finally explains the main choices of the
+ programming design that enable to achieve both genericity
+ and efficiency.}
+}
+
+@InProceedings{ maes.03.dpcool,
+ author = {Francis Maes},
+ title = {Program templates: expression templates applied to program
+ evaluation},
+ booktitle = {Proceedings of the Workshop on Declarative Programming in
+ the Context of Object-Oriented Languages (DP-COOL; in
+ conjunction with PLI)},
+ year = 2003,
+ address = {Uppsala, Sweden},
+ number = {FZJ-ZAM-IB-2003-10},
+ pages = {67--86},
+ editor = {J\"org Striegnitz and Kei Davis},
+ month = aug,
+ series = {John von Neumann Institute for Computing (NIC)},
+ project = {Software},
+ urllrde = {20020619-Seminar-Maes-Report},
+ abstract = {The C++ language provides a two-layer execution model:
+ static execution of meta-programs and dynamic execution of
+ resulting programs. The Expression Templates technique
+ takes advantage of this dual execution model through the
+ construction of C++ types expressing simple arithmetic
+ formulas. Our intent is to extend this technique to a whole
+ programming language. The Tiger language is a small,
+ imperative language with types, variables, arrays, records,
+ ow control structures and nested functions. The rst step is
+ to show how to express a Tiger program as a C++ type. The
+ second step concerns operational analysis which is done
+ through the use of meta-programs. Finally an implementation
+ of our Tiger evaluator is proposed. Our technique goes much
+ deeper than the Expression Templates one. It shows how the
+ generative power of C++ meta-programming can be used in
+ order to compile abstract syntax trees of a fully featured
+ programming language.}
+}
+
+@InProceedings{ maes.04.mpool,
+ author = {Francis Maes},
+ title = {Metagene, a {C++} meta-program generation tool},
+ booktitle = {Proceedings of the Workshop on Multiple Paradigm with OO
+ Languages (MPOOL; in conjunction with ECOOP)},
+ year = 2004,
+ address = {Oslo, Norway},
+ month = jun,
+ project = {Software},
+ urllrde = {200406-MPOOL},
+ abstract = {The C++ language offers a two layer evaluation model.
+ Thus, it is possible to evaluate a program in two steps:
+ the so-called static and dynamic evaluations. Static
+ evaluation is used for reducing the amount of work done at
+ execution-time. Programs executed statically (called
+ metaprograms) are written in C++ through an intensive use
+ of template classes. Due to the complexity of these
+ structures, writing, debugging and maintaining C++
+ meta-programs is a difficult task. Metagene is a program
+ transformation tool which simplifies the development of
+ such programs. Due to the similarities between C++
+ meta-programming and functional programming, the input
+ language of Metagene is an ML language. Given a functional
+ input program, Metagene outputs the corresponding C++
+ meta-program expressed using template classes.}
+}
+
+@InProceedings{ minetto.10.icip,
+ author = {Rodrigo Minetto and Nicolas Thome and Matthieu Cord and
+ Jonathan Fabrizio and Beatriz Marcotegui},
+ title = {SnooperText: A Multiresolution System for Text Detection
+ in Complex Visual Scenes},
+ booktitle = {Proceedings of the IEEE International Conference on Image
+ Processing (ICIP)},
+ pages = {3861--3864},
+ year = 2010,
+ address = {Hong Kong},
+ month = sep,
+ abstract = {Text detection in natural images remains a very
+ challenging task. For instance, in an urban context, the
+ detection is very difficult due to large variations in
+ terms of shape, size, color, orientation, and the image may
+ be blurred or have irregular illumination, etc. In this
+ paper, we describe a robust and accurate multiresolution
+ approach to detect and classify text regions in such
+ scenarios. Based on generation/validation paradigm, we
+ first segment images to detect character regions with a
+ multiresolution algorithm able to manage large character
+ size variations. The segmented regions are then filtered
+ out using shapebased classification, and neighboring
+ characters are merged to generate text hypotheses. A
+ validation step computes a region signature based on
