From 37ed78fdb5f710c586efa6a2299cf711b24acebd Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Wed, 1 Feb 2017 20:55:29 +0000
Subject: [PATCH 01/18] detect ods types: boolean, currency, time and
 percentage. support repeated column and multi-line text in a cell.

---
 CHANGELOG.md           |   3 ++
 horror/ods_formats.ods | Bin 0 -> 9960 bytes
 messytables/ods.py     | 111 +++++++++++++++++++++++++++++++++--------
 messytables/types.py   |  45 ++++++++++++++++-
 test/test_read.py      | 108 +++++++++++++++++++++++++++++++++++++--
 5 files changed, 238 insertions(+), 29 deletions(-)
 create mode 100644 horror/ods_formats.ods

diff --git a/CHANGELOG.md b/CHANGELOG.md
index a3b1d91..62bead6 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,6 @@
+0.15.2 (unreleased)
+* #: detect ods types: boolean, currency, time and percentage. support repeated columns
+
 0.15.1 (29 September 2016)
 * #158: Add CDFV2-unknown to MIMELOOKUP
 * #157: Fix for Python Magic API change
diff --git a/horror/ods_formats.ods b/horror/ods_formats.ods
new file mode 100644
index 0000000000000000000000000000000000000000..fa1f7ff9fceef1586e0b071d503a87f919be99cb
GIT binary patch
literal 9960
zcmdU#bzEFY)2K;s34|aC76=~PT@u`bJ3$Bc8Qhsb2(H21J-AzNw+!y?1b6sg-!Gf&
zzMH-G{&{PD$L35`PgkFQ>U3AjN<zb6LO~%wK}i}>iTRjw`qM!{K|O5uUqM-zSO7py
z)&N~=YjYEQU66^DCB40+0iBgD&;&?lWeu=2u+q1)09b<PfY!DET?3#o005Hxi%kq8
zu|p6X6x74^;@+mBv7LpUrLKuNkRJ5ol+N1H=##9B=o2Jdr2C&d5f>AZzdttKZ+8gL
z_j|{p1jGA-zO0m@FbWC^78VvB9^R{0uc)c1nVFe+cz6T^1SBLR6ciLRG&J<}_07%A
zfk2>(iwhVG4hRT{h=@o?NJvjlFDNJ|D=Vw3t7~g(>*?tk9UYyXo?c#F-rCwaK0dy^
zy?rS0?yk5SoCF00+b1scR?%^4cVg0Zh4Lxv(93DPVQ*!&M;CX=oS*82K;?OoXzpJO
zErlnPaWc5lh?s}zYfrYH(;ChMx86wH8Jt&WHoCGIN+erpI*;)lwzrN$MzI=?iwMn(
z+}agf`7f+vlOweu>+!?9+m<f-5!bHShgf`f*ol~H?K1Mw*+x>IS90dz7|F@TC=wk2
z&m*E@8lJj?0u<{oE!@MFj=`04oZ)f^X2Q%jqgtzV6*d-@=0@}f{e~aYI&3fo&-+<6
z0OW>77Ml83-ThW-@-AFTg}60V`vHNK`${z(Xc8j3AZ+)bagq6&lEa(pz(MGowNDvy
z4!pG;R+4p10O{G+9{J-D;6YQ>UExMdQx$phpm=)em!hV}mQ-A37v|^`N#t4OdcpFS
z!^Hv@BCIZwHy0Z!d}bdsaA)JFBi7fvFs_rk=$i;5=WowM2a9FMJP2+81U3rp`^PrU
z0~ed6SG&Nukb(2e>6~)bPR!L$U{&el@NspXpoogs0dd5mcn)$m=nUP+oO~ZBk(wUo
zGO^Yx;t*$!4Tk$!aIz5OJK^{l9{KT`Nk$Z1%fT|`9R^87PAj-u;0np$J(2#z{b{48
z`C4;D(0@RI-{x9T&~g1JYI%Otogwa8_QG<5lOGL@0`22z2+E*J_H7re#geL9QfBoh
zULOw?s~20>xuU0w8%L0nqS+VR+@a6#>oUzhc*ye~>}H-)23<O{qg@=GcC+6`%<pQ9
zHY6x>v%+rb_Z+>cU5ttBt6F1he`2?iC|5{Hn7*G3(Tr+x(7w)2-{%u;IyddRnv3kb
z#;4+9*^blWEw!C}sqp?u?0CV}di|gqUmRUfBmLlc*!o+sHg<GJS*K@5LFZKb#eSv{
z={;4Lp8>ho%Nz7&Hq$2aTVGaWn+!FI>v<7*_ReFOR+sBtI&KKm&M1yTw0bu}LI=FI
zC{bM|fmGbHx7G@q23q2U1rWfKXyGFn;iP^yvdpyW%F#-Ks7tBbh|7tsiutdDd40X4
zGbJ7@V=VR0c#&Jb+jpQ5*d4toJd1zD8<ZhHGmAGFIyf?R(Bu@8L$^aWB$hBgOBteg
zP2h?uITwk0mY8Zm#kyCF?4dbq<Z2mLvzI*9y!n+<(l7u-I)giZ>xek3j#{?0;9~KZ
zphmV_-2*LI0n9EswZvb4m+N}lT0PpW5PvKW6kY{3$N_~3Rwfbwh(&Zgx5V1e^K2Gm
zdFi30OsmJduFsO-{VrV4-+%0RS9hoxlNjtC{$3qxCCU5rYvuclbYldn9W-(@ZSELA
zU`Zm_U!QR8^oktqxd;yf-AG?eV$&k+YE-2cMz3ugh?d=kIof=Cp&Iu%tf@LSAn&%o
zG-tY{wmm^`{>)^F`Mf;0k5C#RVc`z`++K!#jk}=-?+D(*yw$Qd1T6_>4$N<%l4O0S
z>Tdes?E=i5M>9QzsKbKM>sqZI6*>Z-LRut!SapiZJ{p+^j|W-?Mkb%y*8J9$uEX>O
z@@{QI`${g_%PZB>vybf-R3_U&eHQhw9{Yy}Q2TDwc059>bsOrCra2;uOv8%yAjKud
z%U%H|?K<gp9SFhNEsHdnNY9?5L^q!1`zN9$nme>>d?VQB&$FGWukiI$_qwCKDwXk!
z%+qG&zHiE0FV|$sg3UTQs7KFc3(sBpCB8ZuiNy_c@y(eXZU6{B+N|=BMKkuz_Mba#
zVIygkvy9Aivd#vCRHN~oFM$6)+}-VrH$|W9Yw7tvRk|lUl(?{rP?3O+_kSW40s_Jx
z(eL5T4?(#<0RliE6H6l?ouh^MuKGMMjSJJg$wT<#IJZ`qWkbMdEx3Sx%3>B9Q)B+I
zP&XT?)H9*0@p-7)s%;4~Ec?@e4KT5IBap-O!q#?S>4><E0@%fowj!VFAsdAu#<NAj
zRHrD?y0=NO3fwM6=1L?+V{_YTe?__!Dfr6ZxjXSkjJdve`ly&Na#;KW@<*>$q!(FD
zV7i@Ej{y6IhBW+T&3-&uR9;Wz%9JI;R|IWOy-L*zi?*F<&9od&C}Yzmi?R;q=%wW(
zJYGn^uM8M;uqfTA_TEubP*Z4s5K&Oy<|v?VvMXT<y-Of|)XKs}kw2Pi7VXb6nLrYA
z?H5;SF*B%--xd~-+CubVL`;p{A3%zSjxHhDkaL4{NVkB;fdwH738VM67>XM+P9Q{?
z|7eVDZl`r63UTn^fASXO?}zq$Xr2W&Vwe3j9?53==Q<9Q2x^a3NxB#V&k>0WS&!)K
z*Bidcf<(}VpVRr#N}v&h-QXX`F)l~W$_0N|xofS|+26S{>k1)b`tlCPIZFu2T=c~!
zjjh)H@9>cX&DLEohj{|?-`xpPn2a_g@2H#EdA!yb1goH8u|9}Sqq*){b>~(1R``>y
zzL~LwxCBjQ2@HJ^T0qQPlue85*}(kvh+|HuA0IM!it;F?MNwTF)@o<7gkUJ6b12KS
zE>aJ{Z{!(`?{rvIf~#AJqf%;xEvEAzB|g?=&XA}NAJws3Y5>;d&Yu?{NH=wuwK<&m
zfjeAfrE%hK<;dxS@e6ofroX83#8PhA3cAcDX-CaaHBIR!a;(*7D9pkdwm@Q_Uznyd
zBo7_#^SE{A+xO;ANu%1hAz#$!dYvcE2*4YP$x_HEVjW=Ft(W`iR!BtU)||O*BSsT<
z{ObFw{0=d4Qh5`qH%&|PdS2*Wr%a&NYiu)`Hj<3ta*g!kz=Wbrc>-=ytN{T&0*2nA
zYUwvD8Q2StElcxIpgF(PXI&!-0I+Gg$ZGt1SMzGKXDq&2^`D{nROye40^1~uPVXK!
z&R)3XMzrD~x9yT0JQhK8rZY6z%<bGd7(PYC_n#$vx<;V~H-dMaMB#H-3#&SzhNR!J
zdD{0TI@wL|8gL>8L50c}g0iIiPMg~ixd4Fu4%cC;6t;*y{*Cp~0TG=xqi7f!%nnm*
zQVYY^!)hd@TO=Gtj>ju><Vk)lpBs#D(6>L(9oI3%(rPx{y4Kt^`P)Dy1=YMSvS^eN
zdX%CUlw+g7oS;K?iAHn`h4;ct@&NDmZd6f-tOn2***ej4U>K&A!~*Br784NUsuNEr
z>Ch*;I<$8Ev|%LSdzE=Bl{8O~sw{>MeUV|Nwwy*d(_DdTt!HaTCmXRRZPb`U`kjOH
zNeLd8Btv6j(B*iTB+X8LApaq={LS9eDvR~qBK_8C)^m^XVaC8_84tG$z(=l9XRVWA
zHf(Z}UNe3b3N)44OwNOqMJ{K}%Ke6?o$iBbajA6E1Wq$bEvg~=38yn8V}~|@O3e6u
z4KF+9%4!)}ofmxESA3PB{El|cwkF8VE(>Ex*|kyJ&4lK)EEvxOaON9@_kC6J%5Qc~
z@hzBSK$A&Q&Jo$&t)$N1L@s-8-Tj*0`I_o2oi;#l`bJ$~xRn>qxU@wZ=wdm*2)^kZ
zH{a*JT*?qL^1HimCD$!>F-)m+G~}U)B<^jztw@kF8YXZ!X_2j{T`QQiNjNQBg&9rd
z+T874qx{;7{&5HSX^JabyzMd+Pq$HoFFKefF>s$w58LvKX;XcK8nv@V#R<KvBs_wS
z?d=r<EELo}6%^F(NsD|>T74@^(8D0&9<-|(7Qp37^xHY@R9%Xj5n0zTgUN#3KC(?n
zB4-tQ$!K&41ERRGm^ZZN=M$I9cZa>DP~r#~V*bHGvn&mZVxl_QNm?}6bwjhe>{jAD
zPVIphadT>O-zvyCo|p)YoFtwsABkQ~U9R0aRDmjpU!fA?t|7|!5R51`pTSfsyu>VX
zP#zfZB$jauO@ldl^WjmywYm)~xm^kfi$rs0sb$B1Vg>tL8U^sVob~-NlV%w)TMdbC
zrC=->EeFIvdpNbvil<Tqvp&f{EELYaDkg(}Rt@h}X|qk`^e%`yUcS=4K0Y2#wBNKj
zHLKSan>vpYWF45vwCAWWm}`tdGbHO8KK3DUtsB`8tObVQQiTNSqLS(qq4eo`OC_nh
zHWa~&j&Ikt_AkPfwAf_%ZKpB?_<ro6VRFJCWh2;j+2#3~ap`zn9zy*Y1wv+GR)mDp
z4HvH_8-E=s|3rxiGhVCZM9X0hb06-cGTOD*uepy1&81rq^W)SqiwO{S$O%<;;#|qM
zeoklvy&l}+;BO+`mo77W=X`yPX=&|!+P1<`A}voRTYH8Q5in%3`Bj%AUCOw#z?VW>
zl~!AS+K+TkhQPLxSgVZElnj=G8GZ!8T<<DsVRzrGrh@;aCBCr1+UR?@K|9i*x$rMD
zS1W^Km#*r=JVE{}vK-J&fEaX6AH8<$Z=^Bk_@mmdN_j?65G#b<S!?dPTgd0Xkl6ho
z$na)pl!l>tdmT=4&i7M<mrS~nf}Z~9$Am$l=_$Uug)UBwm+{sNk#T{xEmeE@Qz>1r
zGsDo-Ip{lBNF@Y|3~Ogq!olOM5ldQ2jb(fc-b7<WH_9-GpB%1iB9sOu%o4m)Ls8E^
zTfa~5@<Vu~+eo|+HN1M9e=D?5AWzNhB=Fkm(E)s<+-)|T79$5;`Tl{&5dWdODqR0l
zF>X3*+wi)>P$99+`0JQwIaoGTy{IPi-UsNJ+hf)v80n@IuhJ8W+E(T_9!1FPvuw~U
z#Tu2T9-4D`W>CmL_COYQw)m@$F%h3lNdyldZ<YD7W+W*qrzv*AVj9~dWhl3spibdr
zaQ4BpxR$21Ah&K%T)_%<KzqsCk(H3nhl^pWH1ap7lYjbZ)x6odru}q!>5_Ls^*NhJ
z)e{{Cy)>$<l?Awn6S#p77a&IH*&9b-utgkVc*+D@ENt(Ai_YKx;&jtfbFZoNa$KpD
zs+PWiA>3&tW$jcR>?iRkF8(6S#5@7RMQsWa5wMKqVfK%6j|co?Y2Tx!V(+RzEeWV0
zcrFP}cpULaDqw3z96mYCKSVS}WBk_kR9aX(BJ`VOEY0LfK>_oaj)l;eP;&uAOVm-R
zU-jMA7^C)y21fpxcU(T!Z6Cz{49BNbtTNH3wMR{Dhe#I~$Lu7E-ijCs#_b*YV&<8^
z_C^^_Kx7()WLCE=p~23D<a9<hqhgYk{WAbM#t9(mB>rp^Gz&5v@`}!=X{JC};K(rf
zM%K*vWu@8%*g*3|#DsEIKp|d0xW*qqT`j=a+4Kk=Q`>^B+V*ViAoNWi1_$;ox<d*l
z^6n$)RoM5ey!oWtI1sA(G}RJp;1?eO{M}eiwQP|%^AaadlFz;*$|yd0U-lT=3UMBL
zM74A5GO>6!A~fP*b`CYl;_63Mwlvg~#2NUO<Lc9vt1A?nO%(a{Tq!c=uvooA_pU=C
z#os1o7nCZDbv40%L*`>T+wT;snaEe6RWbO<bp3tPRPg13Vufg6Vu-;`z4opie#O(y
z2GimWxos!+rnyyB=PStjko40nF{#<WbbBYE!u8-4z&NW$Q!93Hf@A8irH5|hTfobM
zCmq%c*kr|*G`;fz#}Ns>G%A6y5=ygLC`Kr#0d5C6T~yq~*!B?3&p!6LNrwXcbg|lr
zWj#lOG*&jsP7to-=gW}TRoXjv^G3e1`grEvqnWHX#-17(-AVPSFFtGr=wPl-VAr8g
z)_6)+SCT`5oO|&R!kH&;&ScN#xmJ4%M=?FzeFkMywG|hugn1q15()MF+k{!&qnSCV
zzs?FrxsS2KK3BRLKQ3)_8);YsfOB@*&H-l>t|wFlPu}KnS?z_7GhdF=+Nq=)rdHTD
z)a*KaFP~0*QGu1j9yNGvFy4&}8zeeQ7wO2y?=lBo=XbtU;3jG}fkNUTOp16EwOiPs
zSW!n-u+&)i<XwBttDB?NRy2xMwQeaRh+#@y^qiHb+z_7S#nh1u9Qx<(cgs0&dj(Cd
zR~JxP+)T7W;H!(KY{6;mEH>TqT?Q1mCv6V1#ad_iHrQrt#;JxR!v55ED9BjWEy;Hq
zwE<zFQa9}k_!xtnyv6ktWa-hj;Pn82!FU;jnUth%PLWsN&|uundNvhj(hl0OW3Wq@
znumtW<NEvNX5B}JSEg(vM&e<W;|;WEM;T8(jU+y9z0y=eiHvR%#%Q<jWk#{Vm{rS5
z=F>K=y2#B0L_CidROt9pZ}+}f(*vitPorU^4C`hUT;;0S2X7(lf#cY#b1+cg+@^2@
z=DTqt>ubJTEr>B%F5dfOG(6u9S2iyq<e(4?uAsx;km^q2ZhEh&Rh8H+%7AqyYE<yZ
zJ(CZ((nVH!=R+69P4}X%e!1%7xOmp!Et*ucz_)kbP=FW33N;5VLc7s$qAF5{ReoEA
z7yKG~9*-Vg1T-DoyN2MQpm3T0?M2{W<_`#RG6(!I;96Cmwwz&m>b9>!V?ZS4WlB~^
ze?oPNYca|O8%`t^@z^bcOrqEK8Ep7e#nFWJrdd;3@(BHA=(C#IA}sB2B57|@;z5*;
zM!u~$-A2z<a1t#To_|<fmti<d#2FyLj^N>Sm}|V~=Cz|!)=n%TeT9yyj~#_GkV(D5
z+O7xpwFV)-8<@S~)D=oZ+$ZNNOqm{v#lbG1i$SoJnTK9zo>vXWo=MgiuHb`JNftnt
zU0l#2y&03ftElS(D-*wj7*FL&mJ`pwuBYZjx~Xa-EPlH;qb8~LA*pKCa(%j$b2;}p
zO|;~@L3J=m0JVe+D^)O$REfHjV++rB6t>o9(iQJ?<H=V6_Ma{cOhbiui?Q*MJ)7;7
zG-VXNMVhpalPBV9#e+Vq83&!OaEvb4=iqmY3a&C(_85v4iC(*GHF%2Oj0ddORWD6`
ztH7)jR%#Pxbc)BSeLeW1EQ`u(>x&uid85aQ(%_S0dwSR2Pi^&&LQcPu2$A%1^!kcH
zYc4$c+Kz{H2-gqE5@jWH!G9hoX(q<RJNbwk*+p@jDScT~sUO$4XW8*7(i*<AwfW*u
zHJ*1Be4%D#@?B>3+y3hc6{7FFJ|&`18?$5@NmJWIx`3A|okhgZ+E_6dtUfu|$QrnD
z=z^29ienC9f`Zf~&!qZI$~15lTVvu8qgS396d>S@6h_durQ+MR$B!q)0U?Etoyy*4
z+r-i(MAdL&z*QM~6Yq3zr0Lly19-T3@{^*%)(0FCi2Pv0MBqZvog471{PS?w+c|fT
zP(MyQd!_E@-GwmH3FlRC*ihxMxemirwdQt8-P7)g9wOV)7J*uX>#91Mw8TGTwd!7m
z8}pi63;9zS_pp)(sfcsK<aEND#SfjYxApn`gXm?GP<<m39fu&|ZLvC^5^jQ%beVc&
z;Lqe@oWNr=g3#oXpX!SCf(Y)M>+{lsq8Mdl#kaSnjEkg4VyA-`GK0OI`=v*w>+NEP
zHZnxRNF16X4W;;CSxS8ns>vGww_kJg+DGOTK%Rf|tt^%ODuCoApN83CT?9|uH4qL~
z;`i7G!D>}w%JJE=u#9am51pAYD^Rmx@YnC;tE}R*t7uWECEB}7*Q)P+qo=O`!l`N}
zDz%$wwN`4%4Ihp;aHXi-M22b3lfCY0Cnb}I+ICKTl&CDKXp2faeljHQ<#E}=UN^yj
zgs5^t9UeV_BWs<g#bjDw!riZf7p!q}6kSn}u5SQ-T1P7Qjh0e~$s83ae(O+)n=odk
zHy~D;lNx!}a?65s!<3mhiouAc*=|N2&aL}x`1yCqLZxlRi6}Fbu@G&VCM;eYs-1%_
zeR)=OJR!;J^KH>wS6fSHG%VP{NqY@1CsA$K+y0gtaRgJ5mYvUUQ%kkjDU)tD3T|(X
z_78HZ@#4>3Ufj_i?k}rugf?!MsJaet&d;yg-9@Zi>nysieY*A%tBdy2lJZwY<UL}z
z5M>wLj*HR+1>1A%kffwUhL&!1UP2fEPE#zlM$(d-eqOs9o>WgM2Ugt<9KNkt8GRg{
zXWueWvf*NnK4-_=%#zt-S8M3V6qzfUWE#<z4vDQQ9CU>6&4FY*UjS)9--rT??ew2~
z40Ks%#lA{pF58X1!Ca~&s9Pr<V;j?zCQTz{x@4*D)9-uTUDAReBF^)8GF-I*S0vMb
z#P4!yJYegNd(4-Xd{aG%JG6ZXjm*e~+J35Si#=vG%{CfxDaqd|=JvUkuv0bFYfeoL
zAy^!!@@7fC0Vq%uXd8?iX#ULWwU96g4%+F;n+uuE8Vuc&uQv4<RdNDz`HeRjp#!Qb
zGySow=`Dqnta~v6uZ<uoIL~D&nLIYnf{wY0g|<YoW^fQhEw8l1Q=Q%Xi_~6E@5lSo
zSi`+>OaImhJK-5$qN>F>6BW&!T%hYvr`*8$nE`HZn?p)#c6$nOejo%)Gdj@y^qsEn
z5=_U<DN*rI;;<xn<|RmXe#~(h4)WR6m4uA?61%eS!1-m&LQQ-0GAFA|w_`ygm~Kt9
z_c2?({KZc5coR#?8=I5N*<|=^5d`6V#oXvlau-6O#!HJD6QYz*DI0n*=d0_T^o=8f
z?KQtskupWzuvfZDep>Ch@KITAcDhF>D5O<A*O1uTRNu5)l>)}1q5S(45OXoNY6vF5
zGcI=(8J2Mt^r0NhGrC%Tyt+&3is;L?v06&4`OO96FScWK6f4B$&Ai8#56Dfm#|f>^
z80t0{%`C4KwX}Lt&`4cQtG|j9n_$TEdd!lfRmhu(5vB}m$-Y?}R!5>@LAJGHhKBD&
zlhNe2hQt%kCvWpNYHOr==35zdEbJIUY7z65;6E-Nm~15XMP~;rN3?G)%G086@!O9I
zE{@O5Qv`I!j>BxNc~oe!c|j(i+l$x}XzLdV>(p=HA3|v$Ns2JWeJHIY`?pZ~NA3;)
z(tQ|E2SrGRj51*QoL&2O^^F~nB(ju*2f{qlV|*rt*-|8tI33thLfsh~>G3Wp(}zy=
z^1SXW!6nh)-MR>T@q#))I#^Xq4oYN7B9=>3+kw#9JIS9tmiJ{8jGWpes}7H4b99TW
zsXC+fY7S6jqExAn0RUb2U48;1rb0+_RF5exs!Vy!DENInKcc{8Bia>Zc@a5{v;xy?
zkue$lOfJwljY|AY>&DPCY^Ksu{jYaVSWCyl3M!tbqbCc#vria#5<5t(RhKt%2KQpZ
z4+T+Ddr1ZMMgx;(S(`Q?U2k^pMS52jLl0%KL{KJpL_?+C;}LMj+bT1c*)}?6J8N_Z
z2Q5q1x`!<$D<+$J0QQ}5q%Y5V+?g~rvUYV0{T_yNZtPE;Le?)WBPb0ct*UU;N+wbB
zUVX^5fz;%x95^vtc7@Xibn>570S=?T4Pnz<AZs7v6t#H9i?}Mo!zX8Hg>{Dd%5@PY
zMQsjT?Tgoku(x1+;pe=9&Y*~7J6_bv2Kij*9c3@?SYPC6@MoAegtuf|9i}{zoW@B`
zN9a$3DO3rf2rx!VAE#0f@(-Nhp{hB9#%lWGt*J&Ihoj;V6qNgY0`&LOCl3?V7P^)u
zh5#Ul&ep)NKce3f%<%NXK93JzVB!ruEUYbMPmO7EGpnyEEd62Jv$@9L2xYm9yshMK
z%S9s0^TFluySn&Y0i`)DiY)rC$$G?$D>kXQ>gZ}Rr@5vkDya$M2o6lanaNbO4wh7$
z+Mj)B*-^{~%hD(v<BoXV)f2ZsIBdjFI?96)5MKrfn4GM*;&KiwAsVX4fwogZoz94O
z17^8$Q%aF^E9DzLRuYAIy$M2;>iGZy<JZbawljY|lH+Vi-CgR`X&42k@#uV~4j!DC
z^Fzs6eZSD~Wls}4?;2(7yr=JMH`;zMPZFt<z_j;ntB%|0ZvB>S@D69GxZGaI{LMhR
z(Pij*`5~c_Xz@c&DL*qFfxUMh_)qu!$a3F%>`!^fNDEq78k!i{+5VB=1~Sp#=d<nf
zbZzhFuzt;r|EY3+@=scZ=2p5Ozz?l|$|+F%Ls`1|`T%plebH97^!j$Tw*NZ|27TY+
z@DIHh>hZ(wUVi^)2=DLyy$U84x<&vXy?_bGLf0DjFEzpa2W4w3Ydh-)C;UI^0YSPT
zJK(>24*#FZv$Zm^1pxmKn*Zc;kd>ABznAw<^?T5=*0ltf|8M>NsjaZ?|EILD;#*pA
z84>zF``0~A3l)1UTZx|5uT#?$MWy!V3fS1@e4`iR_gty~!yxNLQ{G-p$sOh7DVD_H
z)p2t(e`z)aA~a43l0Pp&r$1Cv^kl0<y%KRXpWjv~QKMIGN!j*PO~!^yd<3E5$uVsi
zcZG>yc(P~^VCdjkAbJ*fES<=0;%nik`k$oLZrBY7@}90Qdhhw4*)i3deW(?-6^A!Y
z%ykkQLWl@=82gT(;9|Zy{H6Z%^=kkF?JksR4ySTu-7~&nj(7g=BxAoc>9)g;g5$Qz
z6cJ)NCqwZEkKmYFy&#H7!saB0<#(b-&(L1Jv|@*f?~HI?H?XYZwjA<vjw?LCE&;lA
ziz2C0>C-yCXpM=Y<-Fqs-<ix_P~|Kbyv3&2Wo;#Gly7SEK6(D|;3`n%CcnDxd%`de
z5ALJKm{7l8V{-54U*yM%lfP^IeHF=16~+70->oV6VdX~)|2iZ60M4Isd%rs6*L>gK
zHGdt#KFnDE6ytkCzsz6%UGdj|{V>A)DK_^$`8nSGyWX$#{{sYnD%RuO6aRmM``@*G
zh2CE)m`(84IPpXCcToJZIH<Lk|IL@5i2Q5g{?z&pB!6Dx-(%|Ul=_Lx|5*9aOb=Q=
zxc30fpCbPcR(^xff2-gFGJi_$KN$Tlfd0Oo4~Y3GYyV*ESD^mi+IdOz-~Rd)u)nwS
zfY_g+O!VJ&tp5Vs-&=Y>?oSCN`fp1=0sMO_fAvqtZ>;<P@_+37coKdE$peyq$}Q1v
Y04*yCd*7F#pwRC>*7y9;BYwF09}%!5r2qf`

