1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # LaTeXPlug.pm
|
---|
4 | #
|
---|
5 | # A component of the Greenstone digital library software
|
---|
6 | # from the New Zealand Digital Library Project at the
|
---|
7 | # University of Waikato, New Zealand.
|
---|
8 | #
|
---|
9 | # Written by John McPherson
|
---|
10 | # Copyright (C) 2004 New Zealand Digital Library Project
|
---|
11 | #
|
---|
12 | # This program is free software; you can redistribute it and/or modify
|
---|
13 | # it under the terms of the GNU General Public License as published by
|
---|
14 | # the Free Software Foundation; either version 2 of the License, or
|
---|
15 | # (at your option) any later version.
|
---|
16 | #
|
---|
17 | # This program is distributed in the hope that it will be useful,
|
---|
18 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
20 | # GNU General Public License for more details.
|
---|
21 | #
|
---|
22 | ###########################################################################
|
---|
23 |
|
---|
24 | # todo:
|
---|
25 | # \includegraphics
|
---|
26 | # parse/remove tex \if ... macros
|
---|
27 |
|
---|
28 | package LaTeXPlug;
|
---|
29 |
|
---|
30 | use strict;
|
---|
31 | no strict 'refs'; # so we can print to a handle named by a variable
|
---|
32 |
|
---|
33 | # greenstone packages
|
---|
34 | use BasPlug;
|
---|
35 | use parsargv;
|
---|
36 | use unicode;
|
---|
37 | use util;
|
---|
38 |
|
---|
39 | sub BEGIN {
|
---|
40 | @LaTeXPlug::ISA = ('BasPlug');
|
---|
41 | }
|
---|
42 |
|
---|
43 | sub print_usage {
|
---|
44 | print STDERR "\n usage: plugin LaTeXPlug [options]\n\n";
|
---|
45 | }
|
---|
46 |
|
---|
47 | sub new {
|
---|
48 | my $class = shift (@_);
|
---|
49 | my $self = new BasPlug ($class, @_);
|
---|
50 | $self->{'plugin_type'} = 'LaTeXPlug';
|
---|
51 | my $option_list = $self->{'option_list'};
|
---|
52 | my $options={ 'name' => 'LaTeXPlug',
|
---|
53 | 'desc' => '{LaTeXPlug.desc}',
|
---|
54 | 'abstract' => 'no',
|
---|
55 | 'inherits' => 'yes',
|
---|
56 | 'args' => [
|
---|
57 | ] # no arguments for now...
|
---|
58 | };
|
---|
59 |
|
---|
60 | push( @{$option_list}, $options );
|
---|
61 |
|
---|
62 | if (!parsargv::parse(\@_,
|
---|
63 | "allow_extra_options")) {
|
---|
64 |
|
---|
65 | print STDERR "\nIncorrect options passed to LaTeXPlug, check your collect.cfg configuration file\n";
|
---|
66 | $self->print_txt_usage(""); # don't specify which language bundle
|
---|
67 | die "\n";
|
---|
68 | }
|
---|
69 |
|
---|
70 | $self->{'aux_files'} = {};
|
---|
71 | $self->{'dir_num'} = 0;
|
---|
72 | $self->{'file_num'} = 0;
|
---|
73 | return bless $self, $class;
|
---|
74 | }
|
---|
75 |
|
---|
76 |
|
---|
77 | sub get_default_process_exp {
|
---|
78 | my $self = shift (@_);
|
---|
79 | return q^\.tex$^;
|
---|
80 | }
|
---|
81 |
|
---|
82 | sub get_default_block_exp {
|
---|
83 | # assume any .eps files are part of the latex stuff
|
---|
84 | return '\.(?:eps)$';
|
---|
85 | }
|
---|
86 |
|
---|
87 |
|
---|
88 | sub process {
|
---|
89 | my $self = shift (@_);
|
---|
90 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
|
---|
91 |
|
---|
92 | my $start=substr($$textref, 0, 200); # first 200 bytes
|
---|
93 |
|
---|
94 | if ($start !~ m~\\ (?:documentclass | documentstyle | input | section
|
---|
95 | | chapter | contents | begin) ~x) {
|
---|
96 | # this doesn't look like latex...