+ texture analysis to reject false positives. We evaluate our
+ algorithm in two challenging databases, achieving very good
+ results},
+ keywords = {Text detection, multiresolution, image segmentation,
+ machine learning}
+}
+
+@InProceedings{ perrot.06.nist,
+ author = {Patrick Perrot and R\'eda Dehak and G\'erard Chollet},
+ title = {{ENST-IRCGN} System Description},
+ booktitle = {NIST SRE'06 Workshop: speaker recognition evaluation
+ campaign},
+ year = 2006,
+ address = {San Juan, Puerto Rico},
+ month = jun,
+ urllrde = {200606-NIST-B}
+}
+
+@Misc{ pouillard.05.sud,
+ author = {Akim Demaille and {\relax Th}omas Largillier and Nicolas
+ Pouillard},
+ title = {{ESDF}: A proposal for a more flexible {SDF} handling},
+ note = {Communication to Stratego Users Day 2005},
+ year = 2005,
+ address = {Utrecht {U}niversity, {N}etherland},
+ month = may,
+ project = {Transformers},
+ urllrde = {200505-SUD-esdf},
+ abstract = {By the means on its annotations, Syntax Definition
+ Formalism (SDF) seems to be extensible: the user is tempted
+ to tailor its grammar syntax by adding new annotation
+ kinds. Unfortunately the standard SDF crunching tools from
+ Stratego/XT do not support the extension of SDF, and the
+ user has to develop the whole set of tools for her home
+ grown extension(s). We present the SDF tool set that
+ provides ``weak'' genericity with respect to the grammar
+ grammar: support for arbitrary SDF annotations. We would
+ like to contribute it to Stratego/XT since its components
+ subsume their stock peers. Finally, we present a set of
+ four extensions we find useful.}
+}
+
+@InProceedings{ regisgianas.03.poosc,
+ author = {Yann R\'egis-Gianas and Rapha\"el Poss},
+ title = {On orthogonal specialization in {C++}: dealing with
+ efficiency and algebraic abstraction in {V}aucanson},
+ booktitle = {Proceedings of the Parallel/High-performance
+ Object-Oriented Scientific Computing (POOSC; in conjunction
+ with ECOOP)},
+ year = 2003,
+ number = {FZJ-ZAM-IB-2003-09},
+ pages = {71--82},
+ editor = {J\"org Striegnitz and Kei Davis},
+ series = {John von Neumann Institute for Computing (NIC)},
+ address = {Darmstadt, Germany},
+ month = jul,
+ project = {Vaucanson},
+ urllrde = {200307-Poosc},
+ abstract = {Vaucanson is a C++ generic library for weighted finite
+ state machine manipulation. For the sake of generality, FSM
+ are defined using algebraic structures such as alphabet
+ (for the letters), free monoid (for the words), semiring
+ (for the weights) and series (mapping from words to
+ weights). As usual, what is at stake is to maintain
+ efficiency while providing a high-level layer for the
+ writing of generic algorithms. Yet, one of the
+ particularities of FSM manipulation is the need of a fine
+ grained specialization power on an object which is both an
+ algebraic concept and an intensive computing machine.}
+}
+
+@InProceedings{ ricou.07.adass,
+ author = {Olivier Ricou and Anthony Baillard and Emmanuel Bertin and
+ Frederic Magnard and Chiara Marmo and Yannick Mellier},
+ title = {Web services at {TERAPIX}},
+ booktitle = {Proceedings of the XVII conference on Astronomical Data
+ Analysis Software \& Systems (ADASS)},
+ month = sep,
+ year = 2007,
+ urllrde = {200709-ADASS},
+ abstract = {We present an implementation of V.O.-compliant web
+ services built around software tools developed at the
+ TERAPIX centre. These services allow to operate from a
+ remote site several pipeline tasks dedicated to
+ astronomical data processing on the TERAPIX cluster,
+ including the latest EFIGI morphological analysis tool.}
+}
+
+@InProceedings{ ricou.07.eceg,
+ author = {Olivier Ricou},
+ title = {10 years of confrontation between {French} {Internet}
+ users and their successive governments},
+ booktitle = {Proceedings of the 7th European Conference on e-Government
+ (ECEG)},
+ month = jun,
+ year = 2007,
+ urllrde = {200706-ECEG},
+ abstract = {This paper is a testimony on the relations between the
+ Internet users and their governments in France during the
+ last decade. It shows the complexity of communication