literal 0
HcmV?d00001

diff --git a/messytables/ods.py b/messytables/ods.py
index 7b03d74..e9d56b4 100644
--- a/messytables/ods.py
+++ b/messytables/ods.py
@@ -6,17 +6,42 @@
 
 from messytables.core import RowSet, TableSet, Cell
 from messytables.types import (StringType, DecimalType,
-                               DateType)
+                               DateType, BoolType, CurrencyType,
+                               TimeType, PercentageType)
 
 
-ODS_NAMESPACES_TAG_MATCH = re.compile(b"(<office:document-content[^>]*>)", re.MULTILINE)
-ODS_TABLE_MATCH = re.compile(b".*?(<table:table.*?<\/.*?:table>).*?", re.MULTILINE)
+ODS_NAMESPACES_TAG_MATCH = re.compile(
+    b"(<office:document-content[^>]*>)", re.MULTILINE)
+ODS_TABLE_MATCH = re.compile(
+    b".*?(<table:table.*?<\/.*?:table>).*?", re.MULTILINE)
 ODS_TABLE_NAME = re.compile(b'.*?table:name=\"(.*?)\".*?')
-ODS_ROW_MATCH = re.compile(b".*?(<table:table-row.*?<\/.*?:table-row>).*?", re.MULTILINE)
+ODS_ROW_MATCH = re.compile(
+    b".*?(<table:table-row.*?<\/.*?:table-row>).*?", re.MULTILINE)
+
+NS_OPENDOCUMENT_PTTN = u"urn:oasis:names:tc:opendocument:xmlns:%s"
+NS_CAL_PTTN = u"urn:org:documentfoundation:names:experimental:calc:xmlns:%s"
+NS_OPENDOCUMENT_TABLE = NS_OPENDOCUMENT_PTTN % "table:1.0"
+NS_OPENDOCUMENT_OFFICE = NS_OPENDOCUMENT_PTTN % "office:1.0"
+
+TABLE_CELL = 'table-cell'
+VALUE_TYPE = 'value-type'
+COLUMN_REPEAT = 'number-columns-repeated'
+
+ODS_VALUE_TOKEN = {
+    "float": "value",
+    "date": "date-value",
+    "time": "time-value",
+    "boolean": "boolean-value",
+    "percentage": "value",
+    "currency": "value"
+}
 
 ODS_TYPES = {
     'float': DecimalType(),
-    'date': DateType(None),
+    'date': DateType('%Y-%m-%d'),
+    'boolean': BoolType(),
+    'percentage': PercentageType(),
+    'time': TimeType()
 }
 
 
@@ -102,13 +127,13 @@ def __init__(self, sheet, window=None, namespace_tags=None):
         else:
             namespaces = {
                 "dc": u"http://purl.org/dc/elements/1.1/",
-                "draw": u"urn:oasis:names:tc:opendocument:xmlns:drawing:1.0",
-                "number": u"urn:oasis:names:tc:opendocument:xmlns:datastyle:1.0",
-                "office": u"urn:oasis:names:tc:opendocument:xmlns:office:1.0",
-                "svg": u"urn:oasis:names:tc:opendocument:xmlns:svg-compatible:1.0",
-                "table": u"urn:oasis:names:tc:opendocument:xmlns:table:1.0",
-                "text": u"urn:oasis:names:tc:opendocument:xmlns:text:1.0",
-                "calcext": u"urn:org:documentfoundation:names:experimental:calc:xmlns:calcext:1.0",
+                "draw": NS_OPENDOCUMENT_PTTN % u"drawing:1.0",
+                "number": NS_OPENDOCUMENT_PTTN % u"datastyle:1.0",
+                "office": NS_OPENDOCUMENT_PTTN % u"office:1.0",
+                "svg": NS_OPENDOCUMENT_PTTN % u"svg-compatible:1.0",
+                "table": NS_OPENDOCUMENT_PTTN % u"table:1.0",
+                "text": NS_OPENDOCUMENT_PTTN % u"text:1.0",
+                "calcext": NS_CAL_PTTN % u"calcext:1.0",
             }
 
             ods_header = u"<wrapper {0}>"\
@@ -129,19 +154,61 @@ def raw(self, sample=False):
             block = self.namespace_tags[0] + row + self.namespace_tags[1]
             partial = io.BytesIO(block)
 