|
---|
97 | return undef;
|
---|
98 | }
|
---|
99 | my $outhandle = $self->{'outhandle'};
|
---|
100 | if ($gli) {
|
---|
101 | print STDERR "<Processing n='$file' p='LaTeXPlug'>\n";
|
---|
102 | } elsif ($self->{'verbosity'} > 1) {
|
---|
103 | print $outhandle "LaTeXPlug: processing $file\n"
|
---|
104 | }
|
---|
105 | my $cursection = $doc_obj->get_top_section();
|
---|
106 |
|
---|
107 | ###### clean up text ######
|
---|
108 | $$textref =~ s/\r$//mg; # remove dos ^M
|
---|
109 | $$textref =~ s/%.*$//mg; # remove comments
|
---|
110 |
|
---|
111 | # convert to utf-8 if not already - assume non ascii => iso-8859-1/latin
|
---|
112 |
|
---|
113 | $$textref =~ s@(?<=[[:ascii:]])\xA0+@\xc2\xa0@g; # latin nonbreaking space
|
---|
114 | # check that both sides are ascii, so we don't screw up utf-8 chars
|
---|
115 | $$textref =~ s@ (?<=[[:ascii:]])([\x80-\xff])(?=[[:ascii:]]) @
|
---|
116 | unicode::ascii2utf8($1) @egx; # takes "extended ascii" (ie latin)
|
---|
117 |
|
---|
118 |
|
---|
119 | ###### find metadata ######
|
---|
120 |
|
---|
121 | ## FileFormat metadata ##
|
---|
122 | $doc_obj->add_metadata($cursection, "FileFormat", "LaTeX");
|
---|
123 |
|
---|
124 | ### title metadata ###
|
---|
125 | $$textref =~ m@\\title\s*{(.*?)}@s;
|
---|
126 | my $title = $1;
|
---|
127 | if (!$title) {
|
---|
128 | # no title tag. look for a chapter/section heading
|
---|
129 | $$textref =~ m@\\(?:chapter|section)\s*{(.*?)}@s; # will get 1st match
|
---|
130 | $title = $1;
|
---|
131 | }
|
---|
132 | if (!$title) {
|
---|
133 | # no chapter/section heading tags either... use filename
|
---|
134 | $title = $file;
|
---|
135 | $title =~ s/\.tex$//i;
|
---|
136 | $title =~ s/[-_.]/ /g; # turn punctuation into spaces
|
---|
137 | }
|
---|
138 | if ($title) {
|
---|
139 | $title =~ s@\\\\@ @g; # embedded newlines
|
---|
140 | $title = $self->process_latex($title); # no "-html" for title eg in browser
|
---|
141 | $doc_obj->add_utf8_metadata($cursection, "Title", $title);
|
---|
142 | }
|
---|
143 |
|
---|
144 | ### creator/individual author metadata ###
|
---|
145 | $$textref =~ m@\\author\s*{((?:{.*?}|.*?)+)}\s*$@sm;
|
---|
146 | my $authors=$1;
|
---|
147 | if ($authors) {
|
---|
148 | # take care of "et al."...
|
---|
149 | $authors =~ s/(\s+et\.?\s+al\.?)\s*$//;
|
---|
150 | my $etal=$1;
|
---|
151 | $etal="" if (!defined ($etal));
|
---|
152 |
|
---|
153 | my @authorlist=parse_authors($self, $authors);
|
---|
154 |
|
---|
155 | foreach my $author (@authorlist) {
|
---|
156 | # Add each name to set of Authors
|
---|
157 | $doc_obj->add_utf8_metadata ($cursection, "Author", $author);
|
---|
158 | }
|
---|
159 |
|
---|
160 | # Only want at most one "and" in the Creator field
|
---|
161 | my $creator_str="";
|
---|
162 | if (scalar(@authorlist) > 2) {
|
---|
163 | my $lastauthor=pop @authorlist;
|
---|
164 | $creator_str=join(', ', @authorlist);
|
---|
165 | $creator_str.=" and $lastauthor";
|
---|
166 | } else { # 1 or 2 authors...
|
---|
167 | $creator_str=join(" and ",@authorlist);
|
---|
168 | }
|
---|
169 | $creator_str.=$etal; # if there was "et al."
|
---|
170 | $doc_obj->add_utf8_metadata($cursection, "Creator", $creator_str);
|
---|
171 | }
|
---|
172 | ### end of author metadata ###
|
---|
173 |
|
---|
174 | ###### process latex for the main text ######
|
---|
175 | $$textref =~ s/^.*?\\begin{document}//s;
|
---|
176 | $$textref =~ s/\\end{document}.*?$//s;
|
---|
177 | $$textref = $self->process_latex("-html",$$textref);
|
---|
178 | $doc_obj->add_utf8_text($cursection, $$textref);
|
---|
179 |
|
---|
180 | return 1;
|
---|
181 | }
|
---|
182 |
|
---|
183 |
|
---|
184 | # returns a list of author names
|
---|
185 | sub parse_authors {
|
---|
186 | my $self=shift;
|
---|
187 | my $authors=shift;
|
---|
188 |
|
---|
189 | my $outhandle=$self->{'outhandle'};
|
---|
190 |
|
---|
191 | $authors =~ s/\n/ /g; # remove newlines
|
---|
192 |
|
---|
193 | # some people do this for affiliation footnote/dagger
|
---|
194 | $authors =~ s@\$.*?\$@@g; # remove maths from author :(
|
---|
195 |
|
---|
196 | # und here for german language...
|
---|
197 | # don't use brackets in pattern, else the matched bit becomes
|
---|
198 | # an element in the list!
