diff options
author | Karl Berry <karl@freefriends.org> | 2009-07-24 16:57:38 +0000 |
---|---|---|
committer | Karl Berry <karl@freefriends.org> | 2009-07-24 16:57:38 +0000 |
commit | e5cee0d6a57f5481608d21154b5e4aa06fdc1d05 (patch) | |
tree | d3e1b95c37a31c2721dcfb0f21b540656b4c9395 /Master | |
parent | 7e5630813bec518301a3f0ab57792217f3cc8503 (diff) |
new script mkgrkindex (23jul09)
git-svn-id: svn://tug.org/texlive/trunk@14428 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master')
23 files changed, 875 insertions, 13 deletions
diff --git a/Master/bin/alpha-linux/mkgrkindex b/Master/bin/alpha-linux/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/alpha-linux/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/amd64-freebsd/mkgrkindex b/Master/bin/amd64-freebsd/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/amd64-freebsd/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/i386-freebsd/mkgrkindex b/Master/bin/i386-freebsd/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/i386-freebsd/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/i386-linux/mkgrkindex b/Master/bin/i386-linux/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/i386-linux/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/i386-solaris/mkgrkindex b/Master/bin/i386-solaris/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/i386-solaris/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/mips-irix/mkgrkindex b/Master/bin/mips-irix/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/mips-irix/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/powerpc-aix/mkgrkindex b/Master/bin/powerpc-aix/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/powerpc-aix/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/powerpc-linux/mkgrkindex b/Master/bin/powerpc-linux/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/powerpc-linux/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/sparc-linux/mkgrkindex b/Master/bin/sparc-linux/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/sparc-linux/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/sparc-solaris/mkgrkindex b/Master/bin/sparc-solaris/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/sparc-solaris/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/universal-darwin/mkgrkindex b/Master/bin/universal-darwin/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/universal-darwin/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/bin/win32/mkgrkindex.bat b/Master/bin/win32/mkgrkindex.bat new file mode 100755 index 00000000000..db35d13b51c --- /dev/null +++ b/Master/bin/win32/mkgrkindex.bat @@ -0,0 +1,78 @@ +@echo off
+rem Universal script launcher
+rem
+rem Copyright 2009 by Tomasz M. Trzeciak
+rem Public Domain
+
+rem Make environment changes local
+setlocal enableextensions
+rem Get program/script name
+set progname=%~n0
+rem Check if this is 'sys' version of program
+set TEX_SYS_PROG=
+if /i "%progname:~-4%"=="-sys" (
+ set progname=%progname:~0,-4%
+ set TEX_SYS_PROG=true
+)
+
+rem Default command to execute
+set command=call :noscript
+rem Make sure our dir is on the search path; avoid trailing backslash
+for %%I in ("%~f0\..") do path %%~fI;%path%
+rem Ask kpsewhich about root and texmfsys trees (the first line of output)
+rem and location of the script (the second line of output)
+rem (4NT shell acts wierd with 'if' statements in a 'for' loop,
+rem so better process this output further in a subroutine)
+for /f "tokens=1-3 delims=;" %%I in (
+ 'kpsewhich --expand-var "$SELFAUTOPARENT;$TEXMFSYSCONFIG;$TEXMFSYSVAR" ^
+ --format texmfscripts "%progname%.pl" "%progname%.tlu" "%progname%.rb" "%progname%.py"'
+) do (
+ call :setcmdenv "%%~I" "%%~J" "%%~K"
+)
+
+rem By now we should have the command to execute (whatever that is), so
+rem pass through all the arguments we have and execute it
+%command% %*
+rem Finish with goto :eof (it will preserve the last set errorlevel)
+goto :eof
+
+REM SUBROUTINES
+
+:setcmdenv selfautoparent texmfsysconfig texmfsysvar
+rem If there is only one argument it must be a script name
+if "%~2"=="" goto :setcmd
+rem Otherwise, it is the first line from kpsewhich, so to set up the environment