+ between two worlds that are strangers to each other. Since
+ most of the confrontation occurred over law proposals, it
+ analyses their impact on Internet users and focuses on two
+ examples. These example show the failure of Internet as a
+ political medium. French politicians do not seem to want an
+ active participation of the citizens in decisionmaking
+ processes. In order to end this paper on an optimistic
+ note, the last section enumerates the achievements of
+ egovernment which contributed preparing for a better
+ democracy by increasing transparency, accountability, and
+ education. This might push citizens to ask for more. }
+}
+
+@InProceedings{ ricou.08.eceg,
+ author = {Olivier Ricou},
+ title = {A Survey of {French} Local e-Democracy},
+ booktitle = {Proceedings of the 8th European Conference on e-Government
+ (ECEG)},
+ month = jul,
+ year = 2008,
+ urllrde = {200807-ECEG},
+ abstract = {Since the end of the last century, the Internet has shown
+ that it is a different media, a media of citizen
+ journalists. This paper surveys e-democratic tools used at
+ the local level in France in order to see how the Internet
+ can change our democracy and people's participation. It
+ describes the official tools provided by municipalities and
+ administrations as well as citizens' tools, like blogs,
+ which become more and more important in today's democratic
+ debate. It analyses how they help for more transparency,
+ accountability and participation, which might lead to
+ define new democratic rules.}
+}
+
+@TechReport{ vaucanson.04.techrep,
+ author = {The \textsc{Vaucanson} group},
+ title = {Proposal: an {XML} representation for automata},
+ institution = {EPITA Research and Development Laboratory (LRDE)},
+ year = 2004,
+ number = 0414,
+ address = {France},
+ month = nov,
+ url = {http://www.lrde.epita.fr/cgi-bin/twiki/view/Publications/200414-TR}
+ ,
+ project = {Vaucanson},
+ urllrde = {200414-TR}
+}
+
+@InProceedings{ verna.00.vsmm,
+ author = {Didier Verna},
+ title = {Action recognition: how intelligent virtual environments
+ can ease human-machine interaction},
+ booktitle = {Proceedings of the 6th International Conference on Virtual
+ Systems and MultiMedia (VSMM)---Intelligent Environments
+ Workshop},
+ pages = {703--713},
+ year = 2000,
+ address = {Gifu, Japan},
+ month = oct,
+ publisher = {IOS Press, USA},
+ isbn = {1-58603-108-2},
+ project = {Urbi},
+ urllrde = {200010-Vsmm-2},
+ abstract = {This paper describes a research that has been conducted in
+ the field of cognitive assistance to human-machine
+ interaction in virtual environments. The idea is to design
+ a system which, bearing in mind the actions performed by
+ the operator at present and the current state of the
+ environment, attempts to determine the global operation
+ that the user is in the process of executing, and
+ eventually takes control of the same process in order to
+ complete it automatically. This idea implies the conception
+ of an action recognition mechanism based on a specific
+ knowledge representation model. This mechanism is
+ implemented in a computer demonstrator, known as the TOASt
+ system, which is also presented.}
+}
+
+@InProceedings{ verna.01.sci,
+ author = {Didier Verna},
+ title = {Virtual reality and tele-operation: a common framework},
+ booktitle = {Proceedings of the 5th World Multi-Conference on
+ Systemics, Cybernetics and Informatics (SCI)---Emergent
+ Computing and Virtual Engineering},
+ year = 2001,
+ volume = 3,
+ pages = {499--504},
+ address = {Orlando, Florida, USA},
+ month = jul,
+ editors = {N. Callas and S. Esquivel and J. Burge},
+ project = {Urbi},
+ urllrde = {200107-Sci},
+ abstract = {This paper proposes an overview of a study that
+ conceptually unify the fields of virtual reality and
+ tele-operation, by analyzing the notion of ``assistance''
+ to the operator of a virtual reality or tele-operation
+ system. This analysis demonstrates that cases of assistance
+ that are usually considered to belong to virtual reality
+ are not conceptually different from what has been done in
+ tele-operation since long before virtual reality appeared.