|
---|
199 | my @authorlist = split(/\s+and\s+|\s+und\s+/, $authors);
|
---|
200 | my @formattedlist = ();
|
---|
201 | foreach my $author (@authorlist) {
|
---|
202 | $author =~ s/\s*$//;
|
---|
203 | $author =~ s/^\s*//;
|
---|
204 | # Reformat and add author name
|
---|
205 | next if $author=~ /^\s*$/;
|
---|
206 |
|
---|
207 | # names are "First von Last", "von Last, First"
|
---|
208 | # or "von Last, Jr, First". See the "BibTeXing" manual, page 16
|
---|
209 | my $first="";
|
---|
210 | my $vonlast="";
|
---|
211 | my $jr="";
|
---|
212 |
|
---|
213 | if ($author =~ /,/) {
|
---|
214 | my @parts=split(/,\s*/, $author);
|
---|
215 | $first = pop @parts;
|
---|
216 | if (scalar(@parts) == 2) {
|
---|
217 | $jr = pop @parts;
|
---|
218 | }
|
---|
219 | $vonlast=shift @parts;
|
---|
220 | if (scalar(@parts) > 0) {
|
---|
221 | print $outhandle $self->{'plugin_type'} .
|
---|
222 | ": couldn't parse name $author\n";
|
---|
223 | # but we continue anyway...
|
---|
224 | }
|
---|
225 | } else { # First von Last
|
---|
226 | my @words = split(/ /, $author);
|
---|
227 | while (scalar(@words) > 1 && $words[0] !~ /^[a-z]{2..}/) {
|
---|
228 | $first .= " " . shift (@words);
|
---|
229 | }
|
---|
230 | $first =~ s/^\s//;
|
---|
231 | $vonlast = join (' ', @words); # whatever's left...
|
---|
232 | }
|
---|
233 | my $von="";
|
---|
234 | my $last="";
|
---|
235 | if ($vonlast =~ m/^[a-z]/) { # lowercase implies "von"
|
---|
236 | $vonlast =~ s/^(([a-z]\w+\s+)+)//;
|
---|
237 | $von = $1;
|
---|
238 | if (!defined ($von)) {
|
---|
239 | # some non-English names do start with lowercase
|
---|
240 | # eg "Marie desJardins". Also we can get typos...
|
---|
241 | print $outhandle "BibTexPlug: couldn't parse surname $vonlast\n";
|
---|
242 | $von="";
|
---|
243 | if ($vonlast =~ /^[a-z]+$/) {
|
---|
244 | # if it's all lowercase, uppercase 1st.
|
---|
245 | $vonlast =~ s/^(.)/\u$1/;
|
---|
246 | }
|
---|
247 | }
|
---|
248 | $von =~ s/\s*$//;
|
---|
249 | $last=$vonlast;
|
---|
250 | } else {
|
---|
251 | $last=$vonlast;
|
---|
252 | }
|
---|
253 | my $wholename="$first $von $last $jr";
|
---|
254 | $wholename =~ s/ $//; $wholename =~ s/\s+/ /g;
|
---|
255 | # my $fullname = "$last";
|
---|
256 | # $fullname .= " $jr" if ($jr);
|
---|
257 | # $fullname .= ", $first";
|
---|
258 | # $fullname .= " $von" if ($von);
|
---|
259 | push (@formattedlist, $wholename);
|
---|
260 | }
|
---|
261 | return @formattedlist;
|
---|
262 | }
|
---|
263 |
|
---|
264 |
|
---|
265 | ## following functions based on bibtex plugin ##
|
---|
266 | # not actually used at the moment, but might be useful in future?
|
---|
267 | sub expand_month {
|
---|
268 | my $text=shift;
|
---|
269 |
|
---|
270 | # bibtex style files expand abbreviations for months.
|
---|
271 | # Entries can contain more than one month (eg ' month = jun # "-" # aug, ')
|
---|
272 | $text =~ s/jan/_textmonth01_/g;
|
---|
273 | $text =~ s/feb/_textmonth02_/g;
|
---|
274 | $text =~ s/mar/_textmonth03_/g;
|
---|
275 | $text =~ s/apr/_textmonth04_/g;
|
---|
276 | $text =~ s/may/_textmonth05_/g;
|
---|
277 | $text =~ s/jun/_textmonth06_/g;
|
---|
278 | $text =~ s/jul/_textmonth07_/g;
|
---|
279 | $text =~ s/aug/_textmonth08_/g;
|
---|
280 | $text =~ s/sep/_textmonth09_/g;
|
---|
281 | $text =~ s/oct/_textmonth10_/g;
|
---|
282 | $text =~ s/nov/_textmonth11_/g;
|
---|
283 | $text =~ s/dec/_textmonth12_/g;
|
---|
284 |
|
---|
285 | return $text;
|
---|
286 | }
|
---|
287 |
|
---|
288 |
|
---|
289 | # If you want basic html formatting (eg \emph -> <em>, \bf, etc), give "-html"
|
---|
290 | # as the first argument to this function.
|
---|
291 | #
|
---|
292 | # Convert accented characters, remove { }, interprete some commands....
|
---|
293 | # Note!! This is not comprehensive! Also assumes Latin -> Unicode!
