pyOpenVBA 3.2.0__tar.gz → 3.4.0__tar.gz

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pyOpenVBA
-Version: 3.2.0
+Version: 3.4.0
 Summary: Read and write VBA macros inside Excel, Word, and PowerPoint files in pure Python, no dependencies.
 Project-URL: Homepage, https://github.com/WilliamSmithEdward/pyOpenVBA
 Project-URL: Repository, https://github.com/WilliamSmithEdward/pyOpenVBA
@@ -200,7 +200,8 @@ path controls the format:
 ```python
 from pyopenvba import ExcelFile, WordFile, PowerPointFile
 
-# Excel - macro-enabled workbook (.xlsm) or binary workbook (.xlsb)
+# Excel - macro-enabled workbook (.xlsm), binary workbook (.xlsb),
+# or add-in (.xlam)
 with ExcelFile.create_new("new_book.xlsm") as wb:
     wb.set_module("Module1", 'Sub Hello()\r\n    MsgBox "xlsm"\r\nEnd Sub\r\n')
     wb.save()
@@ -209,6 +210,10 @@ with ExcelFile.create_new("new_book.xlsb") as wb:
     wb.set_module("Module1", 'Sub Hello()\r\n    MsgBox "xlsb"\r\nEnd Sub\r\n')
     wb.save()
 
+with ExcelFile.create_new("new_addin.xlam") as wb:
+    wb.set_module("Module1", 'Sub Hello()\r\n    MsgBox "xlam"\r\nEnd Sub\r\n')
+    wb.save()
+
 # Word - macro-enabled document (.docm)
 with WordFile.create_new("new_doc.docm") as doc:
     doc.set_module("Module1", 'Sub Hello()\r\n    MsgBox "docm"\r\nEnd Sub\r\n')
@@ -329,7 +334,7 @@ modules, `.cls` for class modules and code-behind.
 |-----------|------------------------------|:----:|:-----:|:----------:|
 | `.xlsm`   | Macro-enabled workbook       |  yes |  yes  |    yes     |
 | `.xlsb`   | Binary workbook              |  yes |  yes  |    yes     |
-| `.xlam`   | Macro-enabled add-in         |  yes |  yes  |    no      |
+| `.xlam`   | Macro-enabled add-in         |  yes |  yes  |    yes     |
 | `.xls`    | Legacy (Excel 97-2003)       |  yes |  yes  |    no      |
 
 ### Word

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/README.md

@@ -164,7 +164,8 @@ path controls the format:
 ```python
 from pyopenvba import ExcelFile, WordFile, PowerPointFile
 
-# Excel - macro-enabled workbook (.xlsm) or binary workbook (.xlsb)
+# Excel - macro-enabled workbook (.xlsm), binary workbook (.xlsb),
+# or add-in (.xlam)
 with ExcelFile.create_new("new_book.xlsm") as wb:
     wb.set_module("Module1", 'Sub Hello()\r\n    MsgBox "xlsm"\r\nEnd Sub\r\n')
     wb.save()
@@ -173,6 +174,10 @@ with ExcelFile.create_new("new_book.xlsb") as wb:
     wb.set_module("Module1", 'Sub Hello()\r\n    MsgBox "xlsb"\r\nEnd Sub\r\n')
     wb.save()
 
+with ExcelFile.create_new("new_addin.xlam") as wb:
+    wb.set_module("Module1", 'Sub Hello()\r\n    MsgBox "xlam"\r\nEnd Sub\r\n')
+    wb.save()
+
 # Word - macro-enabled document (.docm)
 with WordFile.create_new("new_doc.docm") as doc:
     doc.set_module("Module1", 'Sub Hello()\r\n    MsgBox "docm"\r\nEnd Sub\r\n')
@@ -293,7 +298,7 @@ modules, `.cls` for class modules and code-behind.
 |-----------|------------------------------|:----:|:-----:|:----------:|
 | `.xlsm`   | Macro-enabled workbook       |  yes |  yes  |    yes     |
 | `.xlsb`   | Binary workbook              |  yes |  yes  |    yes     |
-| `.xlam`   | Macro-enabled add-in         |  yes |  yes  |    no      |
+| `.xlam`   | Macro-enabled add-in         |  yes |  yes  |    yes     |
 | `.xls`    | Legacy (Excel 97-2003)       |  yes |  yes  |    no      |
 
 ### Word

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "pyOpenVBA"
-version = "3.2.0"
+version = "3.4.0"
 description = "Read and write VBA macros inside Excel, Word, and PowerPoint files in pure Python, no dependencies."
 readme = "README.md"
 license = { text = "MIT" }

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/src/pyopenvba/__init__.py

@@ -204,4 +204,4 @@ __all__ = [
     "push_word",
 ]
 
-__version__ = "3.2.0"
+__version__ = "3.4.0"

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/src/pyopenvba/_host.py

@@ -431,6 +431,7 @@ class VBAHostFile:
                         rename_map,
                         add_modules=add_modules_for_project,
                         delete_names=delete_names,
+                        code_page=project.code_page,
                     )
                     cfb.write_stream("PROJECT", new_project)
                 # Rewrite PROJECTwm to enumerate the current module set in

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/src/pyopenvba/_templates/__init__.py