+ With this common framework for virtual reality and
+ tele-operation, we hope to provide a theoretical
+ formalization of many ideas acquired empirically, and hence
+ a basis onto which further discussion could be undertaken
+ in a constructive manner.}
+}
+
+@InProceedings{ verna.06.ecoop,
+ author = {Didier Verna},
+ title = {Beating {C} in Scientific Computing Applications},
+ booktitle = {Third European Lisp Workshop at ECOOP},
+ year = 2006,
+ address = {Nantes, France},
+ month = jul,
+ note = {Best paper award.},
+ project = {Software},
+ urllrde = {200607-ECOOP},
+ abstract = {This paper presents an ongoing research on the behavior
+ and performance of Lisp with respect to C in the context of
+ scientific numerical computing. Several simple image
+ processing algorithms are used to evaluate the performance
+ of pixel access and arithmetic operations in both
+ languages. We demonstrate that the behavior of equivalent
+ Lisp and C code is similar with respect to the choice of
+ data structures and types, and also to external parameters
+ such as hardware optimization. We further demonstrate that
+ properly typed and optimized Lisp code runs as fast as the
+ equivalent C code, or even faster in some cases.}
+}
+
+@Article{ verna.06.ijcs,
+ author = {Didier Verna},
+ title = {How to make Lisp go faster than {C}},
+ journal = {IAENG International Journal of Computer Science},
+ year = 2006,
+ volume = 32,
+ number = 4,
+ month = dec,
+ issn = {1819-656X},
+ project = {Software},
+ urllrde = {200606-IMECS},
+ abstract = {Contrary to popular belief, Lisp code can be very
+ efficient today: it can run as fast as equivalent C code or
+ even faster in some cases. In this paper, we explain how to
+ tune Lisp code for performance by introducing the proper
+ type declarations, using the appropriate data structures
+ and compiler information. We also explain how efficiency is
+ achieved by the compilers. These techniques are applied to
+ simple image processing algorithms in order to demonstrate
+ the announced performance on pixel access and arithmetic
+ operations in both languages.}
+}
+
+@InProceedings{ verna.06.imecs,
+ author = {Didier Verna},
+ title = {How to make Lisp go faster than {C}},
+ booktitle = {Proceedings of the International MultiConference of
+ Engineers and Computer Scientists},
+ year = 2006,
+ address = {Hong Kong},
+ month = jun,
+ organization = {International Association of Engineers},
+ isbn = {988-98671-3-3},
+ project = {Software},
+ urllrde = {200606-IMECS},
+ abstract = {Contrary to popular belief, Lisp code can be very
+ efficient today: it can run as fast as equivalent C code or
+ even faster in some cases. In this paper, we explain how to
+ tune Lisp code for performance by introducing the proper
+ type declarations, using the appropriate data structures
+ and compiler information. We also explain how efficiency is
+ achieved by the compilers. These techniques are applied to
+ simple image processing algorithms in order to demonstrate
+ the announced performance on pixel access and arithmetic
+ operations in both languages.}
+}
+
+@Article{ verna.06.practex,
+ author = {Didier Verna},
+ year = 2006,
+ volume = 2006,