-            for action, elem in etree.iterparse(partial, ('end',)):
-                if elem.tag == '{urn:oasis:names:tc:opendocument:xmlns:table:1.0}table-cell':
-                    cell_type = elem.attrib.get('urn:oasis:names:tc:opendocument:xmlns:office:1.0:value-type')
-                    children = elem.getchildren()
-                    if children:
-                        c = Cell(children[0].text,
-                                 type=ODS_TYPES.get(cell_type, StringType()))
-                        row_data.append(c)
-
-            if not row_data:
+            for action, element in etree.iterparse(partial, ('end',)):
+                if element.tag != _tag(NS_OPENDOCUMENT_TABLE, TABLE_CELL):
+                    continue
+
+                cell_type = element.attrib.get(
+                    _tag(NS_OPENDOCUMENT_OFFICE, VALUE_TYPE))
+                value_token = ODS_VALUE_TOKEN.get(cell_type, 'value')
+                repeat = element.attrib.get(
+                    _tag(NS_OPENDOCUMENT_TABLE, COLUMN_REPEAT))
+                if cell_type == 'string':
+                    cell = _read_text_cell(element)
+                elif cell_type == 'currency':
+                    value = element.attrib.get(
+                        _tag(NS_OPENDOCUMENT_OFFICE, value_token))
+                    currency = element.attrib.get(
+                        _tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
+                    cell = Cell(value + ' ' + currency,
+                                type=CurrencyType())
+                elif cell_type is not None:
+                    value = element.attrib.get(
+                        _tag(NS_OPENDOCUMENT_OFFICE, value_token))
+                    cell = Cell(value,
+                                type=ODS_TYPES.get(cell_type, StringType()))
+                else:
+                    cell = Cell('', type=StringType())
+                if repeat:
+                    number_of_repeat = int(repeat)
+                    row_data += [cell] * number_of_repeat
+                else:
+                    row_data.append(cell)
+
+            empty_cells = [c for c in row_data if c.value == '']
+            if len(empty_cells) == len(row_data):
                 # ignore blank lines
                 continue
 
             del partial
             yield row_data
         del rows
+
+
+def _read_text_cell(element):
+    children = element.getchildren()
+    text_content = []
+    for child in children:
+        if child.text:
+            text_content.append(child.text)
+        else:
+            text_content.append('')
+    if len(text_content) > 0:
+        cell_value = '\n'.join(text_content)
+    else:
+        cell_value = ''
+    return Cell(cell_value, type=StringType())
+
+
+def _tag(namespace, tag):
+    return '{%s}%s' % (namespace, tag)
diff --git a/messytables/types.py b/messytables/types.py
index ba017f3..04dd234 100644
--- a/messytables/types.py
+++ b/messytables/types.py
@@ -103,6 +103,27 @@ def cast(self, value):
             return decimal.Decimal(value)
 
 
+class PercentageType(DecimalType):
+    """ Decimal number, ``decimal.Decimal`` or float numbers. """
+    guessing_weight = 0
+
+    def cast(self, value):
+        result = DecimalType.cast(self, value)
+        if result:
+            result = result/decimal.Decimal('100')
+        return result
+
+
+class CurrencyType(DecimalType):
+    guessing_weight = 0
+    result_type = decimal.Decimal
+
+    def cast(self, value):
+        value_without_currency = value.split(' ')[0]
+        return DecimalType.cast(self,
+                                value_without_currency)
+
+
 class FloatType(DecimalType):
     """ FloatType is deprecated """
     pass
@@ -134,6 +155,25 @@ def cast(self, value):
         raise ValueError
 
 
+class TimeType(CellType):
+    result_type = datetime.time
+
+    def cast(self, value):
+        if isinstance(value, self.result_type):
+            return value
+        if value in ('', None):
+            return None
+        hour = int(value[2:4])
+        minute = int(value[5:7])
+        second = int(value[8:10])
+        if hour > 24:
+            return datetime.timedelta(hours=hour,
+                                      minutes=minute,
+                                      seconds=second)
+        else:
+            return datetime.time(hour, minute, second)
+
+
 class DateType(CellType):
     """ The date type is special in that it also includes a specific
     date format that is used to parse the date, additionally to the
@@ -186,7 +226,7 @@ class DateUtilType(CellType):
 
     def test(self, value):
         if len(value) == 1:
-             return False
+            return False
         return CellType.test(self, value)
 
     def cast(self, value):
@@ -195,7 +235,8 @@ def cast(self, value):
         return parser.parse(value)
 
 
-TYPES = [StringType, DecimalType, IntegerType, DateType, BoolType]
+TYPES = [StringType, DecimalType, IntegerType, DateType, BoolType,
+         TimeType, CurrencyType, PercentageType]
 
 
 def type_guess(rows, types=TYPES, strict=False):
diff --git a/test/test_read.py b/test/test_read.py
index f4b73d1..c09a727 100644
--- a/test/test_read.py
+++ b/test/test_read.py
@@ -1,6 +1,7 @@
 # -*- coding: utf-8 -*-
 import unittest
 
+from decimal import Decimal
 from . import horror_fobj
 from nose.plugins.attrib import attr
 from nose.tools import assert_equal
@@ -22,6 +23,7 @@
 import datetime
 stringy = type(u'')
 
+
 class ReadCsvTest(unittest.TestCase):
     def test_utf8bom_lost(self):
         fh = horror_fobj('utf8bom.csv')
@@ -188,7 +190,8 @@ def rows(skip_policy):
             row_set = table_set.tables[0]
             return row_set
 
-        second = lambda r: r[1].value
+        def second(row):
+            return row[1].value
 
         assert "goodbye" in list(map(second, rows(True)))
         assert "    goodbye" in list(map(second, rows(False)))
@@ -308,9 +311,9 @@ def test_read_large_ods(self):
         assert_equal(6, len(table_set.tables))
         row_set = table_set.tables[0]
         row = next(row_set.raw())
-        assert len(row) == 5, len(row)
+        assert len(row) == 16384, len(row)
         for row in row_set.sample:
-            assert len(row) == 5, len(row)
+            assert len(row) == 16384, len(row)
 
     def test_ods_version_4412(self):
         fh = horror_fobj('loffice-4.4.1.2.ods')
@@ -334,6 +337,94 @@ def test_ods_read_past_blank_lines(self):
         assert_equal(rows[2][0], 'Jane')
         assert_equal(rows[3][0], 'Ian')
 
+    def test_ods_read_all_supported_formats(self):
+        fh = horror_fobj('ods_formats.ods')
+        table_set = ODSTableSet(fh)
+        assert_equal(3, len(table_set.tables))
+        row_set = table_set.tables[0]
+        rows = row_set_to_rows(row_set)
+        assert_equal(rows[0][0], "Date")
+        assert_equal(rows[1][0], "2014-11-11")
+        assert_equal(rows[2][0], "2001-01-01")
+        assert_equal(rows[3][0], '')
+        # time formats
+        assert_equal(rows[0][1], "Time")
+        assert_equal(rows[1][1], "PT11H12M12S")
+        assert_equal(rows[2][1], "PT00H00M12S")
+        assert_equal(rows[4][1], 'PT27H17M54S')
+        assert_equal(rows[5][1], "Other")
+        # boolean
+        assert_equal(rows[0][2], "Boolean")
+        assert_equal(rows[1][2], 'true')
+        assert_equal(rows[2][2], 'false')
+        # Float
+        assert_equal(rows[0][3], "Float")
+        assert_equal(rows[1][3], '11.11')
+        # Currency
+        assert_equal(rows[0][4], "Currency")
+        assert_equal(rows[1][4], '1 GBP')
+        assert_equal(rows[2][4], '-10000 GBP')
+        # Percentage
+        assert_equal(rows[0][5], "Percentage")
+        assert_equal(rows[1][5], '2')
+        # int
+        assert_equal(rows[0][6], "Int")
+        assert_equal(rows[1][6], '3')
+        assert_equal(rows[4][6], '11')
+        # Scientifed not supported
+        assert_equal(rows[1][7], '100000')
+        # Fraction
+        assert_equal(rows[1][8], '1.25')
+        # Text
+        assert_equal(rows[1][9], "abc")
+
+    def test_ods_read_all_supported_formats_casted(self):
+        fh = horror_fobj('ods_formats.ods')
+        table_set = ODSTableSet(fh)
+        assert_equal(3, len(table_set.tables))
+        row_set = table_set.tables[0]
+        rows = cast_row_set_to_rows(row_set)
+        date_format = "%d/%m/%Y"
+        assert_equal(rows[0][0], "Date")
+        assert_equal(rows[1][0].strftime(date_format), "11/11/2014")
+        assert_equal(rows[2][0].strftime(date_format), "01/01/2001")
+        assert_equal(rows[3][0], '')
+        # time formats
+        time_format = "%S:%M:%H"
+        assert_equal(rows[0][1], "Time")
+        assert_equal(rows[1][1].strftime(time_format), "12:12:11")
+        assert_equal(rows[2][1].strftime(time_format), "12:00:00")
+        assert_equal(rows[3][1], 0)
+        assert_equal(rows[4][1], datetime.timedelta(hours=27,
+                                                    minutes=17,
+                                                    seconds=54))
+        assert_equal(rows[5][1], "Other")
+        # boolean
+        assert_equal(rows[0][2], "Boolean")
+        assert_equal(rows[1][2], True)
+        assert_equal(rows[2][2], False)
+        # Float
+        assert_equal(rows[0][3], "Float")
+        assert_equal(rows[1][3], Decimal('11.11'))
+        # Currency
+        assert_equal(rows[0][4], "Currency")
+        assert_equal(rows[1][4], Decimal('1'))
+        assert_equal(rows[2][4], Decimal('-10000'))
+        # Percentage
+        assert_equal(rows[0][5], "Percentage")
+        assert_equal(rows[1][5], Decimal('0.02'))
+        # int
+        assert_equal(rows[0][6], "Int")
+        assert_equal(rows[1][6], 3)
+        assert_equal(rows[4][6], 11)
+        # Scientifed not supported
+        assert_equal(rows[1][7], 100000)
+        # Fraction
+        assert_equal(rows[1][8], 1.25)
+        # Text
+        assert_equal(rows[1][9], "abc")
+
+
 def row_set_to_rows(row_set):
     rows = []
     for row in row_set:
@@ -341,6 +432,13 @@ def row_set_to_rows(row_set):
     return rows
 
 
+def cast_row_set_to_rows(row_set):
+    rows = []
+    for row in row_set:
+        rows.append([cell.type.cast(cell.value) for cell in row])
+    return rows
+
+
 class XlsxBackwardsCompatibilityTest(unittest.TestCase):
     def test_that_xlsx_is_handled_by_xls_table_set(self):
         """
@@ -573,8 +671,8 @@ def setUp(self):
                 PDFTableSet(fh)
             except ImportError:
                 # Optional library isn't installed. Skip the tests.
-                raise SkipTest("pdftables is not installed, skipping PDF tests")
-
+                raise SkipTest(
+                    "pdftables is not installed, skipping PDF tests")
 
     def test_read_simple_pdf(self):
         with horror_fobj('simple.pdf') as fh:

From 56d91521d0b4f0bde30ba4bfb5e60ec8ea20b31c Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Wed, 1 Feb 2017 21:20:48 +0000
Subject: [PATCH 02/18] fix python 2.6 unit test failure on testing ods casted
 decimal values

---
 test/test_read.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/test_read.py b/test/test_read.py
index c09a727..dcb23ee 100644
--- a/test/test_read.py
+++ b/test/test_read.py
@@ -420,7 +420,7 @@ def test_ods_read_all_supported_formats_casted(self):
         # Scientifed not supported
         assert_equal(rows[1][7], 100000)
         # Fraction
-        assert_equal(rows[1][8], 1.25)
+        assert_equal(rows[1][8], Decimal('1.25'))
         # Text
         assert_equal(rows[1][9], "abc")
 

From 9ae2ed6863d36663641d84dcc5b1919df31a0fa4 Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Wed, 1 Feb 2017 21:24:51 +0000
Subject: [PATCH 03/18] fix TimeType to return a datetime.time if hour is less
 than 24

---
 messytables/types.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/messytables/types.py b/messytables/types.py
index 04dd234..3c70ebe 100644
--- a/messytables/types.py
+++ b/messytables/types.py
@@ -166,12 +166,12 @@ def cast(self, value):
         hour = int(value[2:4])
         minute = int(value[5:7])
         second = int(value[8:10])
-        if hour > 24:
+        if hour < 24:
+            return datetime.time(hour, minute, second)
+        else:
             return datetime.timedelta(hours=hour,
                                       minutes=minute,
                                       seconds=second)
-        else:
-            return datetime.time(hour, minute, second)
 
 
 class DateType(CellType):

From 92b95dc4d9b7d5b82d3cb20f43cc7d7ca0b3e945 Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Thu, 2 Feb 2017 09:23:37 +0000
Subject: [PATCH 04/18] performance improvement on
 test/test_read.py:ReadODSTest.test_read_large_ods: 45s down to 9s

---
 messytables/ods.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/messytables/ods.py b/messytables/ods.py
index e9d56b4..f766661 100644
--- a/messytables/ods.py
+++ b/messytables/ods.py
@@ -153,6 +153,7 @@ def raw(self, sample=False):
 
             block = self.namespace_tags[0] + row + self.namespace_tags[1]
             partial = io.BytesIO(block)
+            empty_row = True
 
             for action, element in etree.iterparse(partial, ('end',)):
                 if element.tag != _tag(NS_OPENDOCUMENT_TABLE, TABLE_CELL):
@@ -165,6 +166,8 @@ def raw(self, sample=False):
                     _tag(NS_OPENDOCUMENT_TABLE, COLUMN_REPEAT))
                 if cell_type == 'string':
                     cell = _read_text_cell(element)
+                    if cell.value != '':
+                        empty_row = False
                 elif cell_type == 'currency':
                     value = element.attrib.get(
                         _tag(NS_OPENDOCUMENT_OFFICE, value_token))
@@ -172,11 +175,13 @@ def raw(self, sample=False):
                         _tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
                     cell = Cell(value + ' ' + currency,
                                 type=CurrencyType())
+                    empty_row = False
                 elif cell_type is not None:
                     value = element.attrib.get(
                         _tag(NS_OPENDOCUMENT_OFFICE, value_token))
                     cell = Cell(value,
                                 type=ODS_TYPES.get(cell_type, StringType()))
+                    empty_row = False
                 else:
                     cell = Cell('', type=StringType())
                 if repeat:
@@ -185,8 +190,7 @@ def raw(self, sample=False):
                 else:
                     row_data.append(cell)
 
-            empty_cells = [c for c in row_data if c.value == '']
-            if len(empty_cells) == len(row_data):
+            if empty_row:
                 # ignore blank lines
                 continue
 

From 0fc889f2a7ff8729dc20b5bc611d4fa61d309ecf Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Thu, 2 Feb 2017 09:32:00 +0000
Subject: [PATCH 05/18] code refactoring: reduce the length of ODSRowSet.raw()

---
 messytables/ods.py | 51 +++++++++++++++++++++++++---------------------
 1 file changed, 28 insertions(+), 23 deletions(-)

diff --git a/messytables/ods.py b/messytables/ods.py
index f766661..6f6708b 100644
--- a/messytables/ods.py
+++ b/messytables/ods.py
@@ -159,31 +159,12 @@ def raw(self, sample=False):
                 if element.tag != _tag(NS_OPENDOCUMENT_TABLE, TABLE_CELL):
                     continue
 