+set PERL5LIB=%~f1\tlpkg\tlperl\lib
+set GS_LIB=%~f1\tlpkg\tlgs\lib;%~f1\tlpkg\tlgs\fonts
+path %~f1\tlpkg\tlgs\bin;%~f1\tlpkg\tlperl\bin;%~f1\tlpkg\installer;%~f1\tlpkg\installer\wget;%path%
+if not defined TEX_SYS_PROG goto :eof
+rem Extra stuff for sys version
+set TEXMFCONFIG=%~2
+set TEXMFVAR=%~3
+rem For sys version we might have an executable in the bin dir, so check for it
+if exist "%~f1\bin\win32\%progname%.exe" set command="%~f1\bin\win32\%progname%.exe"
+goto :eof
+
+:setcmd script
+rem Set command based on the script extension
+if /i %~x1==.pl set command=Perl.exe "%~f1"
+if /i %~x1==.tlu set command=TeXLua.exe "%~f1"
+rem For Ruby and Python we additionally check if their interpreter is available
+if /i %~x1==.rb call :chkcmd Ruby.exe "%~f1"
+if /i %~x1==.py call :chkcmd Python.exe "%~f1"
+goto :eof
+
+:chkcmd program script
+set command=%*
+rem If there is no interpreter Ruby or Python, suggest getting one
+if "%~$PATH:1"=="" set command=call :noinst %*
+goto :eof
+
+:noinst program
+echo %1 not found on search path>&2
+echo %~n1 is not distributed with TeX Live and has to be installed separately
+exit /b 1
+
+:noscript
+echo %progname%: no appropriate script or program found>&2
+exit /b 1
diff --git a/Master/bin/x86_64-linux/mkgrkindex b/Master/bin/x86_64-linux/mkgrkindex new file mode 120000 index 00000000000..2dc602dd180 --- /dev/null +++ b/Master/bin/x86_64-linux/mkgrkindex @@ -0,0 +1 @@ +../../texmf-dist/scripts/mkgrkindex/mkgrkindex
\ No newline at end of file diff --git a/Master/texmf-dist/doc/support/mkgrkindex/mkgrkindex.nw b/Master/texmf-dist/doc/support/mkgrkindex/mkgrkindex.nw new file mode 100644 index 00000000000..2b6a619bf15 --- /dev/null +++ b/Master/texmf-dist/doc/support/mkgrkindex/mkgrkindex.nw @@ -0,0 +1,466 @@ +@ The correct generation and typesetting of an index for a Greek document +prepared with LaTeX is not a trivial task at all. First of all, the +program [[makeindex]] cannot handle Greek letters, but even if the [[.idx]] +is writen using the babel latin-transcription, the generated [[.ind]] file +does not follow the order of the Greek alphabet. +On the other hand, [[makeindex]] is doing a fine +job for English. This means that if we let [[makeindex]] think that +the Greek words are actually strange English words, we can use it +to generate correct Greek indices. Let's be more concrete. + +Every line of an [[.idx]] file has the following format: +[[\indexnetry{word}{Page number}]], with the obvious meaning. +If [[word]] is some Greek word, then in order to allow correct sorting +we must sort index entries by using a transliteration that will +correspond<br> +á to a, â to b, ã to c, etc. Now, if we transform the above line +into [[\indexentry{trans. word@word}{Page number}]], it is possible +to get a correct index. So, the first thing we need is a routine that +will perform this transliteration. + +The following routine, [[g2e]], accepts one argument, a word, and returns +a transliteration of it. It can handle properly words writen with the Greek +alphabet and word writen using the babel Greek transliteration. The routine +is very simple: it splits the word into an array of characters, and +creates a new word by transliterating each character of the original word. + +Special care has been taken to correctly process the letter ``Capital Alpha +with accute'', since the ISO8859-7 and the Windows-1253 encodings have +reserved different slots for this character. + +<<Routine that performs transliteration>>= +sub g2e +{ + $word = $_[0]; + $tword = ""; + foreach $_ (split(//, $word)) + { + if (/^a|^A|^á|^Á|^Ü|^\xA2|^\xB6/) { $tword .= "a";} + elsif (/^b|^B|^â|^Â/) { $tword .= "b"; } + elsif (/^d|^D|^ä|^Ä/) { $tword .= "d"; } + elsif (/^e|^E|^å|^Å|^Ý|^¸/) { $tword .= "e"; } + elsif (/^i|^I|^é|^É|^ß|^º|^À/) { $tword .= "i"; } + elsif (/^o|^O|^ï|^Ï|^ü|^¼/) { $tword .= "o"; } + elsif (/^p|^P|^ð|^Ð/) { $tword .= "p"; } + elsif (/^g|^G|^ã|^Ã/) { $tword .= "c"; } + elsif (/^z|^Z|^æ|^Æ/) { $tword .= "f"; } + elsif (/^h|^H|^ç|^Ç|^Þ|^¹/) { $tword .= "g"; } + elsif (/^j|^J|^è|^È/) { $tword .