+ number = 3,
+ month = aug,
+ project = {Software},
+ urllrde = {200608-PracTeX},
+ abstract = {This paper presents {\CurVe}, a curriculum vitae class for
+ \LaTeX2e, in a progressive approach going from a first
+ contact with the class, through concrete examples of
+ customization, and some aspects of advanced usage.}
+}
+
+@InProceedings{ verna.07.imecs,
+ author = {Didier Verna},
+ title = {{CLOS} solutions to binary methods},
+ booktitle = {Proceedings of the International MultiConference of
+ Engineers and Computer Scientists},
+ year = 2007,
+ address = {Hong Kong},
+ month = mar,
+ organization = {International Association of Engineers},
+ project = {Software},
+ urllrde = {200703-IMECS},
+ abstract = {Implementing binary methods in traditional object oriented
+ languages is difficult: numerous problems arise, such as
+ typing (covariance vs. contra-variance of the arguments),
+ polymorphism on multiple arguments (lack of multi-methods)
+ etc. The purpose of this paper is to demonstrate how those
+ problems are either solved, or nonexistent in the Common
+ Lisp Object System (CLOS). Several solutions for different
+ levels of binary methods support in CLOS are proposed. They
+ mainly consist in re-programming a binary method specific
+ object system through the CLOS meta-object protocol.}
+}
+
+@InProceedings{ verna.08.els,
+ author = {Didier Verna},
+ title = {Binary Methods Programming: the {CLOS} Perspective},
+ booktitle = {Proceedings of the First European Lisp Symposium},
+ pages = {91--105},
+ year = 2008,
+ address = {Bordeaux, France},
+ month = may,
+ project = {Software},
+ urllrde = {200805-ELS},
+ abstract = {Implementing binary methods in traditional object-oriented
+ languages is difficult: numerous problems arise regarding
+ the relationship between types and classes in the context
+ of inheritance, or the need for privileged access to the
+ internal representation of objects. Most of these problems
+ occur in the context of statically typed languages that
+ lack multi-methods (polymorphism on multiple arguments).
+ The purpose of this paper is twofold: first, we show why
+ some of these problems are either non-issues, or easily
+ solved in Common Lisp. Then, we demonstrate how the Common
+ Lisp Object System (CLOS) allows us not only to implement
+ binary methods in a straightforward way, but also to
+ support the concept directly, and even enforce it at
+ different levels (usage and implementation).}
+}
+
+@Article{ verna.08.jucs,
+ author = {Didier Verna},
+ title = {Binary Methods Programming: the {CLOS} Perspective
+ (extended version)},
+ journal = {Journal of Universal Computer Science},
+ year = 2008,
+ volume = 14,
+ number = 20,
+ pages = {3389--3411},
+ project = {Software},
+ urllrde = {200811-JUCS},
+ abstract = {Implementing binary methods in traditional object-oriented
+ languages is difficult: numerous problems arise regarding
+ the relationship between types and classes in the context
+ of inheritance, or the need for privileged access to the
+ internal representation of objects. Most of these problems
+ occur in the context of statically typed languages that
+ lack multi-methods (polymorphism on multiple arguments).