-                cell_type = element.attrib.get(
-                    _tag(NS_OPENDOCUMENT_OFFICE, VALUE_TYPE))
-                value_token = ODS_VALUE_TOKEN.get(cell_type, 'value')
+                cell = _read_cell(element)
+                if cell.value != '':
+                    empty_row = False
+
                 repeat = element.attrib.get(
                     _tag(NS_OPENDOCUMENT_TABLE, COLUMN_REPEAT))
-                if cell_type == 'string':
-                    cell = _read_text_cell(element)
-                    if cell.value != '':
-                        empty_row = False
-                elif cell_type == 'currency':
-                    value = element.attrib.get(
-                        _tag(NS_OPENDOCUMENT_OFFICE, value_token))
-                    currency = element.attrib.get(
-                        _tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
-                    cell = Cell(value + ' ' + currency,
-                                type=CurrencyType())
-                    empty_row = False
-                elif cell_type is not None:
-                    value = element.attrib.get(
-                        _tag(NS_OPENDOCUMENT_OFFICE, value_token))
-                    cell = Cell(value,
-                                type=ODS_TYPES.get(cell_type, StringType()))
-                    empty_row = False
-                else:
-                    cell = Cell('', type=StringType())
                 if repeat:
                     number_of_repeat = int(repeat)
                     row_data += [cell] * number_of_repeat
@@ -199,6 +180,30 @@ def raw(self, sample=False):
         del rows
 
 
+def _read_cell(element):
+    cell_type = element.attrib.get(
+        _tag(NS_OPENDOCUMENT_OFFICE, VALUE_TYPE))
+    value_token = ODS_VALUE_TOKEN.get(cell_type, 'value')
+    if cell_type == 'string':
+        cell = _read_text_cell(element)
+    elif cell_type == 'currency':
+        value = element.attrib.get(
+            _tag(NS_OPENDOCUMENT_OFFICE, value_token))
+        currency = element.attrib.get(
+            _tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
+        cell = Cell(value + ' ' + currency,
+                    type=CurrencyType())
+    elif cell_type is not None:
+        value = element.attrib.get(
+            _tag(NS_OPENDOCUMENT_OFFICE, value_token))
+        cell = Cell(value,
+                    type=ODS_TYPES.get(cell_type, StringType()))
+    else:
+        cell = Cell('', type=StringType())
+
+    return cell
+
+
 def _read_text_cell(element):
     children = element.getchildren()
     text_content = []

From 927ed7157f136bd00d9acfa2c4a1bfe5dab1a31e Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Thu, 2 Feb 2017 09:39:18 +0000
Subject: [PATCH 06/18] add a unit test on reading multi line text cell in ods
 format

---
 horror/multilineods.ods | Bin 0 -> 8030 bytes
 test/test_read.py       |   7 +++++++
 2 files changed, 7 insertions(+)
 create mode 100644 horror/multilineods.ods

diff --git a/horror/multilineods.ods b/horror/multilineods.ods
new file mode 100644
index 0000000000000000000000000000000000000000..c03f6456a3144882c3462a9ebbb1ba0700abcc85
GIT binary patch
literal 8030
zcmeHMc|27A*B{x}WGj(`B#f~$vL#u=SW3zmV~jDDnPJ9Sl6}jb$dWACDiS3-5mH$~
z$d;v$Eo(`5W|S}W)A#i}&%e*}nb)1U_uO;d=X}one9k>{4n&unf(-zm0RRMw6}18!
zz@cIQ0DyE5NdQN*Ba+~OL&9-52Q&gsKw~i=SB$k77LG^b#jrRe#u|%oaYSMWVtAZ0
z5^jyRMIs51Ele63nk`Hs%s=A<@$3v1V}nMyIB%ihB|&fm0_lJxUSgd=2p4B(5}3QA
z133VqOHHx?u<w`nPuOi!e-XESjuOBBghe~TQAj*U4NY)_<M5!r3m@7;*v0`1Cm@~w
zMT&b5DV#GF<&4DRt>FJAxtBho2VMSM>hNCrus9bSN%UXr+)EBmfD>H)7giFmScm_)
zl_V(~9D{WD!%Dl=fPzMH0DyFmd}3nj;%J3|qaE-d!uF{c4ucAT=xZ|49i}6)FluY5
z8xen>0RUtG8q&ImtQJuO0LUY>)lZwaC(rarICntUR1J}V_YMhPV7_1}EPU&_ZgyHm
zrmmlE5ryX$y61F~((@J+lC;uKOH*FI28X{l?T_BrjJXp&DqmBrm~;iOvELwVTCM;C
z>vxPVVygS7;Q#nC)T^TU+A4e&{>(fGa;wj&^zy?`cLL~b!qxcLVpP&ZOZvd$AD)+H
zM9~DO3kd{4R+3Cxu~qx)%CQ;zauJDYz258dEAWcT)3QnHkIY{yvBcZbAqxldy`ipk
zVNZ1AXPv@{N-BA(!tb0)77vg-zp&c9yxe6O{oKh|=+%t8g5#)YAA%(<O3^D)5@^Jo
zfqOg5{c&{-9sPL7!L`olBsX4O(@c~}E^f_W|5WMtu$t0C!>p86%#-(wJj&d7_a()u
z&!rnnIJvKi<~c)}-`p}#OsHO-*e!d#)#mT>jw=Q}9_qiwNmWhf+&r2cRZ_RMAA}v}
z34FD@X50iy#bUBRMuBpJE4a=f{LRU6L>%t<5Hu1-39eg7AWmBC*a5l_h?YOmBSbs5
zaQ~n5f6@~RtX09N;m{|A^~4Pd&_1iLp08%<`}dauJ&_B6#Snf4{(*~n*il*L%CDw}
zF5Q(&EKDZI<;|S+NuM|dMjq_pLE8zi>W1voHk+@WgTS&JUf$>HqkJF6+FH@*A(hA}
z+<W!vo3Rm^_cU=>zpe$bi>fW>MHdKMS2vcZSDCN&uNNC?^U13hjv%Q@kDGtw@D__o
zNy?eW_!*5yTWKT+p*Rzo7PQ!-+#WP(YPvjVu$tCx$?CRDbsX+$Ieu#T`m2>ie&Xd3
z`wMT5pAzNh%06Kyw%Vl?BlN)Yg1Cs7Nrg_d_W%{OeOU0@yu4dop1A}a)5yrk;*}L&
zk4O<6puNfKlUlf11}M<9xX#A1*W8O~O1Jry!+uX!bOO@U;7OY$v+rbCL$X^`E9X+T
zM9KiPk&yoSN6pf-^8MPn3ucXrmyt)-`fv2s)i18ArnZ}TE2^Zd>7eik@bq01^Dryd
zM8@~yBKJIVBib0Q0Ocym%d_9O?S~Hdn_jxu#*;kcj~DcSImLyuGdhWc((KdhURY-?
zRZE~%)8ca5FFQZh4!e_T@r_K<l`0JD7t$1tTrHJ+ujK#o+|issHFV!W!Q0hW=g|R`
zPyODEEq%rfG9`YN=LtwpmU$uLd*)okzC1U-0nhTIJZQ$xT&CBjvoaf>PS)6yU3zvd
zfe@}L>36p76XfhzVUmTz?ZHY8s`6PD`^0x`gHa!l8Y~Sy^)>f;9TSnANeAD!Wh1zL
zWKq_of4DONxE+;UG?1;ZU;CKma8F4VO9KOQz$5MxYfVv8&8pa-Yr>_$y(0l1ZRWuT
z-F54Nqf*V%UKjB9G&o%zd3mvEFn#{^$4Of+&&KM*P~5{CI@2Sdl@q})KbwEJriD~4
z^mD#EtYb9iZck9?YJHdz$3bpZtL>#fG_gFcDq?zH`QrV@JSG)~k+(mdZd?k_tTCi;
zq)a%Qo?c0{5;skD@C46<n%jw|Uv%ncFVV*jgXTOgOnn_YXWwD`zBo_JKkbES*R$k#
z0H4yul7oXPik4sce09my*)Le1V@Y*%q1{w>`!+xr0?r+Mr=nGY99^Eh!C6)ABS@+1
zBYk7&82n!GL2VCO8PMB8@c`@W3kFvb%fHOa+_rx9=&4U<{cOR#w=V-RN3O7^wyg2M
z=*=9M0InryweHl1H#8Q*@QC#AhsqJ|t8$;zM#GC3Sp*p={EFUl9!l%Z=oIYl48TRj
zd`rl;B)AJI7e$jpH(~V+i{C{eU$O)%u?I_YEN3xIZ+t%KdY{qJdrEXT4(Wv~Rk=KY
z?4)O<<GEC2IQOO6a1NI_xD=jSu-Tp4QP`g0H=N13Y1FgMNXp>6rB1wlMSR7_G6RSi
z9Vyw|D=}ms#sEB$KtN+qcvAK^V`hkdFVB2!uE6q#K7&BK>>bshHuRY!d#QxX$|~9`
zWO~*QHDjD63RHR<j*PqqN*I0OnGLCpse;bsD|?B+s=RJoM3Ji*jh)H%ev4PAA!j%V
zV@xSuU7O_^pPA@7E*38h0M+w`T)&@Bac_yOxBXS=Qs{76EY0k(XcNz5bzUWQ_e12M
zc#pNB%KnS<W6_2i;)HiPFYZibp*>^n+_`(PbUiGLUhXI`o2n_E;F6PTwNiCb{Rdj;
zGDJR>i)+afg_$3koKEhie^bL5=OhI4CX`A+eh_kfg~IeuU*++R6SlE~fok8S?DPl9
zvh8PcGwG=iWY!Ud#!PvR52#O?w&b`?ZqO}=e&D(&e5whC&Uxq3=2r{DMrc%1m0jxg
z9Q5{b6~8LS=M=xd%yaBPqHnRX=IUW6)apubU6g)ZXuJ_E6wo3h60}jI_(oOaT1yC1
zQ!2MtSv`zAXM=26PTODIDJ(`9x-rCX>!tCSw_QgM)9PW$bszHspBiuERZIBWed`vA
z5W+EsQax=G8H^{BEIs{oz2v$I6Wf4lgfjotF6KL@Pkng{tIIt9yd_cSqn3Hm7&P4?
z&FBy(#YV)TH!`8^p)IO&w7PjYr5OjafoD`NL>W$DYg@S#^F5v!OzPEvzw%t#1dGVE
zT{~T^xp;;*s*v))o5ty>&8ZJ=E?-XIyu;BpA9U`A2VExXb%NHycqMuJIf1mXZG*4f
zm=o4ONen{K+&-I}alncFz899m$;KF5d>PJCCc>ZkefL{52J*Esj&-S-<cD<TpE!^T
z_hc04eB;N?5011=FTAcC2V=R%CHAGHec$b+G*8jP5+it)Rws>2U%Pa@=?DAVh25Zj
zGW1eW3Pa&6r9)FstnUh6t<+a$88qNHHUR0Js5*v=<?|M%|2EGbw{eTw-Y0qEI){3v
z5ySMQhVQgxbt*sf@483P#2o(yjweSnNJpZz)09W7)Z78D8+g_!d?Sw^{~<f<qobu)
zZfSKQd^EvvP9WG#{lMa!gg%=-4Lys73I73ZH|F<>MzbT3wRH<yD3zhZ_z{b-p2TLi
z@^Y}Zp88^-QerHyFi{U`6(=$Cnex@)jx5C%PPTf~dUh2Ize%q1>FiPu<H!A!?b0y&
zC(|8*o<*^$jW)i9qN+AsjumzNNTWu<T#FLXVE@G;k>-#do|rZ+f#Gv_UgxJz+83f3
z`wtGWRehm7L)?Ce9a=#L;*sn>Vk8G2wP1z><??jD#$zFvj_%U-YSMNu@x5{B330m*
zasAtNCnfWEf`<cgYqJlS4Pg3Y*{X(&Y3Lw?8a8eHzFy3<rAm*&U0`T_WzX?5hLVOl
zY+B6`YYJClm3o;@M!ze_csf-kMivwNqh+{pF8K@(xLeK*yttyS)f^Un6sA$7N%xxd
zk~P=nOcKoZo_Epur^M`&FRW8=<{vX5J|>zPnbZvd=H0Pe)hsU(`L#cmse+^J1G*~&
z0%R_gwA&Y%C|M=Hk&*lG^a-DvU4CJMt3&(Dk2uC>jwv*R1p5btchP-{AMxRH0`x{a
zmUrT{+IhlGX@t3Pd1srHGS$@$<2;^~jq%ZscG<BfnrR+7Wo|}m1EFQ2<@c1}Y>0QV
zEd(3dim8M<8Vl$BxE0*{mN$UP5J(vf@R|8skvyK`*zvN!OZHUxO0WNG_n`V|MX3*%
zJoDp=#vg^zeV?>u1V_plU^SVm^CPP#3)Gd}fZ;P*0hB-#wn(F720R{-Cz}&C&<0g+
z&v<kVleQl<OnsXD08n0;zA+e{E#KOxPHxV<kB|3C$^r*)DeHui3KQkYQb>fB!sa7#
z7mj=rH(^xAy*w09npInpj#9z&*i9gt-lOa>Ts!zx%TzTTW7F3}sb?;;pdSyF#*CS$
zuZD`#viRQax9EN`8Cdz<40~mClUHcpTZkM-l(DN~c!}Qfxu|Gg8ak(&)WS$=L!*FW
z9bai2WZr!F@wQdu9P8M61l?pMP{{m}wSN=2iSsIf;<7<<w)@S{vDD}Dl){e<z544>
z&=jlM;|6hGsXa9&G8L5itJ7a|6_0`sr;Q0JyDZ~`Q<LV)f(k`hW70!qQo@(}A(H1=
zEuSY~hmW%evF7ws-n*{G8me~6#DMH@#TQAf9Fv^3lmngJh~!j(cCV(kHs+a|=bO73
zSnjED-iUtAU}?dw1Go?KURnzz<JQVa5N7myvLU*(-qIo+F;i~<3YuN+YHEV2XtiV~
z4YgO9$}RWlvL+wsD(Q9^ZCl!(0_l8QapMP1ll>j%OBBTr0rGNx8xw8FfT;9H$7)R+
zbK5LW<E_vP%>usiFeOwQ%Y=ntNa#0yrifJ_>Xx3IEk_?+|4M<%k)k`1A&hFXZ@CNa
zT1UNnR2txYZOT=awL(}=4x3QIHi@~GFSudCl+<a>{k|$pI{brfLD#p=9>)(CWFuik
z{!jLSg`qJAX&+whQnF(A_%`ErPHEuT(rM(&!>9f9JyU2k9l;FGTU}m$`<iO_HCgg$
zTTn^8NtaywyWFdDY2Oo@7SAcUC6vWC-z|S7)oVi|I~QScOrvlaRdYU?nGCB^ZzX_|
zzI~nBHa*F~<fO8Xt#hBK-Nwq1Ao8?crtg|f6=P>!m|eZ80EO#;dJOW76Pe<#X+&MC
z-9+shy`h0o10JIg6xsJuBn|ug=AhYln=xzV3;*XLs#R<fy?OcVTpR)_;DK+47>3s^
z7c5-N)P|-mf78zWk^6eJ*itsM;pp*P#zt;!%7P&*xzJfXZ+1rQz)GrYkIt8jb1`hx
z11;H|a&l)hEpjnmFMcXi*>^bRZ2|g`4{wk%VpQrtPM`5P*GJ_YFZwuj!9ib^_hq$?
zKw14K7S<Xir1fpMLm09SX)mk0L8QOTP5@1Hswp+{*XXiO%9vLH^laHR<y}|({e)fz
z4~}fIDJkJ!olRFU4He7y$TEHh++G>hbd^&*B<6dD?)7?I`@k=!yn+sfD7^sTGOJUR
zffNK9PR`6`%L3SGv2|1Sa+RXGm#p`1H$@Y~#&VBH`MvKb4a|ko^HW3hystPa3yUQU
zIcsW#4-WSCn+L7e+GI|Lw1rhhWK=%R^%q-`f3{@j*#Eg;$s&<juN=c(QV1=Ok-cjP
zbx~K*J;puZo?2Xx6#1e>sQi$SJ$CaP43XGI9Xg}*z$sMtU0`SH5R?DJgHBFKna;~c
z5>HC@nw~ECr0O%dGz{bmO{Fm87QMCMTF8(g-^tHN&)9!$-0XSHwZe-q=);VA{7Sc!
zq1Rxd*%gl%r9rU1t~|yEjmpNe0i#_!5t~wWYjUJl!*wbY*C-_b@S0eY`u9xG5spFI
zAn^n-XKR}`v2QSb;>`YYiUG)$9#s$}rL$0Tsa<M~OyEUI&_W{1U{yry1;fm|&#9fG
z`5IC~5yeR}aPFqC$2qko(ukACo3(78Oo}L!i<RE%Q+VI=ROD#lk$F46d#S=@ZWv)O
z>{b9!j=`a|C|$@saZ%Brg0Hqq-bst0p*VtumM2^dy*zgHFt}xe*2dJ3Fm*S|<14RH
z*nmRr-N$tBr$&`QPkEz#RKsb{HTx6%xXbi)Ur60rGz8bC&5Sf!R(ANM6SIAhh}NFL
zV1}%Z79T1f&6@iSU5&H#oJDxLbi9}!;=gH`EIDiNx%`yJ=J<wK>n7(&L9wg4gKA3#
z>U-39@xsybngt~H6@RsTOGyR*_z~Z(zgIKAl-VO<Nhvn**s9th32;&pt*Ao$;&nwj
z<B4@!B|Zr;aXwxo27$FEX4y)7=S?(#@_Z`Fbc$FT8#Dr`uwBXp5^KJ2UZN-lukiDh
z5}%7RMq#VytAK$!BJm0Ygu?FXuflF_g)K!tDeewvjJ*<{ErEbj0D;`x+{E0Z#IVjN
zkOUYE25sHt`=!bn@ed!)#o1wt%NhY9)`m&-WjshiOaipcOS164YZQrhw@Fwm_8&#0
zIsBaBmO4psacR)c>upUaXKQPRfB1JzfXFKaLV>K|1US$YjdbJZ<Nf8wT@IJzQ{MJ2
z$qZ%1Eo0CaGyx5F03wJLb7Iw7`HU?RW8=bW3df)nL3?uil-uGbK1v`_DN<QdTtXHo
zE&-I3H<6T7kdjo8loXSY{RQwltD-eRVJ{ge1#vksDX=1F`#y<rPpfv(f7*hyCe8ta
z0$Tqn<SRo=WF(A#Y5$L#zsn#z5X8(+S>g}wU*w;>|13sQu^Nphpb@-Vo*}@k9FRZ+
z)&)aQ;*;Y0NhCH9h~E53llVo$TG=5Hgl(!gALy6!QAo_sW`nW;@!=b6Eq#!*n7EjP
zfGHXyCCR%(Cann=v8Yc10!fQYILZj9xu6}afnYfa>0ie0z$yM(ISN0Y^!C+16WJpF
zJ?^m+Q)0cdCZ~bg`Wm2r0w8f0J-k_qspDk`obw6|o?`3qOoVjs2+q)!QCrq)N7wL*
zpI=+yk9%u#`4sRxw&(HYT*>JNvec;iO7+%?veN-7MRbt0ca}qsB4@6wH_J}FXg(=@
zL?C7~P1q};1zzW@K6zeJJKZ@-06j@=IpyMdU7))A73(Co9)oDFry<lmr4SOo;a*{K
z$~HW2^3YwlFok__`di>i&q6Gv^6^DBz8u*Zzy!xa%<SVQE|6mT1WT)1aZ|!ro>Lv;
z?vE30si7G;bS5&|zzwzV-Z$sowH6vToDM+MEwg=7F_0Q=$mLP5*(m$&8b5GC7!iAS
zzA5Ol2b(W>2wcka7*eMU>*75n$SaIhQL;#g;Fyh0v2bxH&rb4}Xw>aG^Y9d}%wbpv
z3|9M{o)jUqwM4T7i4pQF1t~(3?PCM{y|YF1_O7wr=h~LqU82}-KaqM@JHv!{`e)1O
zFP*Gy(7$!Ac0lj%0s0?(tl#IgBmA?AvU5GL2iQ(u>+dL}&eqQ0A^P*rdH&w#`a2S-
ztFtpM?t!$^_xd{use`pMBKAPR?dpU5cX1l_z}ejw`#a9>v)b4LXQxm0ca+~}B~4sp
ze|mX$-|TmsKRahT-f<w}{H>3+y$rY4_U^(QsRX|>&{V$_<o^`h{TH89x7!(x#Q*iX
zt9XA(?#=*7C4ik_&a}HA@Tc7F)aQ4NZp0n_Pp~9ri`$ahd+pEEXh-8yV!$PZWKyEE
lEw$~M-GPJ@mv;u1^Dj{vqDx8i8US#B_`wn5PtVo^{x1>46-WR8