= "h"; } + elsif (/^k|^K|^ê|^Ê/) { $tword .= "j"; } + elsif (/^l|^L|^ë|^Ë/) { $tword .= "k"; } + elsif (/^m|^M|^ì|^Ì/) { $tword .= "l"; } + elsif (/^n|^N|^í|^Í/) { $tword .= "m"; } + elsif (/^x|^X|^î|^Î/) { $tword .= "n"; } + elsif (/^r|^R|^ñ|^Ñ/) { $tword .= "q"; } + elsif (/^s|^c|^S|^ó|^ò|^Ó/) { $tword .= "r"; } + elsif (/^t|^T|^ô|^Ô/) { $tword .= "s"; } + elsif (/^u|^U|^õ|^Õ|^ý|^¾|^à/) { $tword .= "t"; } + elsif (/^f|^F|^ö|^Ö/) { $tword .= "u"; } + elsif (/^q|^Q|^÷|^×/) { $tword .= "v"; } + elsif (/^y|^Y|^ø|^Ø/) { $tword .= "w"; } + elsif (/^w|^W|^ù|^Ù|^þ|^¿/) { $tword .= "x"; } + elsif (/^'|^`|^~|^<|^>/) { } + else { $tword .= $_ } + } + return $tword; +} + +@ Now, that we have solved one problem we must face the others, namely the +modification of the [[.idx]] file and the subsequent modification of +the [[.ind]] file. (This of course means that we use [[makeindex]] to +actually generate the index.) Before we do anything we must get the +various command line arguments. Our program accepts at most three +command line arguments: [[mkgrkindex.pl [-s A|a] [-l] index]]. +With the [[-s]] switch one can specify which index style he wants to +be used: [[A]] stands for an index where each group of entries that start +with the same letter has as group-header this letter in uppercase form; +and [[a]] means that the group-headers are lowercase letter. +The absence of this switch denotes that the index will be just an ordinary +LaTeX index, i.e., without a group-header. The [[-l]] specifies that this +is an index that has only Latin words. The absence of this switch means +that the index is a Greek one. Finally, [[index]] denotes the name of the +[[.idx]] file; users may omit the file extension. + +The first think we do is to check whether there any command line arguments. +In case there aren't or there more than three, we just print a usage message +and exit. Otherwise, we print the version number and we process the command +line arguments. + +<<Check for command line arguments>>= +$argc = @ARGV; + +if ($argc == 0 || $argc > 4 ) # no command line arguments or more than 3 +{ # arguments + die "Usage: mkgrkindex.pl [-s A|a] [-l] index\n"; +} +else +{ + print "This is mkgrkindex (previously known as mkindex) version 2.0\n"; + <<Process command line arguments>> +} + +@ Since we don't know apriori the number of arguments we set the following +global variables: [[$style]] and [[$is_latin]]. For variable [[$style]] +a negative value means that a lowercase letter will precede each group +of words beginning with the same letter, a positive value means that +the letter will be uppercase and the value 0 means that there will be +no letter preceding each group of words. While for variable [[$is_latin]] +a positive value means that the index is non-Greek script index and a +negative or zero means it is a Greek script index. The last thing we +must do is to check whether the index file exists. + +<<Process command line arguments>>= + + $style = 0; + $is_latin = 0; + <<Get command line arguments>> + <<Check if .idx file exists>> + +@ In order to get the various command line arguments we use a simple +[[while]] loop that check each element of the array [[@ARGV]]. We check +for both switches and we get the name of the file that contains the index. + +<<Get command line arguments>>= + + SWITCHES: + while($_ = $ARGV[0]) + { + shift; + if (/^-s/) + { + if ($ARGV[0] eq "a") + { + $style = -1; + } + elsif ($ARGV[0] eq "A") + { + $style = 1; + } + else + { + die "$ARGV[0]: Illegal argument for switch -s\n"; + } + shift; + } + elsif (/^-l/) + { + $is_latin = 1; + } + elsif (/^-\w+/) + { + die "$_: Illegal command line switch!\n"; + } + else + { + $file = $_; + } + } + die "No index file name specified!\n" if $file eq ""; + +@ In order to check whether the index file exists, we simply use the +[[-e]] operator. First we check if [[$file]] exits, if doesn't then +probably file [[$file.idx]] exists, i.e., the user hasn't typed the +filename extension. In case this doesn't exist, then we are sure there is +no index file. After these checks, variable [[$file]] contains the file name +without extension. + +<<Check if .idx file exists>>= + if (! (-e $file)) + { + die "$file: no such file!