|
---|
294 |
|
---|
295 | # Also, it sucks quite a bit for complicated/nested commands since it doesn't
|
---|
296 | # match { with the corresponding }, only the nearest }
|
---|
297 |
|
---|
298 | sub process_latex {
|
---|
299 | my $self=shift;
|
---|
300 | my $text=shift;
|
---|
301 |
|
---|
302 | my $outhandle=$self->{'outhandle'};
|
---|
303 |
|
---|
304 | my $html_markup=0;
|
---|
305 | if ($text =~ /^\-html/) {
|
---|
306 | $html_markup=1;
|
---|
307 | $text=shift;
|
---|
308 | }
|
---|
309 |
|
---|
310 | if (! $text) {
|
---|
311 | return $text;
|
---|
312 | }
|
---|
313 | # escape html-sensitive characters
|
---|
314 | $text =~ s@&@&@g;
|
---|
315 | $text =~ s@<@<@g;
|
---|
316 | $text =~ s@>@>@g;
|
---|
317 |
|
---|
318 | # do this before accents, since \= means something different in tabbing
|
---|
319 | # also \> is a tab stop too, and \\ is newline
|
---|
320 | sub do_tabbing {
|
---|
321 | my $tabbing=shift;
|
---|
322 | $tabbing =~ s!^.*\\kill\s*$!!mg; # \kill sets tab stops, kills line
|
---|
323 | $tabbing =~ s~\\(?:=|>)~\xc2\xa0~g; # replace with nbsp
|
---|
324 | $tabbing =~ s~[\\][\\](?:\[.*?\])?\s*$~<br/>~mg;
|
---|
325 | return "<br/>" . $tabbing . "<br/>\n";
|
---|
326 | }
|
---|
327 | $text =~ s@\\begin{tabbing}(.*?)\\end{tabbing}@do_tabbing($1)@ges;
|
---|
328 | sub do_tabular {
|
---|
329 | my $tabular=shift;
|
---|
330 | $tabular =~ s~(?<!\\)\s*&\s*~</td><td>~g;
|
---|
331 | $tabular =~ s~[\\][\\]\s*~</td></tr>\n <tr><td>~g;
|
---|
332 | $tabular =~ s~\\hline~~g; # for now...
|
---|
333 | $tabular =~ s~<td>\s*\\multicolumn{(\d+)}{.*?}~<td colspan="$1">~g;
|
---|
334 | return "<table border=\"1\">\n <tr><td>"
|
---|
335 | . $tabular . "</td></tr></table>\n";
|
---|
336 | }
|
---|
337 | $text =~ s@\\begin{tabular}(?:\[.*?\])?{.*?}(.*?)\\end{tabular} @
|
---|
338 | do_tabular($1) @xges;
|
---|
339 |
|
---|
340 | $text =~ s@[\\][\\]\s*\n@ @g; # fold lines ending with \\
|
---|
341 |
|
---|
342 | # process maths mode before accents... things like \, mean different!
|
---|
343 | # maths mode
|
---|
344 | $text =~ s@\$\$(.*?)\$\$
|
---|
345 | @ process_latex_math($html_markup,$1)
|
---|
346 | @xsge; # multi-line maths: $$ .... $$
|
---|
347 |
|
---|
348 | $text =~ s@([^\\])\$(.*?[^\\])\$
|
---|
349 | @$1.process_latex_math($html_markup,$2)@xsge;
|
---|
350 |
|
---|
351 |
|
---|
352 | # is this an amstext environment, or just custom for that input file?
|
---|
353 | $text =~ s@\\begin{(algorithm)}(.*?)\\end{\1}@remove_equals($2)@ges;
|
---|
354 |
|
---|
355 | # convert latex-style accented characters.
|
---|
356 | $self->latex_accents_to_utf8(\$text);
|
---|
357 |
|
---|
358 | # replace quotes with utf-8
|
---|
359 |
|
---|
360 | $text =~ s/``/\xe2\xc0\x9c/g; # Latex-specific, left-dbl quote (“)
|
---|
361 | $text =~ s/''/\xe2\xc0\x9d/g; # Latex-specific, right-dbl quote (”)
|
---|
362 | $text =~ s/`/\xe2\xc0\x98/g; # single left quote
|
---|
363 | $text =~ s/'/\xe2\xc0\x99/g; # single right quote
|
---|
364 |
|
---|
365 | ###### remove/replace latex commands ######
|
---|
366 | ### commands that expand to something that gets displayed ###
|
---|
367 | $text =~ s~\\ldots~…~g;
|
---|
368 | $text =~ s~\\hrule~<hr/>\n~g;
|
---|
369 | $text =~ s~\\maketitle~ ~;
|
---|
370 | ### space commands ###
|
---|
371 | $text =~ s~\\[vh]skip\s+\S+~~g;
|
---|
372 | $text =~ s~\\vspace\*?{.*?}~<div> </div>~g; # vertical space
|
---|
373 | $text =~ s~\\\w+skip~ ~g; # \smallskip \medskip \bigskip \baselineskip etc
|
---|
374 | $text =~ s~\\noindent\b~~g;
|
---|
375 | # newpage, etc