literal 0
HcmV?d00001

diff --git a/test/test_read.py b/test/test_read.py
index dcb23ee..bfaee5b 100644
--- a/test/test_read.py
+++ b/test/test_read.py
@@ -424,6 +424,13 @@ def test_ods_read_all_supported_formats_casted(self):
         # Text
         assert_equal(rows[1][9], "abc")
 
+    def test_ods_read_multi_line_cell(self):
+        fh = horror_fobj('multilineods.ods')
+        table_set = ODSTableSet(fh)
+        row_set = table_set.tables[0]
+        rows = row_set_to_rows(row_set)
+        assert_equal(rows[0][0], '1\n2\n3\n4')
+
 
 def row_set_to_rows(row_set):
     rows = []

From 8eb8691b5802896c06b531b07ca2f4073aad71d1 Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Thu, 2 Feb 2017 09:48:06 +0000
Subject: [PATCH 07/18] code refactoring: restore one liner where possible

---
 messytables/ods.py | 18 ++++++------------
 1 file changed, 6 insertions(+), 12 deletions(-)

diff --git a/messytables/ods.py b/messytables/ods.py
index 6f6708b..3a45109 100644
--- a/messytables/ods.py
+++ b/messytables/ods.py
@@ -181,23 +181,17 @@ def raw(self, sample=False):
 
 
 def _read_cell(element):
-    cell_type = element.attrib.get(
-        _tag(NS_OPENDOCUMENT_OFFICE, VALUE_TYPE))
+    cell_type = element.attrib.get(_tag(NS_OPENDOCUMENT_OFFICE, VALUE_TYPE))
     value_token = ODS_VALUE_TOKEN.get(cell_type, 'value')
     if cell_type == 'string':
         cell = _read_text_cell(element)
     elif cell_type == 'currency':
-        value = element.attrib.get(
-            _tag(NS_OPENDOCUMENT_OFFICE, value_token))
-        currency = element.attrib.get(
-            _tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
-        cell = Cell(value + ' ' + currency,
-                    type=CurrencyType())
+        value = element.attrib.get(_tag(NS_OPENDOCUMENT_OFFICE, value_token))
+        currency = element.attrib.get(_tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
+        cell = Cell(value + ' ' + currency, type=CurrencyType())
     elif cell_type is not None:
-        value = element.attrib.get(
-            _tag(NS_OPENDOCUMENT_OFFICE, value_token))
-        cell = Cell(value,
-                    type=ODS_TYPES.get(cell_type, StringType()))
+        value = element.attrib.get(_tag(NS_OPENDOCUMENT_OFFICE, value_token))
+        cell = Cell(value, type=ODS_TYPES.get(cell_type, StringType()))
     else:
         cell = Cell('', type=StringType())
 

From 15a5bfbe11bda4c4c2e64354bd6905ba1811218c Mon Sep 17 00:00:00 2001
From: Rufus Pollock <rufus.pollock@okfn.org>
Date: Tue, 7 Feb 2017 09:38:33 +0800
Subject: [PATCH 08/18] [types,bug][s]: fixes #163 type_guess error "has no
 len" on DateUtilType.

This is happening when type guessing using the `DataUtilType` (in my case on an xls). It is caused by this code https://github.com/okfn/messytables/blob/master/messytables/types.py#L188

```
    def test(self, value):
        if len(value) == 1:
             return False
        return CellType.test(self, value)
```

The value passed is the value of any cell and may already have been cast e.g. it could be float or a datetime -- neither of which have a len attribute. I suspect in CSV this issue does not show up because everything is a string. However, with xls parsing this can happen as xls already casts cell data.

The solution was based on adapting the test from the DateType above.

I have **not** created a new test case. However, these two tests in dataconverters were exercising the problem and failing. They are now passing:

https://github.com/okfn/dataconverters/blob/85cf1a844a4e6b0b7de349a9c61c7c1c322d34e1/tests/test_xls_json.py#L24
https://github.com/okfn/dataconverters/blob/85cf1a844a4e6b0b7de349a9c61c7c1c322d34e1/tests/test_xls_json.py#L33
---
 messytables/types.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/messytables/types.py b/messytables/types.py
index ba017f3..356811c 100644
--- a/messytables/types.py
+++ b/messytables/types.py
@@ -185,7 +185,10 @@ class DateUtilType(CellType):
     result_type = datetime.datetime
 
     def test(self, value):
-        if len(value) == 1:
+        if not(
+            isinstance(value, datetime.datetime) or
+            (isinstance(value, string_types) and is_date(value))
+            ):
              return False
         return CellType.test(self, value)
 

From b50e6ef201fcd8fe87aa6fc3ecfa0c039826eca3 Mon Sep 17 00:00:00 2001
From: Rufus Pollock <rufus.pollock@okfn.org>
Date: Tue, 7 Feb 2017 09:59:28 +0800
Subject: [PATCH 09/18] [README][s]: fix broken badges at top by replacing with
 shields.io or removing.

---
 README.md | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/README.md b/README.md
index 75667cd..52fc19c 100644
--- a/README.md
+++ b/README.md
@@ -2,11 +2,11 @@
 
 [![Build Status](https://travis-ci.org/okfn/messytables.png?branch=master)](https://travis-ci.org/okfn/messytables)
 [![Coverage Status](https://coveralls.io/repos/okfn/messytables/badge.png?branch=master)](https://coveralls.io/r/okfn/messytables?branch=master)
-[![Latest Version](https://pypip.in/version/messytables/badge.svg)](https://pypi.python.org/pypi/messytables/)
-[![Downloads](https://pypip.in/download/messytables/badge.svg)](https://pypi.python.org/pypi/messytables/)
-[![Supported Python versions](https://pypip.in/py_versions/messytables/badge.svg)](https://pypi.python.org/pypi/ckanserviceprovider/)
-[![Development Status](https://pypip.in/status/messytables/badge.svg)](https://pypi.python.org/pypi/messytables/)
-[![License](https://pypip.in/license/messytables/badge.svg)](https://pypi.python.org/pypi/messytables/)
+[![Latest Version](https://img.shields.io/pypi/v/messytables.svg)](https://pypi.python.org/pypi/messytables/)
+
+<-- download counts not working on pypi atm https://github.com/badges/shields/issues/716
+[![Downloads](https://img.shields.io/pypi/dm/messytables.svg)](https://pypi.python.org/pypi/messytables/)
+-->
 
 A library for dealing with messy tabular data in several formats, guessing types and detecting headers.
 

From b7faca40d2eb888cfe88a0505768a379c1b51152 Mon Sep 17 00:00:00 2001
From: Rufus Pollock <rufus.pollock@okfn.org>
Date: Tue, 7 Feb 2017 10:00:35 +0800
Subject: [PATCH 10/18] [README][xs]: remove download counts badge in html
 comment as html comment not working in markdown on github.

---
 README.md | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/README.md b/README.md
index 52fc19c..5d83641 100644
--- a/README.md
+++ b/README.md
@@ -4,10 +4,6 @@
 [![Coverage Status](https://coveralls.io/repos/okfn/messytables/badge.png?branch=master)](https://coveralls.io/r/okfn/messytables?branch=master)
 [![Latest Version](https://img.shields.io/pypi/v/messytables.svg)](https://pypi.python.org/pypi/messytables/)
 
-<-- download counts not working on pypi atm https://github.com/badges/shields/issues/716
-[![Downloads](https://img.shields.io/pypi/dm/messytables.svg)](https://pypi.python.org/pypi/messytables/)
--->
-
 A library for dealing with messy tabular data in several formats, guessing types and detecting headers.
 
 See the documentation at: https://messytables.readthedocs.io

From 9cb960a03ba045ded4d6469d065a20530d20484e Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Tue, 7 Feb 2017 09:26:21 +0000
Subject: [PATCH 11/18] rebase master branch

---
 CHANGELOG.md           |   3 ++
 horror/ods_formats.ods | Bin 0 -> 9960 bytes
 messytables/ods.py     | 111 +++++++++++++++++++++++++++++++++--------
 messytables/types.py   |  43 +++++++++++++++-
 test/test_read.py      | 108 +++++++++++++++++++++++++++++++++++++--
 5 files changed, 237 insertions(+), 28 deletions(-)
 create mode 100644 horror/ods_formats.ods