@@ -1161,3 +1161,171 @@ _EMPTY_XLSB_B85 = (
 
 EMPTY_XLSB_BYTES: bytes = zlib.decompress(base64.b85decode(_EMPTY_XLSB_B85))
 # --- end XLSB template ---
+
+# --- XLAM template (generated by bake_xlam_template.py) ---
+# Source bytes: 8400 | zlib: 7791 | base85: 9739
+_EMPTY_XLAM_B85 = (
+    "c-nn=1yEhf(k>9(-Q8V7aJS$R+%-5GC%A9i-"
+    "Gh4w9&AH!LU7j&!9755*gVcX_3HeG|Gu|Yt(mD=>+4n3Gu=J?wVL7^SX?M5D8!eC3hLp#PMi"
+    "fa6qFb&6co<OjDZxu$=%Ay-9*#d*~-m`{gb0ZZOV5Q5-"
+    "v>O*?a$pv^txU$v0JM?@0B19Na!*aMm;E@m<I65KWJ*6tBV3Vq#);Eh6}W&M54KyUAVcJay@"
+    "~J$8G?l|r`#q?g@yCr&1=b8Mx9VQA@(f07Rva)5Hj05@=@M6T2aN_F~Qa&f>G@b!j?z}$I5o"
+    "(kPBBsx6+v(VEc2hRiH6l&XQnm*QDF=sR{`J5byBy@T0B&EwIXMm;$yn_5xs4RVJv1wydr?F"
+    "guXKRSnPWS%YHv%)fl-"
+    "$oEajWE8%h9&&H2$ovqT7F>&4j&;oNFGs`t&V&J|bxaN(_#FqiPE@Y-"
+    "QQ4^(S#XCPGLUkwi>#fH%1Vnx=6sq9%>2=9I{3y3BZ<qwh3E2nc(d24*D9DpVqxh3r^T+%dS"
+    "?BoJX7?a?S@;z&;UOWsYa{ebl!C0*oTA#llO30Q@2hFN$RFcZ`egKfh!MAz8kFUdGNH#}!<O"
+    "jh5h{cBB{wK!>xUq(tWP*5l@y@{)pgBu6?Uv2I5cZD}xnD84N8$w}iUe*Q@(%C06Ve8WK2!t"
+    "sIPK&fS$rUTD1Q<mq&3`tY_6h*NrPJMwIibB@w063DY>e@;H8G*U42X*dLX~E8dZ+Zf^EOP!"
+    ">Al^<&-|(sUD=H2?9c?6K3)5EIo6Nc;-yLbF6a<S0X4R`ZD^dVDEL9;m34E-"
+    "8uVJxhm%@hXDi?5dy8L+?K{rSK#KlOxcIAn#O*fMCI=lR#b1VG1ZpBX!PW3AyG)AA@0obreg"
+    "~Dy<c!K3OgM5m(<!D+zTdsfuj?0>ycbh#QhK!cm_U(Su9I}I5)wnPFADu{rO5p{edF^63W^X"
+    "33JT+8=92@57r@ot900KY&n8Vb0Ia;j9Red93t{?MODm*0ASIy~EkG&jlo&Ta!^7wou~E-"
+    "j(c-Zn-mf7rfoM<2W1BdVKl$AXwTdB`1Q6OGk+)43IZD1~=Y;KKTGn{Gf(dE<IYkQNsQZC?>"
+    "78(xs$m~S3^=f3wfWF&tp&rh!m3zGmuLnqy)-"
+    "6zLCK_v3;@)3(%0fC=Fo3$b}#2=@fc{UEVZ&Y*5k!N^P_@2B$e%0iVP`toLcDQ;=mlB<0*1<"
+    "rxW!5JU1ZlExR%TBqd|u^)s?Lu~%4G4e^lr+w!|wbNdk~eoO*XWYiyTLa7N+e~3yGM?OeZ=;"
+    "Kz+oeUb{I(DhvN8BT|AbQV)NOYx0vkg!T@b}^T@MRbkYUsp?-6$x-"
+    "^6OnXD`}I<MkB}|p}ISXafd!?pQs^cprcAu%BjHxW>DoyP_B>#TB!LG+(@QVxqo$>pkiE3?s"
+    "f=l=^7r&SDy4sSKFq`{49J$ooZ$LP7!fH)Dt9O?;;FkGFyEzs<7Yjgn^mr*XLuKj#q~f1)Lm"
+    "g2|a&b5kr{_tOgS0H^ptn-N_v0Ey~YDhlG+q035grDUqAMTq)UEU|Qx)GAqUFED<kFuOc`p9"
+    "<Hi<?;E$q97$`Nn6=kt2(Mhmv~Pe5d6vMVIsJy+&4A-B<1WEBHMdzGdZO#8=DChC&dxd-"
+    "zmY!pu-vK?+>5b)cV@_LSPP%@lFbt#OAq5G3|VVx$0@CpwCQm3Yx3VVH^{=oK2*8a^qh?04x"
+    "32*L>u#jJN6-"
+    ")I=Vit=x0I_4i<QHt$!lVGCBB)xwNvXkK}C%H7Fho8#3al1xytG9Kr=`XC9H}cVNk6mF4BqQ"
+    ";HY;-"
+    "XiAn`@7%tN1ssRC`?hD=whO4rrxYAYQY=Fx8?D1*xAcjLZj2Hp50GBNbIJSf9qf3efU_Zdrp"
+    "~sf(Gtax~%_JPbu0d|ITxd%hQi!Dt9{)o1&Qin$dM;)y0}tnVV0HPAI+G+ZyC%zdV+8fVW^V"
+    "JtIh1aJTOB_H^yq%?F3_oBWXExbq`(cI~#c;9je>Abmky#)9>Iz>oABue-MKDUV#)y$3lyzA"
+    "3TM<1;D5qd=1Tr=Aaz{>PSBuDM~<f}_?p`4$8M?y@);gGApOXL@$NeDVW#I@*!pW_5yKZ@f<"
+    "UpRve~pTSULV+X{dOdftJdq?jL9ufX+ron%ZS?*p&<uAUA_-"
+    "`})+kO2{WBuEeB}=IEabn8cykGsEE#jhqDOsVyhgc~T)(dTDot(LxKWw4k>MI@do+Uh5MyL3"
+    "h;JIsnRVUQTJR*q=4o)ULw8ylfrBLg)wg^}_Xrn}`mJCjlle({}6QmC+ni<j97{6wPKZ9Y}*"
+    "EpmowVWwj$W>1;`~9VG-"