|
---|
376 | $text =~ s~\\(?:clearemptydoublepage|newpage)~~g;
|
---|
377 | ### counters, contents, environments, labels, etc ###
|
---|
378 | $text =~ s~\\(?:addcontentsline){.*?}{.*?}{.*}~~g;
|
---|
379 | $text =~ s~\s*\\begin{itemize}\s*~\n<ul>\n~g;
|
---|
380 | $text =~ s~\s*\\end{itemize}\s*~</li></ul>\n~g;
|
---|
381 | $text =~ s~\s*\\begin{enumerate}\s*~<ol>\n~g;
|
---|
382 | $text =~ s~\s*\\end{enumerate}\s*~</li></ol>\n~g;
|
---|
383 | if ($text =~ s~\s*\\item~</li>\n<li>~g) {
|
---|
384 | # (count for first list item)
|
---|
385 | $text =~ s~<([ou])l>\s*</li>\s*~<$1l>~g;
|
---|
386 | }
|
---|
387 | $text =~ s~\\(?:label|begin|end){.*?}\s*\n?~ ~g; # remove tag and contents
|
---|
388 | $text =~ s~\\(?:tableofcontents|listoffigures)~ ~g;
|
---|
389 | ### font sizes/styles ###
|
---|
390 | $text =~ s~\\(?:tiny|small|footnotesize|normalsize|large|Large|huge|Huge)\b~~g;
|
---|
391 |
|
---|
392 | if ($html_markup) {
|
---|
393 | $text =~ s~\\section\*?{([^\}]+)}\s*\n?~<H1>$1</H1>\n~g;
|
---|
394 | $text =~ s~\\subsection\*?{(.*?)}\s*\n?~<H2>$1</H2>\n~g;
|
---|
395 | $text =~ s~{\\tt\s*(.*?)}~<tt>$1</tt>~g;
|
---|
396 | $text =~ s~\\(?:texttt|tt|ttseries)\s*{(.*?)}~<tt>$1</tt>~g;
|
---|
397 | $text =~ s~\\emph{(.*?)}~<em>$1</em>~g;
|
---|
398 | $text =~ s~{\\(?:em|it)\s*(.*?)}~<em>$1</em>~g;
|
---|
399 | $text =~ s~{\\(?:bf|bfseries)\s*(.*?)}~<strong>$1</strong>~g;
|
---|
400 | $text =~ s~\\(?:textbf|bf|bfseries)\s*{(.*?)}~<strong>$1</strong>~g;
|
---|
401 | } else {
|
---|
402 | # remove tags for text-only
|
---|
403 | $text =~ s~\\(?:textbf|bf|bfseries|em|emph|tt|rm|texttt)\b~ ~g;
|
---|
404 | }
|
---|
405 | $text =~ s ~ {\\sc\s+(.*?)} ~
|
---|
406 | {<span style="font-variant:\ small-caps">$1</span>} ~gx;
|
---|
407 | # ignore these font tags (if there are any left)
|
---|
408 | # sf is sans-serif
|
---|
409 | $text =~ s~\\(?:mdseries|textmd|bfseries|textbf|sffamily|sf|sc)\b~ ~;
|
---|
410 | #### end font-related stuff ####
|
---|
411 |
|
---|
412 | ### remove all other commands with optional arguments... ###
|
---|
413 | # don't remove commands without { }....
|
---|
414 | # $text =~ s~\\\w+(\[.*?\])?\s*~~g;
|
---|
415 | # $text =~ s~\\noopsort{[^}]+\}~~g;
|
---|
416 | # verbatim
|
---|
417 | $text =~ s~\\verb(.)(.*?)\1~verb_text($1)~ge;
|
---|
418 | # remove tags, keep contents for \tag[optional]{contents}
|
---|
419 | while ($text =~ s~\\\w+(\[.*?\])?{([^}]+)}~$2 ~g) {;} # all other commands
|
---|
420 |
|
---|
421 | # remove latex groupings { } (but not \{ or \} )
|
---|
422 | while ($text =~ s/([^\\])[\{\}]/$1/g) {;} # needed for "...}{..."
|
---|
423 | $text =~ s/^\{//; # remove { if first char
|
---|
424 |
|
---|
425 | # latex characters
|
---|
426 | # spaces - nobr space (~), opt break (\-), append ("#" - bibtex only)
|
---|
427 | $text =~ s/([^\\])~+/$1 /g; # non-breaking space "~"
|
---|
428 | # optional break "\-"
|
---|
429 | if ($text =~ m/[^&]\#/) { # concat macros (bibtex) but not HTML codes
|
---|
430 | # the non-macro bits have quotes around them - we just remove quotes
|
---|
431 | # XXX bibtex and latex differ here (for the '#' char)...
|
---|
432 | $text =~ s/([^&])[\"\#]/$1/g;
|
---|
433 | }
|
---|
434 | # dashes. Convert (m|n)-dash into single dash for html.
|
---|
435 | $text =~ s~\-\-+~\-~g;
|
---|
436 |
|
---|
437 | # quoted { } chars
|
---|
438 | $text =~ s~\\{~{~g;
|
---|
439 | $text =~ s~\\}~}~g;
|
---|
440 |
|
---|
441 | # spaces
|
---|
442 | $text =~ s~\\ ~ ~g;
|
---|
443 |
|
---|
444 | # finally to protect against macro language...