diff --git a/CHANGELOG.md b/CHANGELOG.md
index a3b1d91..62bead6 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,6 @@
+0.15.2 (unreleased)
+* #: detect ods types: boolean, currency, time and percentage. support repeated columns
+
 0.15.1 (29 September 2016)
 * #158: Add CDFV2-unknown to MIMELOOKUP
 * #157: Fix for Python Magic API change
diff --git a/horror/ods_formats.ods b/horror/ods_formats.ods
new file mode 100644
index 0000000000000000000000000000000000000000..fa1f7ff9fceef1586e0b071d503a87f919be99cb
GIT binary patch
literal 9960
zcmdU#bzEFY)2K;s34|aC76=~PT@u`bJ3$Bc8Qhsb2(H21J-AzNw+!y?1b6sg-!Gf&
zzMH-G{&{PD$L35`PgkFQ>U3AjN<zb6LO~%wK}i}>iTRjw`qM!{K|O5uUqM-zSO7py
z)&N~=YjYEQU66^DCB40+0iBgD&;&?lWeu=2u+q1)09b<PfY!DET?3#o005Hxi%kq8
zu|p6X6x74^;@+mBv7LpUrLKuNkRJ5ol+N1H=##9B=o2Jdr2C&d5f>AZzdttKZ+8gL
z_j|{p1jGA-zO0m@FbWC^78VvB9^R{0uc)c1nVFe+cz6T^1SBLR6ciLRG&J<}_07%A
zfk2>(iwhVG4hRT{h=@o?NJvjlFDNJ|D=Vw3t7~g(>*?tk9UYyXo?c#F-rCwaK0dy^
zy?rS0?yk5SoCF00+b1scR?%^4cVg0Zh4Lxv(93DPVQ*!&M;CX=oS*82K;?OoXzpJO
zErlnPaWc5lh?s}zYfrYH(;ChMx86wH8Jt&WHoCGIN+erpI*;)lwzrN$MzI=?iwMn(
z+}agf`7f+vlOweu>+!?9+m<f-5!bHShgf`f*ol~H?K1Mw*+x>IS90dz7|F@TC=wk2
z&m*E@8lJj?0u<{oE!@MFj=`04oZ)f^X2Q%jqgtzV6*d-@=0@}f{e~aYI&3fo&-+<6
z0OW>77Ml83-ThW-@-AFTg}60V`vHNK`${z(Xc8j3AZ+)bagq6&lEa(pz(MGowNDvy
z4!pG;R+4p10O{G+9{J-D;6YQ>UExMdQx$phpm=)em!hV}mQ-A37v|^`N#t4OdcpFS
z!^Hv@BCIZwHy0Z!d}bdsaA)JFBi7fvFs_rk=$i;5=WowM2a9FMJP2+81U3rp`^PrU
z0~ed6SG&Nukb(2e>6~)bPR!L$U{&el@NspXpoogs0dd5mcn)$m=nUP+oO~ZBk(wUo
zGO^Yx;t*$!4Tk$!aIz5OJK^{l9{KT`Nk$Z1%fT|`9R^87PAj-u;0np$J(2#z{b{48
z`C4;D(0@RI-{x9T&~g1JYI%Otogwa8_QG<5lOGL@0`22z2+E*J_H7re#geL9QfBoh
zULOw?s~20>xuU0w8%L0nqS+VR+@a6#>oUzhc*ye~>}H-)23<O{qg@=GcC+6`%<pQ9
zHY6x>v%+rb_Z+>cU5ttBt6F1he`2?iC|5{Hn7*G3(Tr+x(7w)2-{%u;IyddRnv3kb
z#;4+9*^blWEw!C}sqp?u?0CV}di|gqUmRUfBmLlc*!o+sHg<GJS*K@5LFZKb#eSv{
z={;4Lp8>ho%Nz7&Hq$2aTVGaWn+!FI>v<7*_ReFOR+sBtI&KKm&M1yTw0bu}LI=FI
zC{bM|fmGbHx7G@q23q2U1rWfKXyGFn;iP^yvdpyW%F#-Ks7tBbh|7tsiutdDd40X4
zGbJ7@V=VR0c#&Jb+jpQ5*d4toJd1zD8<ZhHGmAGFIyf?R(Bu@8L$^aWB$hBgOBteg
zP2h?uITwk0mY8Zm#kyCF?4dbq<Z2mLvzI*9y!n+<(l7u-I)giZ>xek3j#{?0;9~KZ
zphmV_-2*LI0n9EswZvb4m+N}lT0PpW5PvKW6kY{3$N_~3Rwfbwh(&Zgx5V1e^K2Gm
zdFi30OsmJduFsO-{VrV4-+%0RS9hoxlNjtC{$3qxCCU5rYvuclbYldn9W-(@ZSELA
zU`Zm_U!QR8^oktqxd;yf-AG?eV$&k+YE-2cMz3ugh?d=kIof=Cp&Iu%tf@LSAn&%o
zG-tY{wmm^`{>)^F`Mf;0k5C#RVc`z`++K!#jk}=-?+D(*yw$Qd1T6_>4$N<%l4O0S
z>Tdes?E=i5M>9QzsKbKM>sqZI6*>Z-LRut!SapiZJ{p+^j|W-?Mkb%y*8J9$uEX>O
z@@{QI`${g_%PZB>vybf-R3_U&eHQhw9{Yy}Q2TDwc059>bsOrCra2;uOv8%yAjKud
z%U%H|?K<gp9SFhNEsHdnNY9?5L^q!1`zN9$nme>>d?VQB&$FGWukiI$_qwCKDwXk!
z%+qG&zHiE0FV|$sg3UTQs7KFc3(sBpCB8ZuiNy_c@y(eXZU6{B+N|=BMKkuz_Mba#
zVIygkvy9Aivd#vCRHN~oFM$6)+}-VrH$|W9Yw7tvRk|lUl(?{rP?3O+_kSW40s_Jx
z(eL5T4?(#<0RliE6H6l?ouh^MuKGMMjSJJg$wT<#IJZ`qWkbMdEx3Sx%3>B9Q)B+I
zP&XT?)H9*0@p-7)s%;4~Ec?@e4KT5IBap-O!q#?S>4><E0@%fowj!VFAsdAu#<NAj
zRHrD?y0=NO3fwM6=1L?+V{_YTe?__!Dfr6ZxjXSkjJdve`ly&Na#;KW@<*>$q!(FD
zV7i@Ej{y6IhBW+T&3-&uR9;Wz%9JI;R|IWOy-L*zi?*F<&9od&C}Yzmi?R;q=%wW(
zJYGn^uM8M;uqfTA_TEubP*Z4s5K&Oy<|v?VvMXT<y-Of|)XKs}kw2Pi7VXb6nLrYA
z?H5;SF*B%--xd~-+CubVL`;p{A3%zSjxHhDkaL4{NVkB;fdwH738VM67>XM+P9Q{?
z|7eVDZl`r63UTn^fASXO?}zq$Xr2W&Vwe3j9?53==Q<9Q2x^a3NxB#V&k>0WS&!)K
z*Bidcf<(}VpVRr#N}v&h-QXX`F)l~W$_0N|xofS|+26S{>k1)b`tlCPIZFu2T=c~!
zjjh)H@9>cX&DLEohj{|?-`xpPn2a_g@2H#EdA!yb1goH8u|9}Sqq*){b>~(1R``>y
zzL~LwxCBjQ2@HJ^T0qQPlue85*}(kvh+|HuA0IM!it;F?MNwTF)@o<7gkUJ6b12KS
zE>aJ{Z{!(`?{rvIf~#AJqf%;xEvEAzB|g?=&XA}NAJws3Y5>;d&Yu?{NH=wuwK<&m
zfjeAfrE%hK<;dxS@e6ofroX83#8PhA3cAcDX-CaaHBIR!a;(*7D9pkdwm@Q_Uznyd
zBo7_#^SE{A+xO;ANu%1hAz#$!dYvcE2*4YP$x_HEVjW=Ft(W`iR!BtU)||O*BSsT<
z{ObFw{0=d4Qh5`qH%&|PdS2*Wr%a&NYiu)`Hj<3ta*g!kz=Wbrc>-=ytN{T&0*2nA
zYUwvD8Q2StElcxIpgF(PXI&!-0I+Gg$ZGt1SMzGKXDq&2^`D{nROye40^1~uPVXK!
z&R)3XMzrD~x9yT0JQhK8rZY6z%<bGd7(PYC_n#$vx<;V~H-dMaMB#H-3#&SzhNR!J
zdD{0TI@wL|8gL>8L50c}g0iIiPMg~ixd4Fu4%cC;6t;*y{*Cp~0TG=xqi7f!%nnm*
zQVYY^!)hd@TO=Gtj>ju><Vk)lpBs#D(6>L(9oI3%(rPx{y4Kt^`P)Dy1=YMSvS^eN
zdX%CUlw+g7oS;K?iAHn`h4;ct@&NDmZd6f-tOn2***ej4U>K&A!~*Br784NUsuNEr
z>Ch*;I<$8Ev|%LSdzE=Bl{8O~sw{>MeUV|Nwwy*d(_DdTt!HaTCmXRRZPb`U`kjOH
zNeLd8Btv6j(B*iTB+X8LApaq={LS9eDvR~qBK_8C)^m^XVaC8_84tG$z(=l9XRVWA
zHf(Z}UNe3b3N)44OwNOqMJ{K}%Ke6?o$iBbajA6E1Wq$bEvg~=38yn8V}~|@O3e6u
z4KF+9%4!)}ofmxESA3PB{El|cwkF8VE(>Ex*|kyJ&4lK)EEvxOaON9@_kC6J%5Qc~
z@hzBSK$A&Q&Jo$&t)$N1L@s-8-Tj*0`I_o2oi;#l`bJ$~xRn>qxU@wZ=wdm*2)^kZ
zH{a*JT*?qL^1HimCD$!>F-)m+G~}U)B<^jztw@kF8YXZ!X_2j{T`QQiNjNQBg&9rd
z+T874qx{;7{&5HSX^JabyzMd+Pq$HoFFKefF>s$w58LvKX;XcK8nv@V#R<KvBs_wS
z?d=r<EELo}6%^F(NsD|>T74@^(8D0&9<-|(7Qp37^xHY@R9%Xj5n0zTgUN#3KC(?n
zB4-tQ$!K&41ERRGm^ZZN=M$I9cZa>DP~r#~V*bHGvn&mZVxl_QNm?}6bwjhe>{jAD
zPVIphadT>O-zvyCo|p)YoFtwsABkQ~U9R0aRDmjpU!fA?t|7|!5R51`pTSfsyu>VX
zP#zfZB$jauO@ldl^WjmywYm)~xm^kfi$rs0sb$B1Vg>tL8U^sVob~-NlV%w)TMdbC
zrC=->EeFIvdpNbvil<Tqvp&f{EELYaDkg(}Rt@h}X|qk`^e%`yUcS=4K0Y2#wBNKj
zHLKSan>vpYWF45vwCAWWm}`tdGbHO8KK3DUtsB`8tObVQQiTNSqLS(qq4eo`OC_nh
zHWa~&j&Ikt_AkPfwAf_%ZKpB?_<ro6VRFJCWh2;j+2#3~ap`zn9zy*Y1wv+GR)mDp
z4HvH_8-E=s|3rxiGhVCZM9X0hb06-cGTOD*uepy1&81rq^W)SqiwO{S$O%<;;#|qM
zeoklvy&l}+;BO+`mo77W=X`yPX=&|!+P1<`A}voRTYH8Q5in%3`Bj%AUCOw#z?VW>
zl~!AS+K+TkhQPLxSgVZElnj=G8GZ!8T<<DsVRzrGrh@;aCBCr1+UR?@K|9i*x$rMD
zS1W^Km#*r=JVE{}vK-J&fEaX6AH8<$Z=^Bk_@mmdN_j?65G#b<S!?dPTgd0Xkl6ho
z$na)pl!l>tdmT=4&i7M<mrS~nf}Z~9$Am$l=_$Uug)UBwm+{sNk#T{xEmeE@Qz>1r
zGsDo-Ip{lBNF@Y|3~Ogq!olOM5ldQ2jb(fc-b7<WH_9-GpB%1iB9sOu%o4m)Ls8E^
zTfa~5@<Vu~+eo|+HN1M9e=D?5AWzNhB=Fkm(E)s<+-)|T79$5;`Tl{&5dWdODqR0l
zF>X3*+wi)>P$99+`0JQwIaoGTy{IPi-UsNJ+hf)v80n@IuhJ8W+E(T_9!1FPvuw~U
z#Tu2T9-4D`W>CmL_COYQw)m@$F%h3lNdyldZ<YD7W+W*qrzv*AVj9~dWhl3spibdr
zaQ4BpxR$21Ah&K%T)_%<KzqsCk(H3nhl^pWH1ap7lYjbZ)x6odru}q!>5_Ls^*NhJ
z)e{{Cy)>$<l?Awn6S#p77a&IH*&9b-utgkVc*+D@ENt(Ai_YKx;&jtfbFZoNa$KpD
zs+PWiA>3&tW$jcR>?iRkF8(6S#5@7RMQsWa5wMKqVfK%6j|co?Y2Tx!V(+RzEeWV0
zcrFP}cpULaDqw3z96mYCKSVS}WBk_kR9aX(BJ`VOEY0LfK>_oaj)l;eP;&uAOVm-R
zU-jMA7^C)y21fpxcU(T!Z6Cz{49BNbtTNH3wMR{Dhe#I~$Lu7E-ijCs#_b*YV&<8^
z_C^^_Kx7()WLCE=p~23D<a9<hqhgYk{WAbM#t9(mB>rp^Gz&5v@`}!=X{JC};K(rf
zM%K*vWu@8%*g*3|#DsEIKp|d0xW*qqT`j=a+4Kk=Q`>^B+V*ViAoNWi1_$;ox<d*l
z^6n$)RoM5ey!oWtI1sA(G}RJp;1?eO{M}eiwQP|%^AaadlFz;*$|yd0U-lT=3UMBL
zM74A5GO>6!A~fP*b`CYl;_63Mwlvg~#2NUO<Lc9vt1A?nO%(a{Tq!c=uvooA_pU=C
z#os1o7nCZDbv40%L*`>T+wT;snaEe6RWbO<bp3tPRPg13Vufg6Vu-;`z4opie#O(y
z2GimWxos!+rnyyB=PStjko40nF{#<WbbBYE!u8-4z&NW$Q!93Hf@A8irH5|hTfobM
zCmq%c*kr|*G`;fz#}Ns>G%A6y5=ygLC`Kr#0d5C6T~yq~*!B?3&p!6LNrwXcbg|lr
zWj#lOG*&jsP7to-=gW}TRoXjv^G3e1`grEvqnWHX#-17(-AVPSFFtGr=wPl-VAr8g
z)_6)+SCT`5oO|&R!kH&;&ScN#xmJ4%M=?FzeFkMywG|hugn1q15()MF+k{!&qnSCV
zzs?FrxsS2KK3BRLKQ3)_8);YsfOB@*&H-l>t|wFlPu}KnS?z_7GhdF=+Nq=)rdHTD
z)a*KaFP~0*QGu1j9yNGvFy4&}8zeeQ7wO2y?=lBo=XbtU;3jG}fkNUTOp16EwOiPs
zSW!n-u+&)i<XwBttDB?NRy2xMwQeaRh+#@y^qiHb+z_7S#nh1u9Qx<(cgs0&dj(Cd
zR~JxP+)T7W;H!(KY{6;mEH>TqT?Q1mCv6V1#ad_iHrQrt#;JxR!v55ED9BjWEy;Hq
zwE<zFQa9}k_!xtnyv6ktWa-hj;Pn82!FU;jnUth%PLWsN&|uundNvhj(hl0OW3Wq@
znumtW<NEvNX5B}JSEg(vM&e<W;|;WEM;T8(jU+y9z0y=eiHvR%#%Q<jWk#{Vm{rS5
z=F>K=y2#B0L_CidROt9pZ}+}f(*vitPorU^4C`hUT;;0S2X7(lf#cY#b1+cg+@^2@
z=DTqt>ubJTEr>B%F5dfOG(6u9S2iyq<e(4?uAsx;km^q2ZhEh&Rh8H+%7AqyYE<yZ
zJ(CZ((nVH!=R+69P4}X%e!1%7xOmp!Et*ucz_)kbP=FW33N;5VLc7s$qAF5{ReoEA
z7yKG~9*-Vg1T-DoyN2MQpm3T0?M2{W<_`#RG6(!I;96Cmwwz&m>b9>!V?ZS4WlB~^
ze?oPNYca|O8%`t^@z^bcOrqEK8Ep7e#nFWJrdd;3@(BHA=(C#IA}sB2B57|@;z5*;
zM!u~$-A2z<a1t#To_|<fmti<d#2FyLj^N>Sm}|V~=Cz|!)=n%TeT9yyj~#_GkV(D5
z+O7xpwFV)-8<@S~)D=oZ+$ZNNOqm{v#lbG1i$SoJnTK9zo>vXWo=MgiuHb`JNftnt
zU0l#2y&03ftElS(D-*wj7*FL&mJ`pwuBYZjx~Xa-EPlH;qb8~LA*pKCa(%j$b2;}p
zO|;~@L3J=m0JVe+D^)O$REfHjV++rB6t>o9(iQJ?<H=V6_Ma{cOhbiui?Q*MJ)7;7
zG-VXNMVhpalPBV9#e+Vq83&!OaEvb4=iqmY3a&C(_85v4iC(*GHF%2Oj0ddORWD6`
ztH7)jR%#Pxbc)BSeLeW1EQ`u(>x&uid85aQ(%_S0dwSR2Pi^&&LQcPu2$A%1^!kcH
zYc4$c+Kz{H2-gqE5@jWH!G9hoX(q<RJNbwk*+p@jDScT~sUO$4XW8*7(i*<AwfW*u
zHJ*1Be4%D#@?B>3+y3hc6{7FFJ|&`18?$5@NmJWIx`3A|okhgZ+E_6dtUfu|$QrnD
z=z^29ienC9f`Zf~&!qZI$~15lTVvu8qgS396d>S@6h_durQ+MR$B!q)0U?Etoyy*4
z+r-i(MAdL&z*QM~6Yq3zr0Lly19-T3@{^*%)(0FCi2Pv0MBqZvog471{PS?w+c|fT
zP(MyQd!_E@-GwmH3FlRC*ihxMxemirwdQt8-P7)g9wOV)7J*uX>#91Mw8TGTwd!7m
z8}pi63;9zS_pp)(sfcsK<aEND#SfjYxApn`gXm?GP<<m39fu&|ZLvC^5^jQ%beVc&
z;Lqe@oWNr=g3#oXpX!SCf(Y)M>+{lsq8Mdl#kaSnjEkg4VyA-`GK0OI`=v*w>+NEP
zHZnxRNF16X4W;;CSxS8ns>vGww_kJg+DGOTK%Rf|tt^%ODuCoApN83CT?9|uH4qL~
z;`i7G!D>}w%JJE=u#9am51pAYD^Rmx@YnC;tE}R*t7uWECEB}7*Q)P+qo=O`!l`N}
zDz%$wwN`4%4Ihp;aHXi-M22b3lfCY0Cnb}I+ICKTl&CDKXp2faeljHQ<#E}=UN^yj
zgs5^t9UeV_BWs<g#bjDw!riZf7p!q}6kSn}u5SQ-T1P7Qjh0e~$s83ae(O+)n=odk
zHy~D;lNx!}a?65s!<3mhiouAc*=|N2&aL}x`1yCqLZxlRi6}Fbu@G&VCM;eYs-1%_
zeR)=OJR!;J^KH>wS6fSHG%VP{NqY@1CsA$K+y0gtaRgJ5mYvUUQ%kkjDU)tD3T|(X
z_78HZ@#4>3Ufj_i?k}rugf?!MsJaet&d;yg-9@Zi>nysieY*A%tBdy2lJZwY<UL}z
z5M>wLj*HR+1>1A%kffwUhL&!1UP2fEPE#zlM$(d-eqOs9o>WgM2Ugt<9KNkt8GRg{
zXWueWvf*NnK4-_=%#zt-S8M3V6qzfUWE#<z4vDQQ9CU>6&4FY*UjS)9--rT??ew2~
z40Ks%#lA{pF58X1!Ca~&s9Pr<V;j?zCQTz{x@4*D)9-uTUDAReBF^)8GF-I*S0vMb
z#P4!yJYegNd(4-Xd{aG%JG6ZXjm*e~+J35Si#=vG%{CfxDaqd|=JvUkuv0bFYfeoL
zAy^!!@@7fC0Vq%uXd8?iX#ULWwU96g4%+F;n+uuE8Vuc&uQv4<RdNDz`HeRjp#!Qb
zGySow=`Dqnta~v6uZ<uoIL~D&nLIYnf{wY0g|<YoW^fQhEw8l1Q=Q%Xi_~6E@5lSo
zSi`+>OaImhJK-5$qN>F>6BW&!T%hYvr`*8$nE`HZn?p)#c6$nOejo%)Gdj@y^qsEn
z5=_U<DN*rI;;<xn<|RmXe#~(h4)WR6m4uA?61%eS!1-m&LQQ-0GAFA|w_`ygm~Kt9
z_c2?({KZc5coR#?8=I5N*<|=^5d`6V#oXvlau-6O#!HJD6QYz*DI0n*=d0_T^o=8f
z?KQtskupWzuvfZDep>Ch@KITAcDhF>D5O<A*O1uTRNu5)l>)}1q5S(45OXoNY6vF5
zGcI=(8J2Mt^r0NhGrC%Tyt+&3is;L?v06&4`OO96FScWK6f4B$&Ai8#56Dfm#|f>^
z80t0{%`C4KwX}Lt&`4cQtG|j9n_$TEdd!lfRmhu(5vB}m$-Y?}R!5>@LAJGHhKBD&
zlhNe2hQt%kCvWpNYHOr==35zdEbJIUY7z65;6E-Nm~15XMP~;rN3?G)%G086@!O9I
zE{@O5Qv`I!j>BxNc~oe!c|j(i+l$x}XzLdV>(p=HA3|v$Ns2JWeJHIY`?pZ~NA3;)
z(tQ|E2SrGRj51*QoL&2O^^F~nB(ju*2f{qlV|*rt*-|8tI33thLfsh~>G3Wp(}zy=
z^1SXW!6nh)-MR>T@q#))I#^Xq4oYN7B9=>3+kw#9JIS9tmiJ{8jGWpes}7H4b99TW
zsXC+fY7S6jqExAn0RUb2U48;1rb0+_RF5exs!Vy!DENInKcc{8Bia>Zc@a5{v;xy?
zkue$lOfJwljY|AY>&DPCY^Ksu{jYaVSWCyl3M!tbqbCc#vria#5<5t(RhKt%2KQpZ
z4+T+Ddr1ZMMgx;(S(`Q?U2k^pMS52jLl0%KL{KJpL_?+C;}LMj+bT1c*)}?6J8N_Z
z2Q5q1x`!<$D<+$J0QQ}5q%Y5V+?g~rvUYV0{T_yNZtPE;Le?)WBPb0ct*UU;N+wbB
zUVX^5fz;%x95^vtc7@Xibn>570S=?T4Pnz<AZs7v6t#H9i?}Mo!zX8Hg>{Dd%5@PY
zMQsjT?Tgoku(x1+;pe=9&Y*~7J6_bv2Kij*9c3@?SYPC6@MoAegtuf|9i}{zoW@B`
zN9a$3DO3rf2rx!VAE#0f@(-Nhp{hB9#%lWGt*J&Ihoj;V6qNgY0`&LOCl3?V7P^)u
zh5#Ul&ep)NKce3f%<%NXK93JzVB!ruEUYbMPmO7EGpnyEEd62Jv$@9L2xYm9yshMK
z%S9s0^TFluySn&Y0i`)DiY)rC$$G?$D>kXQ>gZ}Rr@5vkDya$M2o6lanaNbO4wh7$
z+Mj)B*-^{~%hD(v<BoXV)f2ZsIBdjFI?96)5MKrfn4GM*;&KiwAsVX4fwogZoz94O
z17^8$Q%aF^E9DzLRuYAIy$M2;>iGZy<JZbawljY|lH+Vi-CgR`X&42k@#uV~4j!DC
z^Fzs6eZSD~Wls}4?;2(7yr=JMH`;zMPZFt<z_j;ntB%|0ZvB>S@D69GxZGaI{LMhR
z(Pij*`5~c_Xz@c&DL*qFfxUMh_)qu!$a3F%>`!^fNDEq78k!i{+5VB=1~Sp#=d<nf
zbZzhFuzt;r|EY3+@=scZ=2p5Ozz?l|$|+F%Ls`1|`T%plebH97^!j$Tw*NZ|27TY+
z@DIHh>hZ(wUVi^)2=DLyy$U84x<&vXy?_bGLf0DjFEzpa2W4w3Ydh-)C;UI^0YSPT
zJK(>24*#FZv$Zm^1pxmKn*Zc;kd>ABznAw<^?T5=*0ltf|8M>NsjaZ?|EILD;#*pA
z84>zF``0~A3l)1UTZx|5uT#?$MWy!V3fS1@e4`iR_gty~!yxNLQ{G-p$sOh7DVD_H
z)p2t(e`z)aA~a43l0Pp&r$1Cv^kl0<y%KRXpWjv~QKMIGN!j*PO~!^yd<3E5$uVsi
zcZG>yc(P~^VCdjkAbJ*fES<=0;%nik`k$oLZrBY7@}90Qdhhw4*)i3deW(?-6^A!Y
z%ykkQLWl@=82gT(;9|Zy{H6Z%^=kkF?JksR4ySTu-7~&nj(7g=BxAoc>9)g;g5$Qz
z6cJ)NCqwZEkKmYFy&#H7!saB0<#(b-&(L1Jv|@*f?~HI?H?XYZwjA<vjw?LCE&;lA
ziz2C0>C-yCXpM=Y<-Fqs-<ix_P~|Kbyv3&2Wo;#Gly7SEK6(D|;3`n%CcnDxd%`de
z5ALJKm{7l8V{-54U*yM%lfP^IeHF=16~+70->oV6VdX~)|2iZ60M4Isd%rs6*L>gK
zHGdt#KFnDE6ytkCzsz6%UGdj|{V>A)DK_^$`8nSGyWX$#{{sYnD%RuO6aRmM``@*G
zh2CE)m`(84IPpXCcToJZIH<Lk|IL@5i2Q5g{?z&pB!6Dx-(%|Ul=_Lx|5*9aOb=Q=
zxc30fpCbPcR(^xff2-gFGJi_$KN$Tlfd0Oo4~Y3GYyV*ESD^mi+IdOz-~Rd)u)nwS
zfY_g+O!VJ&tp5Vs-&=Y>?oSCN`fp1=0sMO_fAvqtZ>;<P@_+37coKdE$peyq$}Q1v
Y04*yCd*7F#pwRC>*7y9;BYwF09}%!5r2qf`