+    "{V)TIg77jOS??l@TD(N0^euhCEO!@3B!d}%BuRIXNM0nx}mC;qQbzpp^3S&5vWDe^#dgviH+"
+    "Zd5-)ED=*}yp5^($}Fsxk03%`~0F$}PIMcQ86*6N;4=w)??S$6U!WoD-mU5JspJ+8@iOZ&*V"
+    "(Q<k$4FUlT!rpZkDF0SifRiGq|0MwT{^i^L@oR22R#xtA9RKOK{?b|Aj3Qu#6BTOXA?OJq*h"
+    "}ovE|yUyhgBDYB-"
+    "&%eD9P3!l8!ienxF6*jKZQQNSQ&+7mY18mEQF2%wyDeN>BzUg<sw?`JOqgQsD8vE$TE#fJ2E"
+    "%pL=*S68$uzK#UFD((8-FpPGeQ7gFN04r=*a2Q?Ie4-sVxjxyRPXXXtazXW57P-"
+    "WVcFW%<S>odM#DEvynRZ*4N&pXm3yQLaAp#4c^khjXsT6r|VS$}a26hnKU@~K&?_F^aqW~Sj"
+    "&6`EeXW0F)sT)1VIRo9pvKyJ@b5@N?tTS6z7mOyt9C}I)_5zb<eG^X_>Nr}UWsk7tn@nY_t$"
+    "~LTdBqLdiROVRx-QiHLLl+@-"
+    "ckIGrypu3>s0fN5Tyn=A<2zt&&#ApQHhV`=Tb|SX)ev@niCdm7fPh$_6EzDpd8bH|cnpds1}"
+    "`Ea5f!c*k9;+d>(cxogiRrVE&_GY8PF=-"
+    "VRliuorqpKqIE)Vm~1_&P}2{(65Iccr?vJ8eThFG3I)L=b%<9mYS8Wd*3|a&p#G;MR)V~W^P"
+    "I2E-0{u2;aWVV*RbJ>9|($&zOs3%-"
+    "V58QrP_p|)KFjSeV;trxncuD{d?uFO`9Zs!$Uz~QU87PdRcjQ8!JbvzuW&T`icHT!d43&h+z"
+    "Mja{siM)&Rn#tGrRLbUc2H{;|_w`<8VI+lx_y@_>lsJIuiDtYR>X?UZ8hx+`KdyiE{n+n)qI"
+    "%9wXgG|1`=tjq2n8L}KdxDQ5P9h~_--jd(lvL0+ARHy3dM56q(G3~%0U6FiiiB;xeDpsIOBC"
+    ">*wZ37#oj=N}2dyEkuco!4lOpB9BZJ&nC98dWbhpDmn<A*RL1qoqOqN3_lRzg*YXoT3IvB*B"
+    "&G8cvH{4MuvCQ~~DTn*gC349b%YWwt@rUbr02H4U_g(R=AFq7@Fq(QaC{pqpisAfVeCnIR2V"
+    "|yhUPZ22=TLaiIQY`Yt`1{73A&eZnsNpxO#DFS@g0XxgyU4)Sc_+0XPThhh%LB6bwy4wes92"
+    "f^To9*p)2zl&d%)0vH(d)nGdo(#gHdbgmi@KC!}UX6;lliHWKF}Cn?3#$)j<$Y=ku>=a32OK"
+    "U=d*Z367av$oKZza`K9v*f<zF<X}g%edgp!kgh$1mC)NV>_&*JT!?+sK(n!>9a%IjVqPDagQ"
+    "$R;S<6|%+-}4Jh>uy+LvERjWFIRdYwBn~q(-l1DwKaD7(e6Lc~~rrE{AA$T*QLwYt%-"
+    "r7)EN%{guL6{DgSB0|4^Aysze~>kha)C<hon&k25ap+@VJ6W<a%dzc8ksS|rzzTX2w9-"
+    "pkc`L<-"
+    "pp07E&y&q1%fy14%mQS0bttS3=YZQSG2^^w4JqV<`9=E9F?{!tDnmxufK9@mH=7Qt76rt=<-"
+    "|Hg7Z%ES<|2`m&MB<ZfM=^0$qQUvyx1*?hB-"
+    "5LLO%kM<tobINTNW5gQhrHGQ7xKB7dIAqX+`FR?rcTu+k~?*!mX$ltxPY6G;NuW^%>h-"
+    "I;rX6Y>`e9etC${K0suE$~uIt<cg~W7!=2zwb}MIg?tkiu|^rj8p#GE;BSUT#Fa%+%tD6r7T"
+    "%mYDsp!a!;FcFml>Eko5VnglQJB`i@_U#G7>x($!U*GMWvK$?N>YE4+ERi7oad-"
+    "8NZEflAXxVqe~|6F{2oE`ey~vOiBFHdqZy79zly~#*hOcS^!T-"
+    "FgfZCO6>MY%at1`!!BHMVB>ZfTW}_bV!;x6*wU}(16v0=|GO{oo80s1*GQ7EU#KL9eIQ9l6W"
+    "ky1pZF!-"
+    "281bryixD<$5wEWEGydwF?Cs+=zN3CJ;f&ud>GDag3Ed^^a6Y;IkHA_Ei9P*WqTFb(#-"
+    "P#`i+s+)ZtzYKIuU%H5V<Mbql=Zt!owp_K*uX=gC@u*gm0$kMOWr1V2%~dYCJ;@=Wr~rzZm#"
+    "WDJPCQT%6|`X*N^y)={n%mMi@d*0`d;0phBgE_aGtMS6l27?{a-Y1OtcNV(SJklu-"
+    ")4QOiyKYuT^pl;3kX_K{HVzwz1>9H*&~_eh#mI8-"
+    "{XrB(X*k>v$DLa9iiYs}6fRB?akS>z#~;a9>K!d@p)mp#_AxwnEUYOpu-"
+    "v$dUx=yjOAmB)iEx5X77-"
+    "k)@9nel8!9?u_MA=~dXu2qPP`tgOw5^M)>XWGNt(Fp^y@27hTB@P)Jc>Grnn>bTh^7svX}BI"
+    "_DMAVtWs!9)yRaK+DQa_SOpM7vtq$WsFRn6bVr@$$=xUNT-"
+    "o{p53W|c@NJc)9or75?M69+5l=k9E;SD;iFOyl#j#Fj{{8K31tf@b2J;?YDsxtls8EjKpluD"
+    "zvSotVE;Qk2%Yw;%OwWkXH(MGGb48W}Rn{RRl;~1BVdcVmia!5%6E@#}MfPmXGCmsm+YM)7H"
+    "I7@KRgDqjk3$ZFB))7c7$WVwZmJk{D}!xh;MTCchY8mL_PE@a)g5M@#oBG2>$>`-"