\n" if $file =~ /.+\..+/; + die "$file.idx: no such file!\n" if (! (-e "$file.idx")); + $index_file = "$file.idx"; + } + else + { + $index_file = $file; + $file = $1 if $index_file =~ /(.+)\..+/; + } + +@ We have now all the information we need in order to modify the index file. +First we must rename the index file, since we are going to modify it. Next, +we open the renamed file for reading and we open a fresh file into which +we will write the modified index. These changes must be done only if this +is a Greek index, i.e., if the value of [[$is_latin]] is equal to zero. +The next think we do is to modify the index +entries. If an index entry appears in the [[\frontmatter]] of a LaTeX +document, then its page number comes out as a Latin number. Since we are +preparing a document with the [[greek]] option of the babel package this +number comes out as an argument of the command [[\textlatin]], e.g., +[[\textlatin{ix}]]. This means that we must modify the page information +too. However, this change requires to rescan the whole index file. So, +we delete the original index file, rename the modified one and then +correct the page numbers, by using the same method. + +<<Modify LaTeX index file>>= + +$old_file="$index_file.old"; + +if ($is_latin == 0) +{ + rename $index_file, $old_file; + open(OLD, "$old_file") || die "Can't open file $old_file\n"; + open(NEW, ">$index_file") || die "Can't open file $index_file\n"; + <<Modify index entries>> + close OLD; + close NEW; + unlink $old_file; +} + +rename $index_file, $old_file; +open(OLD, "$old_file") || die "Can't open file $old_file\n"; +open(NEW, ">$index_file") || die "Can't open file $index_file\n"; + +<<Correct page numbers>> + +close OLD; +close NEW; +unlink $old_file; + +@ If the value of variable [[$is_latin]] is positive there is nothing to +do. But, if it is zero, then we change each index line. Nikos Platis has +suggested a better method to parse individual lines. Since this method +is better, I have decided to use it. The method uses [[split]] to split +what goes inside the curly brackets into simpler parts. This function takes +two arguments, a regular expression and a string, and if the regular expression +matches somewhere in the string, the returns what goes before and after the +substring that matched. The method tries to first split the string at the +[[|]] symbol. Note that the expression [[(?<!")\x7C]] matches any [[|]] that is not +preceeded by a quotation mark (the original algorithm did not take care of +this detail). Then it splits what goes before [[|]] at each occurence of the +[[!]] symbol. Finally, each part of these substrings is splitted at the point where +the [[@]] symbol is found. The last thing is to make all necesary transformations +and to assemble the index entry. + +<<Modify index entries>>= + +while (<OLD>) +{ + #\x7B = {, \x7D = },\x40 = @, \x7C = |, \x21 = !, \x28 = (, \x29 = ) + chomp; + #Nikos Platis has suggested to replace the original code with a better + #parsing method. The code that follows is essentially Nikos's code with + #some minnor modifications, as explained in the documentation. + $newentry = ""; + /^\\indexentry\x7B(.+)\x7D(.+)/; + $fullentry = $1; + $page = $2; + ($indexentry, $format) = split(/(?<!")\x7C/, $fullentry); + @entryparts = split(/(?<!")\x21/, $indexentry); + $numparts = @entryparts; + for ($i = 0; $i < $numparts; $i++) { + ($x, $y) = split(/(?<!")\x40/, @entryparts[$i]); + if ($i > 0) { + $newentry .= "!" + } + $tx = g2e($x); + if ($y) { + $newentry .= "$tx\@$y" + } + else { + $newentry .= "$tx\@$x"; + } + } + print NEW "\\indexentry{$newentry"; + if ($format) { + print NEW "|$format" + } + print NEW "}$page\n" +} + +@ Correcting the page numbers is very easy. We check each line and +if one happens to have the form [[\indexentry{w}{\textlatin{p}}]], +then we replace it with [[\indexentry{w|textlatin}{p}]]. Of course +this does not take care of cases where the [[|]] operator is +already in use, but then we make the assumption that definitions and +such stuff do not appear in the foreword of a document. + +<<Correct page numbers>>= +while(<OLD>) +{ + chomp($_); + if (/\\indexentry\x7B(.