+ The purpose of this paper is twofold: first, we show why
+ some of these problems are either non-issues, or easily
+ solved in Common Lisp. Then, we demonstrate how the Common
+ Lisp Object System (CLOS) allows us not only to implement
+ binary methods in a straightforward way, but also to
+ support the concept directly, and even enforce it at
+ different levels (usage and implementation).}
+}
+
+@InProceedings{ verna.08.lncs,
+ author = {Didier Verna and Charlotte Herzeel and Christophe Rhodes
+ and Hans H\"ubner},
+ title = {Report on the 5th Workshop {ELW} at {ECOOP 2008}},
+ booktitle = {Object-Oriented Technology. ECOOP 2008 Workshop Reader.},
+ pages = {1--6},
+ year = 2008,
+ editor = {Patrick Eugster},
+ volume = 5475,
+ series = {Lecture Notes in Computer Science},
+ month = jul,
+ publisher = {Springer}
+}
+
+@InProceedings{ verna.09.accu,
+ author = {Didier Verna},
+ title = {Revisiting the Visitor: the Just Do It Pattern},
+ booktitle = {Proceedings of the ACCU Conference 2009},
+ year = 2009,
+ address = {Oxford},
+ project = {Software},
+ urllrde = {200904-ACCU},
+ abstract = { },
+ note = {Accepted}
+}
+
+@InProceedings{ verna.09.ilc,
+ author = {Didier Verna},
+ title = {{CLOS} Efficiency: Instantiation},
+ booktitle = {Proceedings of the International Lisp Conference},
+ year = 2009,
+ month = mar,
+ pages = {76--90},
+ organization = {Association of Lisp Users},
+ urllrde = {200903-ILC},
+ abstract = {This article reports the results of an ongoing
+ experimental research on the behavior and performance of
+ CLOS, the Common Lisp Object System. Our purpose is to
+ evaluate the behavior and performance of the 3 most
+ important characteristics of any dynamic Object Oriented
+ system: class instantiation, slot access and dynamic
+ dispatch. This paper describes the results of our
+ experiments on instantiation. We evaluate the efficiency of
+ the instantiation process in both C++ and Lisp under a
+ combination of parameters such as slot types or classes
+ hierarchy. We show that in a non-optimized configuration
+ where safety is given priority on speed, the behavior of
+ C++ and Lisp instantiation can be quite different, which is
+ also the case amongst different Lisp compilers. On the
+ other hand, we demonstrate that when compilation is tuned
+ for speed, instantiation in Lisp becomes faster than in
+ C++.}
+}
+
+@InProceedings{ verna.10.els,
+ author = {Didier Verna},
+ title = {{CLoX}: {C}ommon {L}isp objects for {XEmacs}},
+ booktitle = {Proceedings of the 3rd European Lisp Symposium},
+ year = 2010,
+ address = {Lisbon, Portugal},
+ month = may,
+ project = {Software},
+ urllrde = {201005-ELS},
+ abstract = {CLoX is an ongoing attempt to provide a full Emacs Lisp
+ implementation of the Common Lisp Object System, including
+ its underlying meta-object protocol, for XEmacs. This paper
+ describes the early development stages of this project.
+ CLoX currently consists in a port of Closette to Emacs
+ Lisp, with some additional features, most notably, a deeper
+ integration between types and classes and a comprehensive
+ test suite. All these aspects are described in the paper,
+ and we also provide a feature comparison with an
+ alternative project called Eieio.}
+}
+
+@Article{ verna.10.jucs,
+ author = {Didier Verna},
+ title = {Revisiting the Visitor: the Just Do It Pattern},
+ journal = {Journal of Universal Computer Science},
+ year = 2010,
+ volume = 16,
+ pages = {246--271},
+ project = {Software},
+ urllrde = {201004-JUCS},
+ abstract = {While software design patterns are a generally useful
+ concept, they are often (and mistakenly) seen as ready-made
+ universal recipes for solving common problems. In a way,
+ the danger is that programmers stop thinking about their
+ actual problem, and start looking for pre-cooked solutions
+ in some design pattern book instead. What people usually
+ forget about design patterns is that the underlying
+ programming language plays a major role in the exact shape
+ such or such pattern will have on the surface. The purpose
+ of this paper is twofold: we show why design pattern
+ expression is intimately linked to the expressiveness of
+ the programming language in use, and we also demonstrate
+ how a blind application of them can in fact lead to very
+ poorly designed code.}
+}
+
+@InProceedings{ verna.10.tug,
+ author = {Didier Verna},
+ title = {Classes, Styles, Conflicts: the Biological Realm of
+ {\LaTeX}},
+ booktitle = {TUGboat},
+ pages = {162--172},
+ year = 2010,
+ editor = {Barbara Beeton and Karl Berry},
+ volume = 31,
+ number = 2,
+ project = {Software},
+ urllrde = {201006-TUG},
+ abstract = {The \LaTeX{} world is composed of thousands of software
+ components, most notably classes and styles. Classes and
+ styles are born, evolve or die, interact with each other,
+ compete or cooperate, very much as living organisms do at
+ the cellular level. This paper attempts to draw an extended
+ analogy between the \LaTeX{} biotope and cellular biology.