+    "0%E=?x`wjpkPEg%N|jR($sus|FdC^GlVV!SOIv4W;7KU=ayedCS{}(pP)U4Te9Mq{<Mm@fv-"
+    "xM0pGCvbjT1k8z<rcj$~=#veO4)%pz!He>IyU+wlS%Qej^Hmj_MXY7FI3$v!!=A-O`1m#c-"
+    "_hQ&RhcY^mGFu`DNU__d>o)X_#7KW$Us5k=%sP~@t6u9e1_D9Cl>k2?t|#X6ab+<5By0XV8M"
+    "o&b8CSmiO}BrQEBZc3fGVK@+?ZW^zGDzCD!G|PxtlXsIr$%nZP`6(7#HUacwRnovoa&Jh*=A"
+    "TFT;x#*=xO_1iUU(4Ll>u}V8?nyva9gOPvA_YtAC8;iaR&)K1p6XTuRWT2V>UF6%+^JfQCE9"
+    "W8lL?Y!o-h<dB6euYUv8*vx^;cDxbO@%#*EKA$}`sY<-MI3O-"
+    "OmsB<_{2wvO#B%F6|T+Bo4MeJz3!3W<rUomx{3|ybhYZ@Bgn4`z6=0?Z>E!O<l3Vf9Zt6nos"
+    "r@x#dbzXub+TQ`v&E4C<%I&YvSegd5U*W`jN%$Tp!q&<3zMRlWXxnBJBO4Gkg?2Hw<VGGwB&"
+    "E(SEk27{GPp<%y<5)MclWwqz43cbwUC!=I#i1zOMt0rj`;1165F5lAaMQ<8|2_ir3EG8F_dq"
+    "6=*BY%6p@#{BNTkOzLQLqM$mlgt!`n4&709}QMPTTsSdGPKRcr8?YH$u^-g-"
+    "J=+v=?1T>Uv_%U~@?kOV8uf11f)7056&FM;hKjgy&SMA=|JF*(c8~31>*T9gCJHiP1vQGP&%"
+    "&8YO3}y|gG8ZS>604stxLVr)b$OFrb2{U~cyJjAy1cIAN^tc+IA(r{$%8g$;#Muo1ej8jd_s"
+    "ZqyD}M6xVcjJbf*Z5TzcZMR7EA+)4CgchUbZ4j8jVIR9j_NgTV+NYrT<@fJ{_3Qmjbv#^+Re"
+    "NRAH9MTJV;{0hWWJM#M2nuttHwpll;hZ^sZ(647eJ8-"
+    "kIm(VkNPc*)cOXLhoOThsysWq{+Z6Xhhas+hAeB|op13oy~0Dat$b&_z#N;m_OF-viBRZ>(r"
+    "`GHbDRGkf9x9_LKpIfUmAs}{BREd3v4KfqPZV=JC;6N3U#k0YXFKycelm3M$_ai91lPI^CUv"
+    "fyXOq=nhQJxUK$0xUwG+(PzoPw>=2t%^@99W)am~?xlAiehWS|dzH)r&-"
+    "d!^9^OW*Yn$yr1^Ogur#C_A=0c>ujQ9>0VaY%3A|(73x0Wk2MT%|CnAGaMc_CMGJcJxJK^WL"
+    "fhwHrz2C@Q;CjVk!nOZPbjVnvd)nRcz*iiRp_6-eXb1Ls_2Y%N!@iIR=%ueW#-"
+    "9cy>xfUNL$LfK8^#(#B!1kSi5oHW=K+4d*n{GwFa3zWYG5l#E{-XXePn`-"
+    "oS;k@^1MEFGKp5Q{LYj*wfrh%@ts0W#P_lZtHZCeFng}AsaIB^2D~_3`^ujEaXK$(T+a2lM6"
+    "9&&RE4@m<+R2mR-jFk$bBvd~9fG!Wb*dZ8m-"
+    "JIYl=@fx=<ty2hS2N@Y3!os{r$8Je>2Qb}dbJ6qqyK+{k6`4iuR;Sn{(8M;pzyCJJLr<V}Or"
+    "@-->yY(wEN65p3*y$}o>!y|PYE4aX+c-UR@N$Co-pQhqHZz{tGO8x=t%k(%Vemwd*r>BqY>^"
+    ">ZfA=uv^(yi0sLw(6;tbt)>fa>Hc=E^2b_<~D(q<_oy3HCSo5~#7<ZZXWJZR~=FBR<++=nBx"
+    "qScmS9)SSgW1cH}6_W&XZEuCx7WNGGxo@Kg)uqI~Q8bLXQZB~oxY*L~u|3&-"
+    "L??Sj)I#veBkpzraL9eahZ@w}O0nH*qrb|x0-"
+    "FboQ+c?LVe>sse%;yy3R&vR){$L$i@5N0@PBAVIMr^PS8NtE&!vFj3)J!OeRy<3*#Cid@6eH"
+    "%liSD@DmRBC#|`(>l$nq?XI!1QHM*Rht@=unapR`D_@hGdW4x=1dTIrCawKhE6Ek_iyky5J{"
+    "fTz!r-{eAo|PwWd+wBsFRk(8AInO-X0ke4EuDt#WOytdxKr-"
+    "R^NE<UmtuA`MWrqaSt%X{qurqA_<8KT6+im!W$I;)rPM30vF@t)Znt>NwIw^^xsHLa4)gb`l"
+    "#N0mLty*;XQbKgTZjE;^~2{-tt&A>$LnlQX~bu-"
+    "D!vcNj&iq(>U3CIy|7)NIN}pmYpt6LKjoW_5_{gYhYQ}d)@9x`ght*qz;1E&slT@J&9>1RUq"
+    "QN8M>{yNqvP~!GH`hAxk%G-"
+    "?~?oH=+%9KKlSxDs?GX+aOtJYe)q+q`otmr^Gc54mWj!hnXqV-"
+    "4{~i)ITg#O;W}=J{?2UR@Z-"
+    "H%Chw&98??9oX@1j%;y!8V^&YK?GE$JIbG+lB9DQ@7i`>kI@hNcxGJ16RtNgNJnNNdVAIw#L"
+    "OfT3XQFol7B6XN?F^}KI*!+;aGC$(DqbPP~28u$v*lY=H2xQS{u(UR)?YwK~><UBEW4*|vfz"
+    "8}q;yXV!E9%Y$w_ztRnfV#l0l_qY)aBeiDg;yh3K`R-"
+    "UH$!^oapVlw9L)_IE=hQ;R3k#TFBlk-O0BD5#mh2xkfF>L_*eMpAE&^O;HPkvvTh9>-"
+    "m3<a0`^-$Ho_=^O-O!?!{H)(-"