|
---|
445 | # greenstone-specific
|
---|
446 | $text =~ s~\[~&\#91;~g;
|
---|
447 | $text =~ s~\]~&\#93;~g;
|
---|
448 | $text =~ s~(?<!\\)([\\_])~\\$1~g;
|
---|
449 |
|
---|
450 | if ($html_markup) {
|
---|
451 | $text =~ s~\n{2,}~\n</p>\n<p>~g;
|
---|
452 | return "<p>$text</p>";
|
---|
453 | }
|
---|
454 |
|
---|
455 | return $text;
|
---|
456 | }
|
---|
457 |
|
---|
458 | # only used by process_latex for \verb....
|
---|
459 | sub verb_text {
|
---|
460 | my $verbatim=shift;
|
---|
461 | $verbatim =~ s/([{}_])/\\$1/g;
|
---|
462 | return $verbatim;
|
---|
463 | }
|
---|
464 | # only used by process_latex_math
|
---|
465 | # returns a unicode char if applicable, otherwise ascii
|
---|
466 | sub math_fraction {
|
---|
467 | my $num=$1;
|
---|
468 | my $denom=$2;
|
---|
469 |
|
---|
470 | if ($num==1 && $denom==2) {return chr(0xc2).chr(0xbd)}
|
---|
471 | if ($num==1 && $denom==4) {return chr(0xc2).chr(0xbc)}
|
---|
472 | if ($num==3 && $denom==4) {return chr(0xc2).chr(0xbe)}
|
---|
473 | return "$num/$denom";
|
---|
474 | }
|
---|
475 |
|
---|
476 | sub process_latex_math {
|
---|
477 |
|
---|
478 | my $text = pop; # if given one or two args, this is the last one...
|
---|
479 | my $html_markup=pop; # if given two args, this is the first one else undef
|
---|
480 |
|
---|
481 | $text =~ s~\\,~ ~g; # forces a space?
|
---|
482 | $text =~ s~\\infty~infinity~g; # or unicode 0x221E...
|
---|
483 |
|
---|
484 | # use this one when more things can read 3-byte utf8 values like this!
|
---|
485 | # $text =~ s~\\cup\b~\xe2\xc8\xaa~g; # union operator - unicode 0x222a
|
---|
486 | $text =~ s~\\cup\b~ U ~g;
|
---|
487 |
|
---|
488 | $text =~ s~\\frac\s*{(.+?)}{(.+?)}~math_fraction($1,$2)~ge;
|
---|
489 |
|
---|
490 | if ($html_markup) {
|
---|
491 | $text =~ s~\^{(.*?)}~<sup>$1</sup>~g; # a^b superscript
|
---|
492 | $text =~ s~\^([^\{])~<sup>$1</sup>~g;
|
---|
493 | $text =~ s~\_{(.*?)}~<sub>$1</sub>~g; # a_b subscript
|
---|
494 | $text =~ s~\_([^\{])~<sub>$1</sub>~g;
|
---|
495 |
|
---|
496 | $text =~ s~\\ldots~…~g; # use html named entity for now
|
---|
497 |
|
---|
498 | # put all other command names in italics for now
|
---|
499 | $text =~ s~\\([\w]+)~<i>$1</i> ~g;
|
---|
500 | }
|
---|
501 |
|
---|
502 | # special cases, for some input files
|
---|
503 | if ($text =~ m~^\\\w+$~) {
|
---|
504 | $text="{" . $text . "}";
|
---|
505 | }
|
---|
506 |
|
---|
507 | return $text;
|
---|
508 | }
|
---|
509 |
|
---|
510 |
|
---|
511 |
|
---|
512 | sub latex_accents_to_utf8 {
|
---|
513 |
|
---|
514 | # note - this is really ugly, but it works. There may be a prettier way
|
---|
515 | # of mapping latex accented chars to utf8, but we just brute force it here.
|
---|
516 | # Also, this isn't complete - not every single possible accented letter
|
---|
517 | # is in here yet, but most of the common ones are.