+)\x7D\x7B\\textlatin\s*?\x7B(\w+)\x7D\x7D/) + { + print NEW "\\indexentry{$1|textlatin}{$2}\n"; + } + else + { + print NEW "$_\n"; + } +} + +@ The [[.ind]] file will be generated by invoking program [[makeindex]]. +In case the name of the program is different in the user's installation, +he/she must change the assignment at the beginning of the program. +As we said, the user can choose between three different types of indices. +Indices of type "A" and "a", assume that LaTeX and [[makeindex]] have access +to files [[Alpha.ist]] and [[alpha.ist]]. So, make sure these files are +properly installed. Depending on the value of variable [[$style]], +we generate the appropriate index. + +<<Generate .ind file>>= + +if ($style < 0) +{ + system("$makeindex -s alpha.ist $file"); +} +elsif ($style > 0) +{ + system("$makeindex -s Alpha.ist $file"); +} +else +{ + system("$makeindex $file"); +} + +@ Now the we have created the [[.ind]] file, if the user has chosen either +the [[A]] or [[a]] option we must modify the file so that the English +letters become Greek ones. The mechanism is very simple: we employ the +inverse of the mapping we employed in subroutine [[g2e]]. In order to +do this we use a hash table. In case the [[$is_latin == 1]], then the only +thing we need to do is to translate the word ``symbol'' into Greek. Before, +we do anything further we must open the file and to rename the old index +file. + +<<Modify the .ind file>>= + +<<Declare hash table>> +$ind_file = "$file.ind"; +$old_file="$ind_file.old"; + +if ($is_latin == 0) +{ + rename $ind_file, $old_file; + open(OLD, "$old_file") || die "Can't open file $old_file\n"; + open(NEW, ">$ind_file") || die "Can't open file $ind_file\n"; + <<Correct header letters>> + close OLD; + close NEW; + unlink $old_file; +} + +rename $ind_file, $old_file; +open(OLD, "$old_file") || die "Can't open file $old_file\n"; +open(NEW, ">$ind_file") || die "Can't open file $ind_file\n"; +<<Translate the word symbol>> +close OLD; +close NEW; +unlink $old_file; + +@ We declare a hash table in which we will store the inverse transliteration +table. + +<<Declare hash table>>= + +%e2g = ( + 'a' => 'á', 'A' => 'Á', + 'b' => 'â', 'B' => 'Â', + 'c' => 'ã', 'C' => 'Ã', + 'd' => 'ä', 'D' => 'Ä', + 'e' => 'å', 'E' => 'Å', + 'f' => 'æ', 'F' => 'Æ', + 'g' => 'ç', 'G' => 'Ç', + 'h' => 'è', 'H' => 'È', + 'i' => 'é', 'I' => 'É', + 'j' => 'ê', 'J' => 'Ê', + 'k' => 'ë', 'K' => 'Ë', + 'l' => 'ì', 'L' => 'Ì', + 'm' => 'í', 'M' => 'Í', + 'n' => 'î', 'N' => 'Î', + 'o' => 'ï', 'O' => 'Ï', + 'p' => 'ð', 'P' => 'Ð', + 'q' => 'ñ', 'Q' => 'Ñ', + 'r' => 'óv', 'R' => 'Ó', + 's' => 'ô', 'S' => 'Ô', + 't' => 'õ', 'T' => 'Õ', + 'u' => 'ö', 'U' => 'Ö', + 'v' => '÷', 'V' => '×', + 'w' => 'ø', 'W' => 'Ø', + 'x' => 'ù', 'X' => 'Ù' + ); + +@ In order to translate the word symbol we must just scan the entire +[[.ind]] file and to find the word [[symbols]] or [[Symbols]]. This +isn't difficult since each header line is of the form +[[{\hfil word \hfil}]], where [[word]] is either the a letter, the word +[[symbols]] or the word [[Symbols]]. Once we +find the word we replace it with its translation. + +<<Translate the word symbol>>= + +while (<OLD>) +{ + if (/^\x7B\\hfil (\w)ymbols \\hfil\x7D/) + { + if ($1 eq "s") + { + print NEW "{\\hfil \\textgreek{\\textbf{s'umbola}}"; + print NEW "\\hfil}\\nopagebreak\n"; + } + elsif ($1 eq "S") + { + print NEW "{\\hfil \\textgreek{\\textbf{S'umbola}}"; + print NEW "\\hfil}\\nopagebreak\n"; + } + else + { + die "Illegal header $1 in .ind file\n"; + } + } + else + { + print NEW; + } +} + +@ Correcting the header letters means to scan the whole index file and +replace the English letters with Greek ones. We find lines of the form +[[{\hfil a \hfil}\nopagebreak]] and we replace the English letter +with the corresponding Greek by using the hash [[%e2g]]. + +<<Correct header letters>>= + +while (<OLD>) +{ + if (/^{\\hfil (\w?) \\hfil}/) + { + $lettergr = $e2g{$1}; + print NEW "{\\hfil $lettergr \\hfil}\\nopagebreak\n"; + } + else + { + print NEW ; + } +} + +@ Let's summarize. This program is makes it possible to use +program [[makeindex]] as if this program was writen for Greek +people only! The program works as follows: (a) it check the command line +arguments, (b) it modifies the LaTeX generated index, (c) generates the +[[.ind]] according to the user's wishes, and (d) corrects the [[.ind]] file +if necessary, i.e., if the user has asked for some alphabetic headers +in his/her index. + +<<*>>= +#!