+    "*S1u&vf6pQ6a6MiPT64rRTT_K#TX$<AFrxKA;5KcAQbB{}m(b4sI($Cz<zhya(EZIwk@4Jym"
+    "9a^*ZSJVM`{1^wUiP>U#`#-"
+    "X3hARy#)fWSUxjD!bE<+mYyDSLq=q(7C7nLb8ad&Akz^v20x7~l(MT%Dj$aK6iPYcSc79~GR"
+    "NXt_uaufLEJP>AK|f;3W>XBsBg(AlaEQ7p(kw}ad&(L_nrN+W&fwINe8RSxf!ed|t{jZi8<1"
+    "lgLl9rxG!O}5JxNswwGS>KMI>kG$6())xqQRKFFd-"
+    "2!jHv!$nR<?VIX?gjhIqpuW!aHx4D0o_bL+(@xIDXKBF@Rpi8_R0k*CSsp<~=8qYjuuJY3x("
+    "sECw?N&0@m)@KItu1WD2KGR|aXE=u?Nm!;wHF`$|yS*a)?4}tJWr#}?BRpqO&99VKYlr*0@8"
+    "|tJvQ<v>;eoI%I0VXYi^OVLTU;@^%vsp5EJgf4rVkXcS*$?eBk?Iz8N^@NG5FTHQlLrsGsv?"
+    ">i_cuo7o!(e>{TNRRWmy;ur1)cfYU9VZ)tou?ZYkb8_{!ntBf4*iwQ$uXNDfuQ-HlWkwhm53"
+    "mCw}E?v*FW)jv=*t16dRw3o@bceRJZKmyqC6i!t=kAT=m2T=#-"
+    "Ha{_Qsv=+@LC=dpS`DFtL@T)?*#jGeSt;}9L4j?`NwaOb)cAA0zZ7&88Pc3-"
+    "i%R>DXvk4m6bg#yXs~1sPIJVZHF?ozZ=`zymHrspzT>eX$GAQ6<taTzWAAPb>RwK0HCPi>8o"
+    ">4qHIMm2>vIjCQgl05)mo)WYwS_>O&K$k$RRq{Xy)9&R)3swl|@9T;ou@3G6~L&mGkzU^TbU"
+    "9A$;V}=$&o|JQI1G2iYH`!Bi_kAd{Pxax8dB-`dmI+V~(RyHnb-sm&HN4Ycz%M{dG5DGm>_S"
+    "I?aF)jsTc*yY$%R{>8ks}~%rgnqqw&<uo|T0nu=;K4=zhcj^y&5$2QCzM~y7ya?%;^EM~8z_"
+    "ieFnQ|}i@_6qIPX0m5b7Bx{#110IW(?YGoz`iz+xTNsb<($lyrU>(!BrYd99leo(nG=DD#Nx"
+    "QnN&+iZv0g)DPz<^7Xqv#DeDQi2tdS@Z}V|gUc?p@QxwQ2Fo}xNhBv2`WQ;E@(Y0on(!8Rw<"
+    "4<ulu?h97u$irMs<)bl`r|?*8CbuSg?OjOptC4HkV?JYkv^`BwG{QF#@(Y_9rBMC{kFB@kX}"
+    "a9^y4RQT7SQHbLlw;>7z1@cqv`8=Sw9IrsY&`C(9nmV*KO3zN)Zg$022Wnu*5H^vC-"
+    "Q1NXZ@f~1>AV3uFTT|>zS97e)Q;1x6bA0!Pwiv|x;f!W`MSKTk;Q19@V=tV1#`i+h*~9!D%q"
+    ";;HFOl5Yjl%-qYW3HNG%nf5zLyghc!uzRL3iUgEhsHzU`ArBOt|lS%uXwZHcBAhG9M_W-"
+    "+Z_`@CCdl%y34vQ!~eW2TxJLi}!;k1e<de*|r7G$6#y*NkP(5meUGN(*^^2?C_x@S?EV?gbs"
+    "l0mXV6KR-"
+    "0b!l({KpN<H=sIQSz~Bf3)c+@Q~x`EU)`eCU>81%@4w$O0`137_aRtdb;lzXI1!t(~LJCwPM"
+    "p*M^g8LwNK&D{jN%o4_OkK}ovH*#^B2GD16=D^yO|NVa+ce%SV{|9Qi4b=kCg>#v|g4d)*sU"
+    "v2!m(iP;(QcBZZbbiH@bQGCWvDYM$ApLTACV`-&4n-"
+    "p%MV6o?lSk(!U+o`w^U8ThM7t=^GcjpfDqS-94~w-w>0DYDin8Ce)(<mf85638v<Kc?-"
+    "Bbj=(z2O65@z6wma$&wP|W|SWiw~z|Inbx-"
+    "<@7)P~h2{dvxGwT<#Y}ptv>{EsM|MS19?FM*5k0H9cJE`L8#BLLK`(I`VyJ0wI)QiJGlOTaE"
+    "bG0;u7-b*@PgUv=A^nfv?BPcL$YBb6EBhM>atgZpbiw?)@TG_;=LdJDJW;T2h1vI-5nZ-"
+    "Lr`RPJ$goV-BW@B`ZVs@!n0<qUY3(Q4uDq)DzY>>s22e7GNmlt&93^^3@E_s<vQC=2Iq2)i1"
+    "AWbjX;OJQMn)?$Jv#v@khCzIqoAljj>@ifD-L({kj$KDE-"
+    "Eu>UGyjhV<Gh5#+Y}>huGPu!9Zr8v<OdaY63q<%_$tHRM99^XrCL#-psFZ--KT5n%DanoeK*"
+    "cj|&jlY5;?*mm;#`TQWB^t|yvXF49RNE}yweyv0Tt&g%#HB2EK3n|S>n~zx+sm~Q@c(2<i43"
+    "j{tL?sxY;@0S|m72FVQ9?ROgF7>rX*-"
+    "_r1GTnP0;D`q6|i{+ZwZn=!cNqRv}%{jrAW?tb*1LQN4G1{do8T|WN5R}eJRe=jD#7Vx^(`H"
+    "z59=zmo{Ujtv)1OEY5!u+r5;A`;fD#brwHTeHu&EmD3*W~>_a)J^5-"
+    "&Fo<DX)p(f28;#{a0G}wSd=&@;?G{F<+(4uc5C~lz*T`IR7PKc@2N<6#s!Ull&JK`5OM(+W&"
+    "zklKyYoS5t(8|LZOC%kBU2LlN@7KK?(+6cj@"
+)
+
+EMPTY_XLAM_BYTES: bytes = zlib.decompress(base64.b85decode(_EMPTY_XLAM_B85))
+# --- end XLAM template ---