|
---|
518 |
|
---|
519 | my %utf8_chars =
|
---|
520 | (
|
---|
521 | # acutes
|
---|
522 | '\'a' => chr(0xc3).chr(0xa1),
|
---|
523 | '\'c' => chr(0xc4).chr(0x87),
|
---|
524 | '\'e' => chr(0xc3).chr(0xa9),
|
---|
525 | '\'i' => chr(0xc3).chr(0xad),
|
---|
526 | '\'l' => chr(0xc3).chr(0xba),
|
---|
527 | '\'n' => chr(0xc3).chr(0x84),
|
---|
528 | '\'o' => chr(0xc3).chr(0xb3),
|
---|
529 | '\'r' => chr(0xc5).chr(0x95),
|
---|
530 | '\'s' => chr(0xc5).chr(0x9b),
|
---|
531 | '\'u' => chr(0xc3).chr(0xba),
|
---|
532 | '\'y' => chr(0xc3).chr(0xbd),
|
---|
533 | '\'z' => chr(0xc5).chr(0xba),
|
---|
534 | # graves
|
---|
535 | '`a' => chr(0xc3).chr(0xa0),
|
---|
536 | '`A' => chr(0xc3).chr(0x80),
|
---|
537 | '`e' => chr(0xc3).chr(0xa8),
|
---|
538 | '`E' => chr(0xc3).chr(0x88),
|
---|
539 | '`i' => chr(0xc3).chr(0xac),
|
---|
540 | '`I' => chr(0xc3).chr(0x8c),
|
---|
541 | '`o' => chr(0xc3).chr(0xb2),
|
---|
542 | '`O' => chr(0xc3).chr(0x92),
|
---|
543 | '`u' => chr(0xc3).chr(0xb9),
|
---|
544 | '`U' => chr(0xc3).chr(0x99),
|
---|
545 | # circumflex
|
---|
546 | '^a' => chr(0xc3).chr(0xa2),
|
---|
547 | '^A' => chr(0xc3).chr(0x82),
|
---|
548 | '^c' => chr(0xc4).chr(0x89),
|
---|
549 | '^C' => chr(0xc4).chr(0x88),
|
---|
550 | '^e' => chr(0xc3).chr(0xaa),
|
---|
551 | '^E' => chr(0xc3).chr(0x8a),
|
---|
552 | '^g' => chr(0xc4).chr(0x9d),
|
---|
553 | '^G' => chr(0xc4).chr(0x9c),
|
---|
554 | '^h' => chr(0xc4).chr(0xa5),
|
---|
555 | '^H' => chr(0xc4).chr(0xa4),
|
---|
556 | '^i' => chr(0xc3).chr(0xae),
|
---|
557 | '^I' => chr(0xc3).chr(0x8e),
|
---|
558 | '^j' => chr(0xc4).chr(0xb5),
|
---|
559 | '^J' => chr(0xc4).chr(0xb4),
|
---|
560 | '^o' => chr(0xc3).chr(0xb4),
|
---|
561 | '^O' => chr(0xc3).chr(0x94),
|
---|
562 | '^s' => chr(0xc5).chr(0x9d),
|
---|
563 | '^S' => chr(0xc5).chr(0x9c),
|
---|
564 | '^u' => chr(0xc3).chr(0xa2),
|
---|
565 | '^U' => chr(0xc3).chr(0xbb),
|
---|
566 | '^w' => chr(0xc5).chr(0xb5),
|
---|
567 | '^W' => chr(0xc5).chr(0xb4),
|
---|
568 | '^y' => chr(0xc5).chr(0xb7),
|
---|
569 | '^Y' => chr(0xc5).chr(0xb6),
|
---|
570 |
|
---|
571 | # diaeresis
|
---|
572 | '"a' => chr(0xc3).chr(0xa4),
|
---|
573 | '"A' => chr(0xc3).chr(0x84),
|
---|
574 | '"e' => chr(0xc3).chr(0xab),
|
---|
575 | '"E' => chr(0xc3).chr(0x8b),
|
---|
576 | '"i' => chr(0xc3).chr(0xaf),
|
---|
577 | '"I' => chr(0xc3).chr(0x8f),
|
---|
578 | '"\\\\i' => chr(0xc3).chr(0xaf),
|
---|
579 | '"\\\\I' => chr(0xc3).chr(0x8f),
|
---|
580 | '"o' => chr(0xc3).chr(0xb6),
|
---|
581 | '"O' => chr(0xc3).chr(0x96),
|
---|
582 | '"u' => chr(0xc3).chr(0xbc),
|
---|
583 | '"U' => chr(0xc3).chr(0x9c),
|
---|
584 | '"y' => chr(0xc3).chr(0xbf),
|
---|
585 | '"Y' => chr(0xc3).chr(0xb8),
|
---|
586 | # tilde
|
---|
587 | '~A' => chr(0xc3).chr(0x83),
|
---|
588 | '~N' => chr(0xc3).chr(0x91),
|
---|
589 | '~O' => chr(0xc3).chr(0x95),
|
---|
590 | '~a' => chr(0xc3).chr(0xa3),
|
---|
591 | '~n' => chr(0xc3).chr(0xb1),
|
---|
592 | '~o' => chr(0xc3).chr(0xb5),
|
---|
593 | # caron - handled specially
|
---|
594 | # double acute
|
---|
595 | # ring
|
---|
596 | # dot
|
---|
597 | '.c' => chr(0xc4).chr(0x8b),
|
---|
598 | '.C' => chr(0xc4).chr(0x8a),
|
---|
599 | '.e' => chr(0xc4).chr(0x97),
|
---|
600 | '.E' => chr(0xc4).chr(0x96),
|
---|
601 | '.g' => chr(0xc4).chr(0xa1),
|
---|
602 | '.G' => chr(0xc4).chr(0xa0),
|
---|
603 | '.I' => chr(0xc4).chr(0xb0),
|
---|
604 | '.z' => chr(0xc5).chr(0xbc),
|
---|
605 | '.Z' => chr(0xc5).chr(0xbb),
|
---|
606 | # macron
|
---|
607 | '=a' => chr(0xc4).chr(0x81),
|
---|
608 | '=A' => chr(0xc4).chr(0x80),
|
---|
609 | '=e' => chr(0xc4).chr(0x93),
|
---|
610 | '=E' => chr(0xc4).chr(0x92),
|
---|
611 | '=i' => chr(0xc4).chr(0xab),
|
---|
612 | '=I' => chr(0xc4).chr(0xaa),
|
---|
613 | '=o' => chr(0xc4).chr(0x8d),
|
---|
614 | '=O' => chr(0xc4).chr(0x8c),
|
---|
615 | '=u' => chr(0xc4).chr(0xab),
|
---|
616 | '=U' => chr(0xc4).chr(0xaa),
|
---|
617 |
|
---|
618 | # stroke - handled specially - see below
|
---|
619 |
|
---|
620 | # cedilla - handled specially
|
---|
621 | );
|
---|
622 |
|
---|
623 | # these are one letter latex commands - we make sure they're not a longer
|
---|
624 | # command name. eg {\d} is d+stroke, so careful of \d
|
---|
625 | my %special_utf8_chars =
|
---|
626 | (
|
---|
627 | # breve
|
---|
628 | 'u g' => chr(0xc4).chr(0x9f),
|
---|
629 | 'u G' => chr(0xc4).chr(0x9e),
|
---|
630 | 'u i' => chr(0xc4).chr(0xad),
|
---|
631 | 'u I' => chr(0xc4).chr(0xac),
|
---|
632 | 'u o' => chr(0xc5).chr(0x8f),
|
---|
633 | 'u O' => chr(0xc5).chr(0x8e),
|
---|
634 | 'u u' => chr(0xc5).chr(0xad),
|
---|
635 | 'u U' => chr(0xc5).chr(0xac),
|
---|
636 | 'u z' => chr(0xc5).chr(0xbe), # !!! no such char, but common mistake
|
---|
637 | 'u Z' => chr(0xc5).chr(0xbd), # used instead of v Z !!!