+ By considering \LaTeX{} documents as living organisms and
+ styles as viruses that infect them, we are able to exhibit
+ a set of behavioral patterns common to both worlds. We
+ analyze infection methods, types and cures, and we show how
+ \LaTeX{} or cellular organisms are able to survive in a
+ world of perpetual war.}
+}
+
+@InProceedings{ verna.11.onward,
+ author = {Didier Verna},
+ title = {Biological Realms in Computer Science: the Way You Don't
+ (Want To) Think About Them},
+ booktitle = {Onward! 2011},
+ year = {2011},
+ note = {Accepted},
+ abstract = {In biology, evolution is usually seen as a tinkering
+ process, different from what an engineer does when he plans
+ the development of his systems. Recently, studies have
+ shown that even in biology, there is a part of good
+ engineering. As computer scientists, we have much more
+ difficulty to admit that there is also a great deal of
+ tinkering in what we do, and that our software systems
+ behave more and more like biological realms every day. This
+ essay relates my personal experience about this
+ discovery.}
+}
+
+@InProceedings{ verna.11.tug,
+ author = {Didier Verna},
+ title = {Towards {\LaTeX} Coding Standards},
+ booktitle = {TUGboat},
+ OPTpages = {00--00},
+ year = 2011,
+ OPTeditor = {Barbara Beeton and Karl Berry},
+ OPTvolume = 00,
+ OPTnumber = 0,
+ project = {Software},
+ OPTurllrde = {201100-TUG},
+ note = {Accepted},
+ abstract = {Because \LaTeX is only a macro-expansion system, the
+ language does not impose any kind of good software
+ engineering practice, program structure or coding style.
+ Maybe because in the \LaTeX world, collaboration is not so
+ widespread, the idea of some LaTeX Coding Standards is not
+ so pressing as with other programming languages. Over the
+ years, the permanent flow of personal development
+ experiences contributed to shape our own taste in terms of
+ coding style. In this paper, we report on all these
+ experiences and describe what we think are good programming
+ practices.}
+}
+
+@InProceedings{ xue.03.icip,
+ author = {Heru Xue and {\relax Th}ierry G\'eraud and Alexandre
+ Duret-Lutz},
+ title = {Multi-band segmentation using morphological clustering and
+ fusion application to color image segmentation},
+ booktitle = {Proceedings of the IEEE International Conference on Image
+ Processing (ICIP)},
+ year = 2003,
+ pages = {353--356},
+ volume = 1,
+ address = {Barcelona, Spain},
+ month = sep,
+ project = {Image},
+ urllrde = {200309-Icip},
+ abstract = {In this paper we propose a novel approach for color image
+ segmentation. Our approach is based on segmentation of
+ subsets of bands using mathematical morphology followed by
+ the fusion of the resulting segmentation channels. For
+ color images the band subsets are chosen as RG, RB and GB
+ pairs, whose 2D histograms are processed as projections of
+ a 3D histogram. The segmentations in 2D color spaces are
+ obtained using the watershed algorithm. These 2D
+ segmentations are then combined to obtain a final result
+ using a region split-and-merge process. The CIE L a b color
+ space is used to measure the color distance. Our approach
+ results in improved performance and can be generalized for
+ multi-band segmentation of images such as multi-spectral
+ satellite images information.}
+}
+
+@InProceedings{ yoruk.04.eusipco,
+ author = {Erdem Yoruk and Ender Konukoglu and Bulent Sankur and
+ J\'er\^ome Darbon},
+ title = {Person authentication based on hand shape},
+ booktitle = {Proceedings of 12th European Signal Processing Conference
+ (EUSIPCO)},
+ year = 2004,