/usr/bin/perl +# +#(c) Copyright 1998-2009 Apostolos Syropoulos +# asyropoulos@yahoo.com +# +# The LaTeX Project Public License (lppl) +# This software is copyright but you are granted a license which gives you, +# the "user" of the software, legal permission to copy, distribute, and/or +# modify the software. However, if you modify the software and then distribute +# it (even just locally) you must change the name of the software, or use other +# technical means to avoid confusion. +# +$makeindex = "makeindex"; #name of the index generation utility +<<Routine that performs transliteration>> +<<Check for command line arguments>> +<<Modify LaTeX index file>> +<<Generate .ind file>> +if ($style != 0) +{ + <<Modify the .ind file>> +} +__END__ diff --git a/Master/texmf-dist/doc/support/mkgrkindex/mkgrkindex.pdf b/Master/texmf-dist/doc/support/mkgrkindex/mkgrkindex.pdf Binary files differnew file mode 100644 index 00000000000..ee91a5b3c73 --- /dev/null +++ b/Master/texmf-dist/doc/support/mkgrkindex/mkgrkindex.pdf diff --git a/Master/texmf-dist/makeindex/mkgrkindex/Alpha.ist b/Master/texmf-dist/makeindex/mkgrkindex/Alpha.ist new file mode 100644 index 00000000000..0fac84f8a50 --- /dev/null +++ b/Master/texmf-dist/makeindex/mkgrkindex/Alpha.ist @@ -0,0 +1,4 @@ +heading_prefix "{\\hfil " +heading_suffix " \\hfil}\\nopagebreak\n" + +headings_flag 1 diff --git a/Master/texmf-dist/makeindex/mkgrkindex/alpha.ist b/Master/texmf-dist/makeindex/mkgrkindex/alpha.ist new file mode 100644 index 00000000000..c36a6fc262f --- /dev/null +++ b/Master/texmf-dist/makeindex/mkgrkindex/alpha.ist @@ -0,0 +1,4 @@ +heading_prefix "{\\hfil " +heading_suffix " \\hfil}\\nopagebreak\n" + +headings_flag -1 diff --git a/Master/texmf-dist/scripts/mkgrkindex/mkgrkindex b/Master/texmf-dist/scripts/mkgrkindex/mkgrkindex new file mode 100755 index 00000000000..a4d30438f1b --- /dev/null +++ b/Master/texmf-dist/scripts/mkgrkindex/mkgrkindex @@ -0,0 +1,289 @@ +#!/usr/bin/env perl +# +#(c) Copyright 1998-2009 Apostolos Syropoulos +# asyropoulos@yahoo.com +# +# The LaTeX Project Public License (lppl) +# This software is copyright but you are granted a license which gives you, +# the "user" of the software, legal permission to copy, distribute, and/or +# modify the software. However, if you modify the software and then distribute +# it (even just locally) you must change the name of the software, or use other +# technical means to avoid confusion. +# +$makeindex = "makeindex"; #name of the index generation utility +sub g2e +{ + $word = $_[0]; + $tword = ""; + foreach $_ (split(//, $word)) + { + if (/^a|^A|^á|^Á|^Ü|^\xA2|^\xB6/) { $tword .= "a";} + elsif (/^b|^B|^â|^Â/) { $tword .= "b"; } + elsif (/^d|^D|^ä|^Ä/) { $tword .= "d"; } + elsif (/^e|^E|^å|^Å|^Ý|^¸/) { $tword .= "e"; } + elsif (/^i|^I|^é|^É|^ß|^º|^À/) { $tword .= "i"; } + elsif (/^o|^O|^ï|^Ï|^ü|^¼/) { $tword .= "o"; } + elsif (/^p|^P|^ð|^Ð/) { $tword .= "p"; } + elsif (/^g|^G|^ã|^Ã/) { $tword .= "c"; } + elsif (/^z|^Z|^æ|^Æ/) { $tword .= "f"; } + elsif (/^h|^H|^ç|^Ç|^Þ|^¹/) { $tword .= "g"; } + elsif (/^j|^J|^è|^È/) { $tword .= "h"; } + elsif (/^k|^K|^ê|^Ê/) { $tword .= "j"; } + elsif (/^l|^L|^ë|^Ë/) { $tword .= "k"; } + elsif (/^m|^M|^ì|^Ì/) { $tword .= "l"; } + elsif (/^n|^N|^í|^Í/) { $tword .= "m"; } + elsif (/^x|^X|^î|^Î/) { $tword .= "n"; } + elsif (/^r|^R|^ñ|^Ñ/) { $tword .= "q"; } + elsif (/^s|^c|^S|^ó|^ò|^Ó/) { $tword .= "r"; } + elsif (/^t|^T|^ô|^Ô/) { $tword .= "s"; } + elsif (/^u|^U|^õ|^Õ|^ý|^¾|^à/) { $tword .= "t"; } + elsif (/^f|^F|^ö|^Ö/) { $tword .= "u"; } + elsif (/^q|^Q|^÷|^×/) { $tword .= "v"; } + elsif (/^y|^Y|^ø|^Ø/) { $tword .