literal 0
HcmV?d00001

diff --git a/messytables/ods.py b/messytables/ods.py
index 7b03d74..e9d56b4 100644
--- a/messytables/ods.py
+++ b/messytables/ods.py
@@ -6,17 +6,42 @@
 
 from messytables.core import RowSet, TableSet, Cell
 from messytables.types import (StringType, DecimalType,
-                               DateType)
+                               DateType, BoolType, CurrencyType,
+                               TimeType, PercentageType)
 
 
-ODS_NAMESPACES_TAG_MATCH = re.compile(b"(<office:document-content[^>]*>)", re.MULTILINE)
-ODS_TABLE_MATCH = re.compile(b".*?(<table:table.*?<\/.*?:table>).*?", re.MULTILINE)
+ODS_NAMESPACES_TAG_MATCH = re.compile(
+    b"(<office:document-content[^>]*>)", re.MULTILINE)
+ODS_TABLE_MATCH = re.compile(
+    b".*?(<table:table.*?<\/.*?:table>).*?", re.MULTILINE)
 ODS_TABLE_NAME = re.compile(b'.*?table:name=\"(.*?)\".*?')
-ODS_ROW_MATCH = re.compile(b".*?(<table:table-row.*?<\/.*?:table-row>).*?", re.MULTILINE)
+ODS_ROW_MATCH = re.compile(
+    b".*?(<table:table-row.*?<\/.*?:table-row>).*?", re.MULTILINE)
+
+NS_OPENDOCUMENT_PTTN = u"urn:oasis:names:tc:opendocument:xmlns:%s"
+NS_CAL_PTTN = u"urn:org:documentfoundation:names:experimental:calc:xmlns:%s"
+NS_OPENDOCUMENT_TABLE = NS_OPENDOCUMENT_PTTN % "table:1.0"
+NS_OPENDOCUMENT_OFFICE = NS_OPENDOCUMENT_PTTN % "office:1.0"
+
+TABLE_CELL = 'table-cell'
+VALUE_TYPE = 'value-type'
+COLUMN_REPEAT = 'number-columns-repeated'
+
+ODS_VALUE_TOKEN = {
+    "float": "value",
+    "date": "date-value",
+    "time": "time-value",
+    "boolean": "boolean-value",
+    "percentage": "value",
+    "currency": "value"
+}
 
 ODS_TYPES = {
     'float': DecimalType(),
-    'date': DateType(None),
+    'date': DateType('%Y-%m-%d'),
+    'boolean': BoolType(),
+    'percentage': PercentageType(),
+    'time': TimeType()
 }
 
 
@@ -102,13 +127,13 @@ def __init__(self, sheet, window=None, namespace_tags=None):
         else:
             namespaces = {
                 "dc": u"http://purl.org/dc/elements/1.1/",
-                "draw": u"urn:oasis:names:tc:opendocument:xmlns:drawing:1.0",
-                "number": u"urn:oasis:names:tc:opendocument:xmlns:datastyle:1.0",
-                "office": u"urn:oasis:names:tc:opendocument:xmlns:office:1.0",
-                "svg": u"urn:oasis:names:tc:opendocument:xmlns:svg-compatible:1.0",
-                "table": u"urn:oasis:names:tc:opendocument:xmlns:table:1.0",
-                "text": u"urn:oasis:names:tc:opendocument:xmlns:text:1.0",
-                "calcext": u"urn:org:documentfoundation:names:experimental:calc:xmlns:calcext:1.0",
+                "draw": NS_OPENDOCUMENT_PTTN % u"drawing:1.0",
+                "number": NS_OPENDOCUMENT_PTTN % u"datastyle:1.0",
+                "office": NS_OPENDOCUMENT_PTTN % u"office:1.0",
+                "svg": NS_OPENDOCUMENT_PTTN % u"svg-compatible:1.0",
+                "table": NS_OPENDOCUMENT_PTTN % u"table:1.0",
+                "text": NS_OPENDOCUMENT_PTTN % u"text:1.0",
+                "calcext": NS_CAL_PTTN % u"calcext:1.0",
             }
 
             ods_header = u"<wrapper {0}>"\
@@ -129,19 +154,61 @@ def raw(self, sample=False):
             block = self.namespace_tags[0] + row + self.namespace_tags[1]
             partial = io.BytesIO(block)
 
-            for action, elem in etree.iterparse(partial, ('end',)):
-                if elem.tag == '{urn:oasis:names:tc:opendocument:xmlns:table:1.0}table-cell':
-                    cell_type = elem.attrib.get('urn:oasis:names:tc:opendocument:xmlns:office:1.0:value-type')
-                    children = elem.getchildren()
-                    if children:
-                        c = Cell(children[0].text,
-                                 type=ODS_TYPES.get(cell_type, StringType()))
-                        row_data.append(c)
-
-            if not row_data:
+            for action, element in etree.iterparse(partial, ('end',)):
+                if element.tag != _tag(NS_OPENDOCUMENT_TABLE, TABLE_CELL):
+                    continue
+
+                cell_type = element.attrib.get(
+                    _tag(NS_OPENDOCUMENT_OFFICE, VALUE_TYPE))
+                value_token = ODS_VALUE_TOKEN.get(cell_type, 'value')
+                repeat = element.attrib.get(
+                    _tag(NS_OPENDOCUMENT_TABLE, COLUMN_REPEAT))
+                if cell_type == 'string':
+                    cell = _read_text_cell(element)
+                elif cell_type == 'currency':
+                    value = element.attrib.get(
+                        _tag(NS_OPENDOCUMENT_OFFICE, value_token))
+                    currency = element.attrib.get(
+                        _tag(NS_OPENDOCUMENT_OFFICE, 'currency'))
+                    cell = Cell(value + ' ' + currency,
+                                type=CurrencyType())
+                elif cell_type is not None:
+                    value = element.attrib.get(
+                        _tag(NS_OPENDOCUMENT_OFFICE, value_token))
+                    cell = Cell(value,
+                                type=ODS_TYPES.get(cell_type, StringType()))
+                else:
+                    cell = Cell('', type=StringType())
+                if repeat:
+                    number_of_repeat = int(repeat)
+                    row_data += [cell] * number_of_repeat
+                else:
+                    row_data.append(cell)
+
+            empty_cells = [c for c in row_data if c.value == '']
+            if len(empty_cells) == len(row_data):
                 # ignore blank lines
                 continue
 
             del partial
             yield row_data
         del rows
+
+
+def _read_text_cell(element):
+    children = element.getchildren()
+    text_content = []
+    for child in children:
+        if child.text:
+            text_content.append(child.text)
+        else:
+            text_content.append('')
+    if len(text_content) > 0:
+        cell_value = '\n'.join(text_content)
+    else:
+        cell_value = ''
+    return Cell(cell_value, type=StringType())
+
+
+def _tag(namespace, tag):
+    return '{%s}%s' % (namespace, tag)
diff --git a/messytables/types.py b/messytables/types.py
index 356811c..39ff098 100644
--- a/messytables/types.py
+++ b/messytables/types.py
@@ -103,6 +103,27 @@ def cast(self, value):
             return decimal.Decimal(value)
 