+ address = {Vienna, Austria},
+ month = sep,
+ project = {Image},
+ urllrde = {200409-EUSIPCO},
+ abstract = {The problem of person identification based on their hand
+ images has been addressed. The system is based on the
+ images of the right hands of the subjects, captured by a
+ flatbed scanner in an unconstrained pose. In a
+ preprocessing stage of the algorithm, the silhouettes of
+ hand images are registered to a fixed pose, which involves
+ both rotation and translation of the hand and, separately,
+ of the individual fingers. Independent component features
+ of the hand silhouette images are used for recognition. The
+ classification performance is found to be very satisfactory
+ and it was shown that, at least for groups of one hundred
+ subjects, hand-based recognition is a viable secure access
+ control scheme.}
+}
+
+@Article{ yoruk.06.itip,
+ author = {Erdem Y\"or\"uk and Ender Konukoglu and B\"ulent Sankur
+ and J\'er\^ome Darbon},
+ title = {Shape-based hand recognition},
+ journal = {IEEE Transactions on Image Processing},
+ year = 2006,
+ volume = 15,
+ number = 7,
+ pages = {1803--1815},
+ month = jul,
+ project = {Image},
+ urllrde = {2006XX-ITIP},
+ abstract = {The problem of person recognition and verification based
+ on their hand images has been addressed. The system is
+ based on the images of the right hands of the subjects,
+ captured by a flatbed scanner in an unconstrained pose at
+ 45 dpi. In a preprocessing stage of the algorithm, the
+ silhouettes of hand images are registered to a fixed pose,
+ which involves both rotation and translation of the hand
+ and, separately, of the individual fingers. Two feature
+ sets have been comparatively assessed, Hausdorff distance
+ of the hand contours and independent component features of
+ the hand silhouette images. Both the classification and the
+ verification performances are found to be very satisfactory
+ as it was shown that, at least for groups of about five
+ hundred subjects, hand-based recognition is a viable secure
+ access control scheme.}
+}
+
+%% Local Variables:
+%% fill-column: 76
+%% ispell-local-dictionary: "american"
+%% End:
diff --git a/doc/mainpage.hh b/doc/mainpage.dox
similarity index 80%
rename from doc/mainpage.hh
rename to doc/mainpage.dox
index 116c69a..8f49ca3 100644
--- a/doc/mainpage.hh
+++ b/doc/mainpage.dox
@@ -13,8 +13,8 @@
<tr>
<td>
\li What is Olena ?
- \li \ref bibliography
- \li Contributors
+ \li <a class="el" href="http://www.lrde.epita.fr/cgi-bin/twiki/view/Olena/Publications">Bibliography</a>
+ \li <a class="el" href="a00028.html">Contributors</a>
</td>
<td>
\li \ref moduleslist
@@ -23,7 +23,7 @@
\li Swilena
</td>
<td>
- \li Online Demos
+ \li <a class="el" href="http://www.lrde.epita.fr/cgi-bin/twiki/view/Olena/Demos">Online Demos</a>
\li Papers Related Programs
\li Olena Powered Programs
</td>
diff --git a/doc/modules_list.hh b/doc/modules_list.dox
similarity index 100%
rename from doc/modules_list.hh
rename to doc/modules_list.dox
diff --git a/doc/olena.qhcp b/doc/olena.qhcp
deleted file mode 100644
index 595d34e..0000000
--- a/doc/olena.qhcp
+++ /dev/null
@@ -1,12 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<QHelpCollectionProject version="1.0">
- <assistant>
- <startPage>qthelp://fr.epita.lrde.olena/milena-2.0/index.html</startPage>
- </assistant>
- <docFiles>
- <register>
- <file>../milena/doc/user-refman/html/milena.qch</file>
- <file>../scribo/doc/user-refman/html/scribo.qch</file>
- </register>
- </docFiles>
-</QHelpCollectionProject>
\ No newline at end of file
--
1.7.2.5
12 years, 2 months
- 1
- 0