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/src/pyopenvba/excel.py

@@ -55,10 +55,11 @@ class ExcelFile(VBAHostFile):
         VBA project (``ThisWorkbook``, ``Sheet1``, and a bare ``Module1``)
         and return an open :class:`ExcelFile` for it.
 
-        Supported extensions: ``.xlsm`` (default) and ``.xlsb``.
+        Supported extensions: ``.xlsm`` (default), ``.xlsb``, and
+        ``.xlam`` (Excel add-in).
 
         The bytes are decoded from a baked-in template captured from a
-        freshly Excel-authored workbook, so the resulting file opens
+        freshly Excel-authored file, so the resulting file opens
         cleanly in Excel without any "found a problem" repair prompt.
 
         ``path`` is overwritten if it already exists.
@@ -68,6 +69,9 @@ class ExcelFile(VBAHostFile):
         if suffix == ".xlsb":
             from pyopenvba._templates import EMPTY_XLSB_BYTES
             template = EMPTY_XLSB_BYTES
+        elif suffix == ".xlam":
+            from pyopenvba._templates import EMPTY_XLAM_BYTES
+            template = EMPTY_XLAM_BYTES
         else:
             from pyopenvba._templates import EMPTY_XLSM_BYTES
             template = EMPTY_XLSM_BYTES

{pyopenvba-3.2.0 → pyopenvba-3.4.0}/src/pyopenvba/vba.py

@@ -24,7 +24,10 @@ Critical implementation traps (guide section 31)
 
 from __future__ import annotations
 
+import codecs
 import struct
+import unicodedata
+import warnings
 from collections.abc import Callable, Iterable
 from dataclasses import dataclass, field
 from enum import Enum
@@ -451,6 +454,22 @@ class VBAReference:
     libid_secondary: str = ""   # twiddled / relative libid where applicable
 
 
+def _prefer_unicode(ansi: str, unicode_value: str) -> str:
+    """Return the authoritative value of an ANSI / UTF-16 dir record pair.
+
+    The UTF-16 partner record is lossless by construction, while the ANSI
+    record is limited to the project's code page.  When the two disagree
+    the ANSI side is the damaged one -- either its bytes were undecodable
+    (U+FFFD) or the writer had already substituted ``?`` for characters
+    outside the page -- so the Unicode record wins (GitHub issue #12).
+    Well-formed files carry identical values and are unaffected.
+    """
+    cleaned = unicode_value.rstrip("\x00")
+    if cleaned and cleaned != ansi:
+        return cleaned
+    return ansi
+
+
 def _parse_dir_stream(raw: bytes) -> tuple[_DirInfo, list[_ModuleInfo]]:
     """
     Parse a decompressed dir stream.
@@ -663,16 +682,28 @@ def _parse_dir_stream(raw: bytes) -> tuple[_DirInfo, list[_ModuleInfo]]:
             # Do not commit: REFERENCEORIGINAL is always followed by REFERENCECONTROL.
 
         # ------------- MODULE records -----------------------------
+        # Each MBCS record is followed by its UTF-16 partner.  When the
+        # MBCS decode was lossy -- the project's code page could not
+        # represent the name, or the file declares a page we could not
+        # resolve -- the Unicode record is the authoritative lossless
+        # value, so prefer it (GitHub issue #12).
         elif record_id == 0x0047 and current is not None: # MODULENAMEUNICODE
             current.name_unicode = data.decode("utf-16-le", errors="replace")
+            current.name = _prefer_unicode(current.name, current.name_unicode)
         elif record_id == 0x001A and current is not None: # MODULESTREAMNAME
             current.stream_name = data.decode(_enc(), errors="replace")
         elif record_id == 0x0032 and current is not None: # MODULESTREAMNAME unicode
             current.stream_name_unicode = data.decode("utf-16-le", errors="replace")
+            current.stream_name = _prefer_unicode(
+                current.stream_name, current.stream_name_unicode
+            )
         elif record_id == 0x001C and current is not None: # MODULEDOCSTRING (MBCS)
             current.doc_string = data.decode(_enc(), errors="replace")
         elif record_id == 0x0048 and current is not None: # MODULEDOCSTRINGUNICODE
             current.doc_string_unicode = data.decode("utf-16-le", errors="replace")
+            current.doc_string = _prefer_unicode(
+                current.doc_string, current.doc_string_unicode
+            )
         elif record_id == 0x0031 and current is not None: # MODULEOFFSET
             if len(data) >= 4:
                 current.text_offset = int(struct.unpack_from("<I", data, 0)[0])
@@ -1122,15 +1153,148 @@ class VBAProject:
 # Write-back helpers
 # ---------------------------------------------------------------------------
 