 
+class PercentageType(DecimalType):
+    """ Decimal number, ``decimal.Decimal`` or float numbers. """
+    guessing_weight = 0
+
+    def cast(self, value):
+        result = DecimalType.cast(self, value)
+        if result:
+            result = result/decimal.Decimal('100')
+        return result
+
+
+class CurrencyType(DecimalType):
+    guessing_weight = 0
+    result_type = decimal.Decimal
+
+    def cast(self, value):
+        value_without_currency = value.split(' ')[0]
+        return DecimalType.cast(self,
+                                value_without_currency)
+
+
 class FloatType(DecimalType):
     """ FloatType is deprecated """
     pass
@@ -134,6 +155,25 @@ def cast(self, value):
         raise ValueError
 
 
+class TimeType(CellType):
+    result_type = datetime.time
+
+    def cast(self, value):
+        if isinstance(value, self.result_type):
+            return value
+        if value in ('', None):
+            return None
+        hour = int(value[2:4])
+        minute = int(value[5:7])
+        second = int(value[8:10])
+        if hour > 24:
+            return datetime.timedelta(hours=hour,
+                                      minutes=minute,
+                                      seconds=second)
+        else:
+            return datetime.time(hour, minute, second)
+
+
 class DateType(CellType):
     """ The date type is special in that it also includes a specific
     date format that is used to parse the date, additionally to the
@@ -198,7 +238,8 @@ def cast(self, value):
         return parser.parse(value)
 
 
-TYPES = [StringType, DecimalType, IntegerType, DateType, BoolType]
+TYPES = [StringType, DecimalType, IntegerType, DateType, BoolType,
+         TimeType, CurrencyType, PercentageType]
 
 
 def type_guess(rows, types=TYPES, strict=False):
diff --git a/test/test_read.py b/test/test_read.py
index f4b73d1..c09a727 100644
--- a/test/test_read.py
+++ b/test/test_read.py
@@ -1,6 +1,7 @@
 # -*- coding: utf-8 -*-
 import unittest
 
+from decimal import Decimal
 from . import horror_fobj
 from nose.plugins.attrib import attr
 from nose.tools import assert_equal
@@ -22,6 +23,7 @@
 import datetime
 stringy = type(u'')
 
+
 class ReadCsvTest(unittest.TestCase):
     def test_utf8bom_lost(self):
         fh = horror_fobj('utf8bom.csv')
@@ -188,7 +190,8 @@ def rows(skip_policy):
             row_set = table_set.tables[0]
             return row_set
 
-        second = lambda r: r[1].value
+        def second(row):
+            return row[1].value
 
         assert "goodbye" in list(map(second, rows(True)))
         assert "    goodbye" in list(map(second, rows(False)))
@@ -308,9 +311,9 @@ def test_read_large_ods(self):
         assert_equal(6, len(table_set.tables))
         row_set = table_set.tables[0]
         row = next(row_set.raw())
-        assert len(row) == 5, len(row)
+        assert len(row) == 16384, len(row)
         for row in row_set.sample:
-            assert len(row) == 5, len(row)
+            assert len(row) == 16384, len(row)
 
     def test_ods_version_4412(self):
         fh = horror_fobj('loffice-4.4.1.2.ods')
@@ -334,6 +337,94 @@ def test_ods_read_past_blank_lines(self):
         assert_equal(rows[2][0], 'Jane')
         assert_equal(rows[3][0], 'Ian')
 
+    def test_ods_read_all_supported_formats(self):
+        fh = horror_fobj('ods_formats.ods')
+        table_set = ODSTableSet(fh)
+        assert_equal(3, len(table_set.tables))
+        row_set = table_set.tables[0]
+        rows = row_set_to_rows(row_set)
+        assert_equal(rows[0][0], "Date")
+        assert_equal(rows[1][0], "2014-11-11")
+        assert_equal(rows[2][0], "2001-01-01")
+        assert_equal(rows[3][0], '')
+        # time formats
+        assert_equal(rows[0][1], "Time")
+        assert_equal(rows[1][1], "PT11H12M12S")
+        assert_equal(rows[2][1], "PT00H00M12S")
+        assert_equal(rows[4][1], 'PT27H17M54S')
+        assert_equal(rows[5][1], "Other")
+        # boolean
+        assert_equal(rows[0][2], "Boolean")
+        assert_equal(rows[1][2], 'true')
+        assert_equal(rows[2][2], 'false')
+        # Float
+        assert_equal(rows[0][3], "Float")
+        assert_equal(rows[1][3], '11.11')
+        # Currency
+        assert_equal(rows[0][4], "Currency")
+        assert_equal(rows[1][4], '1 GBP')
+        assert_equal(rows[2][4], '-10000 GBP')
+        # Percentage
+        assert_equal(rows[0][5], "Percentage")
+        assert_equal(rows[1][5], '2')
+        # int
+        assert_equal(rows[0][6], "Int")
+        assert_equal(rows[1][6], '3')
+        assert_equal(rows[4][6], '11')
+        # Scientifed not supported
+        assert_equal(rows[1][7], '100000')
+        # Fraction
+        assert_equal(rows[1][8], '1.25')
+        # Text
+        assert_equal(rows[1][9], "abc")
+
+    def test_ods_read_all_supported_formats_casted(self):
+        fh = horror_fobj('ods_formats.ods')
+        table_set = ODSTableSet(fh)
+        assert_equal(3, len(table_set.tables))
+        row_set = table_set.tables[0]
+        rows = cast_row_set_to_rows(row_set)
+        date_format = "%d/%m/%Y"
+        assert_equal(rows[0][0], "Date")
+        assert_equal(rows[1][0].strftime(date_format), "11/11/2014")
+        assert_equal(rows[2][0].strftime(date_format), "01/01/2001")
+        assert_equal(rows[3][0], '')
+        # time formats
+        time_format = "%S:%M:%H"
+        assert_equal(rows[0][1], "Time")
+        assert_equal(rows[1][1].strftime(time_format), "12:12:11")
+        assert_equal(rows[2][1].strftime(time_format), "12:00:00")
+        assert_equal(rows[3][1], 0)
+        assert_equal(rows[4][1], datetime.timedelta(hours=27,
+                                                    minutes=17,
+                                                    seconds=54))
+        assert_equal(rows[5][1], "Other")
+        # boolean
+        assert_equal(rows[0][2], "Boolean")
+        assert_equal(rows[1][2], True)
+        assert_equal(rows[2][2], False)
+        # Float
+        assert_equal(rows[0][3], "Float")
+        assert_equal(rows[1][3], Decimal('11.11'))
+        # Currency
+        assert_equal(rows[0][4], "Currency")
+        assert_equal(rows[1][4], Decimal('1'))
+        assert_equal(rows[2][4], Decimal('-10000'))
+        # Percentage
+        assert_equal(rows[0][5], "Percentage")
+        assert_equal(rows[1][5], Decimal('0.02'))
+        # int
+        assert_equal(rows[0][6], "Int")
+        assert_equal(rows[1][6], 3)
+        assert_equal(rows[4][6], 11)
+        # Scientifed not supported
+        assert_equal(rows[1][7], 100000)
+        # Fraction
+        assert_equal(rows[1][8], 1.25)
+        # Text
+        assert_equal(rows[1][9], "abc")
+
+
 def row_set_to_rows(row_set):
     rows = []
     for row in row_set:
@@ -341,6 +432,13 @@ def row_set_to_rows(row_set):
     return rows
 
 
+def cast_row_set_to_rows(row_set):
+    rows = []
+    for row in row_set:
+        rows.append([cell.type.cast(cell.value) for cell in row])
+    return rows
+
+
 class XlsxBackwardsCompatibilityTest(unittest.TestCase):
     def test_that_xlsx_is_handled_by_xls_table_set(self):
         """
@@ -573,8 +671,8 @@ def setUp(self):
                 PDFTableSet(fh)
             except ImportError:
                 # Optional library isn't installed. Skip the tests.
-                raise SkipTest("pdftables is not installed, skipping PDF tests")
-
+                raise SkipTest(
+                    "pdftables is not installed, skipping PDF tests")
 
     def test_read_simple_pdf(self):
         with horror_fobj('simple.pdf') as fh:

From a01e9895e8fb546538992237e7e7cfe15c7be6c3 Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Tue, 7 Feb 2017 09:44:08 +0000
Subject: [PATCH 12/18] remove magic value and optimize empty row detection

---
 messytables/ods.py | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/messytables/ods.py b/messytables/ods.py
index 3a45109..ea7c86e 100644
--- a/messytables/ods.py
+++ b/messytables/ods.py
@@ -26,6 +26,7 @@
 TABLE_CELL = 'table-cell'
 VALUE_TYPE = 'value-type'
 COLUMN_REPEAT = 'number-columns-repeated'
+EMPTY_CELL_VALUE = ''
 
 ODS_VALUE_TOKEN = {
     "float": "value",
@@ -160,7 +161,7 @@ def raw(self, sample=False):
                     continue
 
                 cell = _read_cell(element)
-                if cell.value != '':
+                if empty_row is True and cell.value != EMPTY_CELL_VALUE:
                     empty_row = False
 
                 repeat = element.attrib.get(
@@ -193,7 +194,7 @@ def _read_cell(element):
         value = element.attrib.get(_tag(NS_OPENDOCUMENT_OFFICE, value_token))
         cell = Cell(value, type=ODS_TYPES.get(cell_type, StringType()))
     else:
-        cell = Cell('', type=StringType())
+        cell = Cell(EMPTY_CELL_VALUE, type=StringType())
 
     return cell
 
@@ -205,11 +206,11 @@ def _read_text_cell(element):
         if child.text:
             text_content.append(child.text)
         else:
-            text_content.append('')
+            text_content.append(EMPTY_CELL_VALUE)
     if len(text_content) > 0:
         cell_value = '\n'.join(text_content)
     else:
-        cell_value = ''
+        cell_value = EMPTY_CELL_VALUE
     return Cell(cell_value, type=StringType())
 
 

From 385d229dc6d2f81be91f7d926ae33ba4a280d90a Mon Sep 17 00:00:00 2001
From: chfw <wangc_2011@hotmail.com>
Date: Tue, 7 Feb 2017 10:10:06 +0000
Subject: [PATCH 13/18] better comment for scientific notation

---
 test/test_read.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/test/test_read.py b/test/test_read.py
index bfaee5b..ec4dbdc 100644
--- a/test/test_read.py
+++ b/test/test_read.py
@@ -371,7 +371,7 @@ def test_ods_read_all_supported_formats(self):
         assert_equal(rows[0][6], "Int")
         assert_equal(rows[1][6], '3')
         assert_equal(rows[4][6], '11')
-        # Scientifed not supported
+        # Scientific value is used but its notation is not
         assert_equal(rows[1][7], '100000')
         # Fraction
         assert_equal(rows[1][8], '1.25')
@@ -417,7 +417,7 @@ def test_ods_read_all_supported_formats_casted(self):
         assert_equal(rows[0][6], "Int")
         assert_equal(rows[1][6], 3)
         assert_equal(rows[4][6], 11)
-        # Scientifed not supported
+        # Scientific value is used but its notation is not
         assert_equal(rows[1][7], 100000)
         # Fraction
         assert_equal(rows[1][8], Decimal('1.25'))

From edf6fef643990a07dcd986fe0c4bf554c75944c1 Mon Sep 17 00:00:00 2001
From: Steven Maude <StevenMaude@users.noreply.github.com>
Date: Wed, 8 Feb 2017 13:40:37 +0000
Subject: [PATCH 14/18] Bump setup.py to 0.15.2 version for release

---
 setup.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/setup.py b/setup.py
index 16fdb73..40fe7e0 100644
--- a/setup.py
+++ b/setup.py
@@ -19,7 +19,7 @@
 
 setup(
     name='messytables',
-    version='0.15.1',
+    version='0.15.2',
     description="Parse messy tabular data in various formats",
     long_description=long_desc,
     classifiers=[

From 0eb9555d888af2145d36ee15cbaceaf7a3559411 Mon Sep 17 00:00:00 2001
From: Steven Maude <StevenMaude@users.noreply.github.com>
Date: Wed, 8 Feb 2017 13:42:27 +0000
Subject: [PATCH 15/18] Update CHANGELOG.md for 0.15.2

Add change not listed, and add date.
---
 CHANGELOG.md | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 62bead6..5e8f8cc 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,6 @@
-0.15.2 (unreleased)
-* #: detect ods types: boolean, currency, time and percentage. support repeated columns
+0.15.2 (8 February 2017)
+* #165: detect ods types: boolean, currency, time and percentage. support repeated columns
+* #160: Correct spelling of separator in source
 
 0.15.1 (29 September 2016)
 * #158: Add CDFV2-unknown to MIMELOOKUP

From a4cf5e57855955da1cff4eb26d0448250ef3897c Mon Sep 17 00:00:00 2001
From: Krzysiek Madejski <krzysztof.madejski@epf.org.pl>
Date: Fri, 24 Mar 2017 21:56:52 +0100
Subject: [PATCH 16/18] fixes #168

---
 setup.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/setup.py b/setup.py
index 40fe7e0..4f8f8ed 100644
--- a/setup.py
+++ b/setup.py
@@ -45,7 +45,8 @@
         'python-dateutil>=1.5.0',
         'lxml>=3.2',
         'requests',
-        'html5lib',
+        'six>=1.9', # until messytables->html5lib releases https://github.com/html5lib/html5lib-python/pull/301  
+        'html5lib',        
         'json-table-schema>=0.2, <=0.2.1'
     ],
     extras_require={'pdf': ['pdftables>=0.0.4']},

From ba346e098abd30ea3baabc7791fe885f09f2aa95 Mon Sep 17 00:00:00 2001
From: Simeon Walker <simeon-walker@users.noreply.github.com>
Date: Mon, 24 Apr 2017 15:40:59 +0100
Subject: [PATCH 17/18] Add another Excel mime type

---
 messytables/any.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/messytables/any.py b/messytables/any.py
index c497391..fd9dfc5 100644
--- a/messytables/any.py
+++ b/messytables/any.py
@@ -25,6 +25,7 @@
               'text/plain': 'CSV',  # could be TAB.
               'application/CDFV2-corrupt': 'XLS',
               'application/CDFV2-unknown': 'XLS',
+              'application/CDFV2': 'XLS',
               'application/vnd.oasis.opendocument.spreadsheet': 'ODS',
               'application/x-vnd.oasis.opendocument.spreadsheet': 'ODS',
               }

From 93f49b1367793e7459e3dcb07793828f4e0218d4 Mon Sep 17 00:00:00 2001
From: Sam Hatchett <samhatchett@gmail.com>
Date: Tue, 2 May 2017 14:08:14 -0400
Subject: [PATCH 18/18] fixes documentation for headers_guess

implementation uses column count compared to the column count mode (not row count)
---
 messytables/headers.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/messytables/headers.py b/messytables/headers.py
index 4434618..a50ebc7 100644
--- a/messytables/headers.py
+++ b/messytables/headers.py
@@ -20,7 +20,7 @@ def column_count_modal(rows):
 def headers_guess(rows, tolerance=1):
     """ Guess the offset and names of the headers of the row set.
     This will attempt to locate the first row within ``tolerance``
-    of the mode of the number of rows in the row set sample.
+    of the mode of the number of columns in the row set sample.
 
     The return value is a tuple of the offset of the header row
     and the names of the columns.