= "w"; } + elsif (/^w|^W|^ù|^Ù|^þ|^¿/) { $tword .= "x"; } + elsif (/^'|^`|^~|^<|^>/) { } + else { $tword .= $_ } + } + return $tword; +} + +$argc = @ARGV; + +if ($argc == 0 || $argc > 4 ) # no command line arguments or more than 3 +{ # arguments + die "Usage: mkgrkindex.pl [-s A|a] [-l] index\n"; +} +else +{ + print "This is mkgrkindex (previously known as mkindex) version 2.0\n"; + + $style = 0; + $is_latin = 0; + + SWITCHES: + while($_ = $ARGV[0]) + { + shift; + if (/^-s/) + { + if ($ARGV[0] eq "a") + { + $style = -1; + } + elsif ($ARGV[0] eq "A") + { + $style = 1; + } + else + { + die "$ARGV[0]: Illegal argument for switch -s\n"; + } + shift; + } + elsif (/^-l/) + { + $is_latin = 1; + } + elsif (/^-\w+/) + { + die "$_: Illegal command line switch!\n"; + } + else + { + $file = $_; + } + } + die "No index file name specified!\n" if $file eq ""; + + if (! (-e $file)) + { + die "$file: no such file!\n" if $file =~ /.+\..+/; + die "$file.idx: no such file!\n" if (! (-e "$file.idx")); + $index_file = "$file.idx"; + } + else + { + $index_file = $file; + $file = $1 if $index_file =~ /(.+)\..+/; + } + + +} + + +$old_file="$index_file.old"; + +if ($is_latin == 0) +{ + rename $index_file, $old_file; + open(OLD, "$old_file") || die "Can't open file $old_file\n"; + open(NEW, ">$index_file") || die "Can't open file $index_file\n"; + + while (<OLD>) + { + #\x7B = {, \x7D = },\x40 = @, \x7C = |, \x21 = !, \x28 = (, \x29 = ) + chomp; + #Nikos Platis has suggested to replace the original code with a better + #parsing method. The code that follows is essentially Nikos's code with + #some minnor modifications, as explained in the documentation. + $newentry = ""; + /^\\indexentry\x7B(.+)\x7D(.+)/; + $fullentry = $1; + $page = $2; + ($indexentry, $format) = split(/(?<!")\x7C/, $fullentry); + @entryparts = split(/(?<!")\x21/, $indexentry); + $numparts = @entryparts; + for ($i = 0; $i < $numparts; $i++) { + ($x, $y) = split(/(?<!")\x40/, @entryparts[$i]); + if ($i > 0) { + $newentry .= "!" + } + $tx = g2e($x); + if ($y) { + $newentry .= "$tx\@$y" + } + else { + $newentry .= "$tx\@$x"; + } + } + print NEW "\\indexentry{$newentry"; + if ($format) { + print NEW "|$format" + } + print NEW "}$page\n" + } + + close OLD; + close NEW; + unlink $old_file; +} + +rename $index_file, $old_file; +open(OLD, "$old_file") || die "Can't open file $old_file\n"; +open(NEW, ">$index_file") || die "Can't open file $index_file\n"; + +while(<OLD>) +{ + chomp($_); + if (/\\indexentry\x7B(.+)\x7D\x7B\\textlatin\s*?\x7B(\w+)\x7D\x7D/) + { + print NEW "\\indexentry{$1|textlatin}{$2}\n"; + } + else + { + print NEW "$_\n"; + } +} + + +close OLD; +close NEW; +unlink $old_file; + + +if ($style < 0) +{ + system("$makeindex -s alpha.ist $file"); +} +elsif ($style > 0) +{ + system("$makeindex -s Alpha.ist $file"); +} +else +{ + system("$makeindex $file"); +} + +if ($style != 0) +{ + + + %e2g = ( + 'a' => 'á', 'A' => 'Á', + 'b' => 'â', 'B' => 'Â', + 'c' => 'ã', 'C' => 'Ã', + 'd' => 'ä', 'D' => 'Ä', + 'e' => 'å', 'E' => 'Å', + 'f' => 'æ', 'F' => 'Æ', + 'g' => 'ç', 'G' => 'Ç', + 'h' => 'è', 'H' => 'È', + 'i' => 'é', 'I' => 'É', + 'j' => 'ê', 'J' => 'Ê', + 'k' => 'ë', 'K' => 'Ë', + 'l' => 'ì', 'L' => 'Ì', + 'm' => 'í', 'M' => 'Í', + 'n' => 'î', 'N' => 'Î', + 'o' => 'ï', 'O' => 'Ï', + 'p' => 'ð', 'P' => 'Ð', + 'q' => 'ñ', 'Q' => 'Ñ', + 'r' => 'óv', 'R' => 'Ó', + 's' => 'ô', 'S' => 'Ô', + 't' => 'õ', 'T' => 'Õ', + 'u' => 'ö', 'U' => 'Ö', + 'v' => '÷', 'V' => '×', + 'w' => 'ø', 'W' => 'Ø', + 'x' => 'ù', 'X' => 'Ù' + ); + + $ind_file = "$file.ind"; + $old_file="$ind_file.old"; + + if ($is_latin == 0) + { + rename $ind_file, $old_file; + open(OLD, "$old_file") || die "Can't open file $old_file\n"; + open(NEW, ">$ind_file") || die "Can't open file $ind_file\n"; + + while (<OLD>) + { + if (/^{\\hfil (\w?) \\hfil}/) + { + $lettergr = $e2g{$1}; + print NEW "{\\hfil $lettergr \\hfil}\\nopagebreak\n"; + } + else + { + print NEW ; + } + } + + close OLD; + close NEW; + unlink $old_file; + } + + rename $ind_file, $old_file; + open(OLD, "$old_file") || die "Can't open file $old_file\n"; + open(NEW, ">$ind_file") || die "Can't open file $ind_file\n"; + + while (<OLD>) + { + if (/^\x7B\\hfil (\w)ymbols \\hfil\x7D/) + { + if ($1 eq "s") + { + print NEW "{\\hfil \\textgreek{\\textbf{s'umbola}}"; + print NEW "\\hfil}\\nopagebreak\n"; + } + elsif ($1 eq "S") + { + print NEW "{\\hfil \\textgreek{\\textbf{S'umbola}}"; + print NEW "\\hfil}\\nopagebreak\n"; + } + else + { + die "Illegal header $1 in .ind file\n"; + } + } + else + { + print NEW; + } + } + + close OLD; + close NEW; + unlink $old_file; + +} +__END__ diff --git a/Master/tlpkg/bin/tlpkg-ctan-check b/Master/tlpkg/bin/tlpkg-ctan-check index c5b858e70ba..23027c41e98 100755 --- a/Master/tlpkg/bin/tlpkg-ctan-check +++ b/Master/tlpkg/bin/tlpkg-ctan-check @@ -46,7 +46,7 @@ my @WorkingTLP = qw( boxhandler bpchem breakurl bullcntr bussproofs bytefield calctab calligra calrsfs captcont caption carlisle casyl catechis carolmin-ps - cbcoptic cbfonts ccaption cclicenses cd cdpbundl cell cellspace + cbcoptic cbfonts cc-pl ccaption cclicenses cd cdpbundl cell cellspace changebar changelayout changepage changes chappg chapterfolder chem-journal chembst chemcompounds chemstyle cherokee chess-problem-diagrams chessboard chessfss @@ -164,7 +164,7 @@ my @WorkingTLP = qw( mff mfnfss mfpic mfpic4ode mftinc mh mhchem mhequ mhs microtype midnight minibox minipage-marginpar minitoc minutes - mkjobtexmf mkpattern + mkgrkindex mkjobtexmf mkpattern mla-paper mlist mmap mnsymbol moderncv modref modroman mongolian-babel montex morefloats moreverb movie15 mp3d mparhack diff --git a/Master/tlpkg/bin/tlpkginfo b/Master/tlpkg/bin/tlpkginfo index 7b9718a422c..49598411774 100755 --- a/Master/tlpkg/bin/tlpkginfo +++ b/Master/tlpkg/bin/tlpkginfo @@ -55,22 +55,28 @@ sub main # sub find_ctan_dir { - my ($me) = @_; - return 1 unless $me; + my ($pkgname) = @_; + return 1 unless $pkgname; # use explicit catalogue name if given in tlpsrc. - $me = &tlpsrc_find_catalogue ($me) || $me; - - # ctan (and latex-tds) bundles all ams packages together, - # but we use separate tlp names. - $me = "amslatex" if $me =~ /^($amslatex_tds_pkgs)$/; + my $me = &tlpsrc_find_catalogue ($pkgname); + if (! $me) { + if ($pkgname =~ /^($amslatex_tds_pkgs)$/) { + # ctan (and latex-tds) bundles all ams packages together, + # but we use separate tlp names. + $me = "amslatex" ; + } elsif ($pkgname =~ /^($cjk_pkgs)$/) { + $me = "cjk"; # likewise CJK + } elsif ($pkgname eq "mkgrkindex") { + $me = "greek-makeindex"; # use directory name + } else { + $me = $pkgname; # usually the same + } + } - # Likewise CJK. - $me = "cjk" if $me =~ /^($cjk_pkgs)$/; - # The CTAN path to a package is sometimes stored in the TeX Catalogue, # but old entries don't have it, etc. Still, we want to use it if present. - my $ctan_dir = &catalogue_find_ctan_path ($me) unless $me eq "cc-pl"; + my $ctan_dir = &catalogue_find_ctan_path ($me); if (! $ctan_dir) { # fall back on many special cases (my $menobin = $me) =~ s,^bin-,,; diff --git a/Master/tlpkg/libexec/ctan2tds b/Master/tlpkg/libexec/ctan2tds index 4ba39ef60e2..b58b31fd3d5 100755 --- a/Master/tlpkg/libexec/ctan2tds +++ b/Master/tlpkg/libexec/ctan2tds @@ -1065,6 +1065,7 @@ $standardtex='\.(cfg|sty|clo|ldf|cls|def|fd|cmap|4ht)$'; 'metauml', 'metapost', 'mfpic', 'generic', 'midnight', 'generic', + 'mkgrkindex', 'support', 'mkpattern', 'plain', 'mkjobtexmf', 'generic', 'mpattern', 'metapost', @@ -1467,6 +1468,7 @@ $standardxmt='\.xmt'; 'fragmaster' => 'fragmaster.pl', 'latex2man' => 'latex2man$', 'latexmk' => '\.pl$', + 'mkgrkindex' => 'mkgrkindex$', 'mkjobtexmf' => '\.pl$', 'pdfcrop' => '\.pl$', 'perltex' => '\.pl$', diff --git a/Master/tlpkg/tlpsrc/collection-langgreek.tlpsrc b/Master/tlpkg/tlpsrc/collection-langgreek.tlpsrc index 6a0deed58ae..fe0a5948a53 100644 --- a/Master/tlpkg/tlpsrc/collection-langgreek.tlpsrc +++ b/Master/tlpkg/tlpsrc/collection-langgreek.tlpsrc @@ -16,6 +16,7 @@ depend kdgreek depend kerkis depend levy depend lgreek +depend mkgrkindex depend teubner depend xgreek depend yannisgr diff --git a/Master/tlpkg/tlpsrc/mkgrkindex.tlpsrc b/Master/tlpkg/tlpsrc/mkgrkindex.tlpsrc new file mode 100644 index 00000000000..e69de29bb2d --- /dev/null +++ b/Master/tlpkg/tlpsrc/mkgrkindex.tlpsrc |