+# Windows code-page identifiers whose portable Python codec is not
+# spelled ``cp<N>``.
+#
+# These aliases are not merely a convenience: on Windows, CPython falls
+# through to the operating system's code-page registry, so ``cp28592``
+# and friends resolve there but raise LookupError on Linux and macOS.
+# Relying on the ``cp<N>`` spelling therefore makes a project's text
+# decode correctly on one platform and turn into latin-1 mojibake on
+# another -- a platform-dependent data bug (caught by the cross-OS
+# language-matrix CI job on its first run; see GitHub issue #13).
+# Consulting this table first keeps behavior identical everywhere.
+#
+# Pages VBA hosts write that Python does spell ``cp<N>`` portably (874,
+# 932, 936, 949, 950, 1250-1258, 1361, 65001) are deliberately absent.
+_CODEPAGE_ALIASES: dict[int, str] = {
+    # Macintosh
+    10000: "mac_roman",
+    10004: "mac_arabic",
+    10006: "mac_greek",
+    10007: "mac_cyrillic",
+    10010: "mac_romanian",
+    10029: "mac_latin2",
+    10079: "mac_iceland",
+    10081: "mac_turkish",
+    10082: "mac_croatian",
+    # KOI8
+    20866: "koi8_r",
+    21866: "koi8_u",
+    # ISO 8859
+    28591: "iso8859_1",
+    28592: "iso8859_2",
+    28593: "iso8859_3",
+    28594: "iso8859_4",
+    28595: "iso8859_5",
+    28596: "iso8859_6",
+    28597: "iso8859_7",
+    28598: "iso8859_8",
+    28599: "iso8859_9",
+    28603: "iso8859_13",
+    28605: "iso8859_15",
+    # ISO 2022 / EUC / GB / misc
+    50220: "iso2022_jp",
+    50225: "iso2022_kr",
+    51932: "euc_jp",
+    51936: "gb2312",
+    51949: "euc_kr",
+    52936: "hz",
+    54936: "gb18030",
+    65000: "utf_7",
+}
+
+
 def _encoding_for_codepage(code_page: int) -> str:
+    """Map a PROJECTCODEPAGE value to a Python codec name.
+
+    Falls back to ``latin-1`` for pages Python cannot resolve, warning
+    so the degradation is visible: silently decoding a foreign code page
+    as latin-1 produces mojibake that survives round-trip checks (see
+    GitHub issue #12).
+    """
+    alias = _CODEPAGE_ALIASES.get(code_page)
+    if alias is not None:
+        return alias
+    encoding = f"cp{code_page}"
     try:
-        encoding = f"cp{code_page}"
-        "".encode(encoding)
+        codecs.lookup(encoding)
     except LookupError:
-        encoding = "latin-1"
+        warnings.warn(
+            f"No Python codec for VBA code page {code_page}; falling back "
+            "to latin-1.  Text outside the Latin-1 range will be mojibake. "
+            "Please report this code page at "
+            "https://github.com/WilliamSmithEdward/pyOpenVBA/issues",
+            UserWarning,
+            stacklevel=3,
+        )
+        return "latin-1"
     return encoding
 
 
+def encode_mbcs(text: str, encoding: str) -> bytes:
+    """Encode ``text`` to an MBCS code page, salvaging composed characters.
+
+    Equivalent to ``text.encode(encoding, errors="replace")`` except for
+    characters the codec cannot represent directly but *can* represent as
+    a base character plus combining marks.  Python's charmap codecs do no
+    composition, so Vietnamese text destined for cp1258 loses every
+    stacked-diacritic character:
+
+        'Tiếng Việt'.encode('cp1258', errors='replace')
+        -> b'Ti?ng Vi?t'
+
+    cp1258 stores ``ệ`` as precomposed ``ê`` (0xEA) plus a combining
+    dot-below byte (0xF2), which is *not* the character's canonical
+    decomposition, so NFD alone does not help either.  For each
+    unmappable character this helper decomposes to NFD and tries folding
+    each combining mark back into the base in turn, keeping the first
+    combination the codec accepts and emitting the remaining marks as
+    combining bytes.  Decoding those bytes yields a canonically
+    equivalent (NFD-ish) string, so compare with NFC normalization.
+
+    Characters that remain unmappable become ``?``, matching
+    ``errors="replace"``.  Text that the codec already encodes directly
+    is returned byte-for-byte unchanged, so this never alters output for
+    projects that work today.
+    """
+    try:
+        return text.encode(encoding)
+    except UnicodeEncodeError:
+        pass
+    out = bytearray()
+    for ch in text:
+        try:
+            out += ch.encode(encoding)
+            continue
+        except UnicodeEncodeError:
+            pass
+        decomposed = unicodedata.normalize("NFD", ch)
+        if len(decomposed) > 1:
+            base, marks = decomposed[0], list(decomposed[1:])
+            folded: bytes | None = None
+            for i, mark in enumerate(marks):
+                combined = unicodedata.normalize("NFC", base + mark)
+                if len(combined) != 1:
+                    continue   # this mark does not compose with the base
+                rest = "".join(marks[:i] + marks[i + 1:])
+                try:
+                    folded = combined.encode(encoding) + rest.encode(encoding)
+                except UnicodeEncodeError:
+                    continue
+                break
+            if folded is not None:
+                out += folded
+                continue
+            try:
+                out += decomposed.encode(encoding)
+                continue
+            except UnicodeEncodeError:
+                pass
+        out += ch.encode(encoding, errors="replace")
+    return bytes(out)
+
+
 def split_attribute_header(source: str) -> tuple[str, str]:
     """
     Split a VBA module source into ``(attribute_header, body)``.
@@ -1316,7 +1480,7 @@ def rebuild_module_stream(module: VBAModule, code_page: int) -> bytes:
     invalidation logic that Office performs on the prefix remains valid.
     """
     encoding = _encoding_for_codepage(code_page)
-    source_bytes = module.source.encode(encoding, errors="replace")
+    source_bytes = encode_mbcs(module.source, encoding)
     compressed = compress(source_bytes)
     return module.prefix_bytes + compressed
 
@@ -1368,11 +1532,11 @@ def serialize_dir_modules_section(project: VBAProject) -> bytes:
         name_u = m.name_unicode or m.name
         stream = m.stream_name or m.name
         stream_u = m.stream_name_unicode or stream
-        out += _pack_record(0x0019, name.encode(enc, errors="replace"))
+        out += _pack_record(0x0019, encode_mbcs(name, enc))
         out += _pack_record(0x0047, name_u.encode("utf-16-le"))
-        out += _pack_record(0x001A, stream.encode(enc, errors="replace"))
+        out += _pack_record(0x001A, encode_mbcs(stream, enc))
         out += _pack_record(0x0032, stream_u.encode("utf-16-le"))
-        out += _pack_record(0x001C, m.doc_string.encode(enc, errors="replace"))
+        out += _pack_record(0x001C, encode_mbcs(m.doc_string, enc))
         out += _pack_record(0x0048, m.doc_string_unicode.encode("utf-16-le"))
         out += _pack_record(0x0031, struct.pack("<I", m.text_offset))
         out += _pack_record(0x001E, struct.pack("<I", m.help_context))
@@ -1432,6 +1596,7 @@ def serialize_project_stream(
     *,
     add_modules: list[tuple[str, str]] | None = None,
     delete_names: set[str] | None = None,
+    code_page: int = 1252,
 ) -> bytes:
     """
     Rewrite a PROJECT stream's plain-text body to apply pending mutations.
@@ -1454,7 +1619,13 @@ def serialize_project_stream(
     ``[Host Extender Info]``) is preserved byte-for-byte except for the
     targeted substitutions.
 
-    Returns cp1252-encoded bytes with CRLF line endings.
+    ``code_page``:
+        PROJECTCODEPAGE of the owning project.  The PROJECT stream is
+        code-page ANSI per [MS-OVBA] 2.3.1, so its ``Module=`` /
+        ``Document=`` declarations carry module names in that page.
+        Defaults to 1252 for callers that have no project handy.
+
+    Returns code-page-encoded bytes with CRLF line endings.
     """
     add_modules = add_modules or []
     delete_names = delete_names or set()
@@ -1462,10 +1633,8 @@ def serialize_project_stream(
     # callers may invoke this to run the dedup/normalization pass on a
     # PROJECT stream that was previously corrupted by buggy writes (e.g.
     # duplicate ``Module=NAME`` lines from a delete-then-readd flow).
-    try:
-        text = raw.decode("cp1252", errors="replace")
-    except LookupError:
-        text = raw.decode("latin-1", errors="replace")
+    encoding = _encoding_for_codepage(code_page)
+    text = raw.decode(encoding, errors="replace")
 
     rename_ci: dict[str, str] = {k.casefold(): v for k, v in rename_map.items()}
     delete_ci: set[str] = {n.casefold() for n in delete_names}
@@ -1579,7 +1748,7 @@ def serialize_project_stream(
                 for name, _ in ws_new:
                     seen_workspace.add(name.casefold())
 
-    return ("\r\n".join(out_lines) + "\r\n").encode("cp1252", errors="replace")
+    return encode_mbcs("\r\n".join(out_lines) + "\r\n", encoding)
 
 
 def _project_section_end(lines: list[str]) -> int:
@@ -1725,7 +1894,7 @@ def parse_vba_project(cfb: CFB) -> VBAProject:
         project_stream_raw = None
     if project_stream_raw is not None:
         try:
-            ps = parse_project_stream(project_stream_raw)
+            ps = parse_project_stream(project_stream_raw, code_page=code_page)
             project.protection = ps.protection
         except VBAProjectError:
             pass
@@ -1767,17 +1936,16 @@ class ProjectStream:
     host_extender_info: list[str] = field(default_factory=lambda: [])
 
 
-def parse_project_stream(raw: bytes) -> ProjectStream:
+def parse_project_stream(raw: bytes, *, code_page: int = 1252) -> ProjectStream:
     """Parse the plain-text PROJECT stream.
 
-    The PROJECT stream is Windows-1252 encoded key=value text terminated
-    by CRLF.  Empty lines separate the project information block, the
-    ``[Host Extender Info]`` block, and the ``[Workspace]`` block.
+    The PROJECT stream is code-page ANSI key=value text terminated by
+    CRLF ([MS-OVBA] 2.3.1); pass the project's PROJECTCODEPAGE as
+    ``code_page`` so non-Latin module names decode correctly.  Empty
+    lines separate the project information block, the ``[Host Extender
+    Info]`` block, and the ``[Workspace]`` block.
     """
-    try:
-        text = raw.decode("cp1252", errors="replace")
-    except LookupError:
-        text = raw.decode("latin-1", errors="replace")
+    text = raw.decode(_encoding_for_codepage(code_page), errors="replace")
 
     out = ProjectStream()
     section = "project"  # "project" | "host_extender" | "workspace"
@@ -1846,9 +2014,15 @@ def parse_project_stream(raw: bytes) -> ProjectStream:
 # PROJECTwm stream parser ([MS-OVBA] 2.3.4.4)
 # ---------------------------------------------------------------------------
 
-def parse_projectwm(raw: bytes) -> list[tuple[str, str]]:
+def parse_projectwm(
+    raw: bytes, *, code_page: int = 1252
+) -> list[tuple[str, str]]:
     """Parse the PROJECTwm stream into (mbcs_name, unicode_name) pairs.
 
+    ``code_page`` selects the MBCS codec for the first name in each pair;
+    pass the project's PROJECTCODEPAGE so non-Latin names decode
+    correctly.
+
     Format:
         loop:
             ModuleName_MBCS    : null-terminated bytes
@@ -1876,10 +2050,9 @@ def parse_projectwm(raw: bytes) -> list[tuple[str, str]]:
             pos += 2
         unicode_name = raw[start:pos].decode("utf-16-le", errors="replace")
         pos += 2   # skip the u16=0 terminator
-        try:
-            mbcs_name = mbcs.decode("cp1252", errors="replace")
-        except LookupError:
-            mbcs_name = mbcs.decode("latin-1", errors="replace")
+        mbcs_name = mbcs.decode(
+            _encoding_for_codepage(code_page), errors="replace"
+        )
         out.append((mbcs_name, unicode_name))
     return out
 
@@ -1903,7 +2076,7 @@ def serialize_projectwm(
         if not mbcs:
             # An empty MBCS name would terminate the stream prematurely.
             continue
-        buf += mbcs.encode(enc, errors="replace")
+        buf += encode_mbcs(mbcs, enc)
         buf += b"\x00"
         buf += unicode_name.encode("utf-16-le", errors="replace")
         buf += b"\x00\x00"