|
---|
638 | # caron
|
---|
639 | 'v c' => chr(0xc4).chr(0x8d),
|
---|
640 | 'v C' => chr(0xc4).chr(0x8c),
|
---|
641 | 'v n' => chr(0xc5).chr(0x88),
|
---|
642 | 'v N' => chr(0xc5).chr(0x87),
|
---|
643 | 'v s' => chr(0xc5).chr(0xa1),
|
---|
644 | 'v S' => chr(0xc5).chr(0xa5),
|
---|
645 | 'v z' => chr(0xc5).chr(0xbe),
|
---|
646 | 'v Z' => chr(0xc5).chr(0xbd),
|
---|
647 | # cedilla
|
---|
648 | 'c c' => chr(0xc3).chr(0xa7),
|
---|
649 | 'c C' => chr(0xc3).chr(0x87),
|
---|
650 | 'c g' => chr(0xc4).chr(0xa3),
|
---|
651 | 'c G' => chr(0xc4).chr(0xa2),
|
---|
652 | 'c k' => chr(0xc4).chr(0xb7),
|
---|
653 | 'c K' => chr(0xc4).chr(0xb6),
|
---|
654 | 'c l' => chr(0xc4).chr(0xbc),
|
---|
655 | 'c L' => chr(0xc4).chr(0xbb),
|
---|
656 | 'c n' => chr(0xc5).chr(0x86),
|
---|
657 | 'c N' => chr(0xc5).chr(0x85),
|
---|
658 | 'c r' => chr(0xc5).chr(0x97),
|
---|
659 | 'c R' => chr(0xc5).chr(0x96),
|
---|
660 | 'c s' => chr(0xc5).chr(0x9f),
|
---|
661 | 'c S' => chr(0xc5).chr(0x9e),
|
---|
662 | 'c t' => chr(0xc5).chr(0xa3),
|
---|
663 | 'c T' => chr(0xc5).chr(0xa2),
|
---|
664 | # double acute / Hungarian accent
|
---|
665 | 'H O' => chr(0xc5).chr(0x90),
|
---|
666 | 'H o' => chr(0xc5).chr(0x91),
|
---|
667 | 'H U' => chr(0xc5).chr(0xb0),
|
---|
668 | 'H u' => chr(0xc5).chr(0xb1),
|
---|
669 |
|
---|
670 | # stroke
|
---|
671 | 'd' => chr(0xc4).chr(0x91),
|
---|
672 | 'D' => chr(0xc4).chr(0x90),
|
---|
673 | 'h' => chr(0xc4).chr(0xa7),
|
---|
674 | # 'H' => chr(0xc4).chr(0xa6), # !! this normally(!!?) means Hung. umlaut
|
---|
675 | 'i' => chr(0xc4).chr(0xb1), # dotless lowercase i
|
---|
676 | 'l' => chr(0xc5).chr(0x82),
|
---|
677 | 'L' => chr(0xc5).chr(0x81),
|
---|
678 | 'o' => chr(0xc3).chr(0xb8),
|
---|
679 | 'O' => chr(0xc3).chr(0x98),
|
---|
680 | 't' => chr(0xc5).chr(0xa7),
|
---|
681 | 'T' => chr(0xc5).chr(0xa6),
|
---|
682 |
|
---|
683 | # other special chars
|
---|
684 | 'ss' => chr(0xc3).chr(0x9f), # german ss/szlig/sharp s
|
---|
685 | 'aa' =>,chr(0xc3).chr(0xa5), # scandanavian/latin a with ring
|
---|
686 | );
|
---|
687 |
|
---|
688 | my $self=shift;
|
---|
689 | my $textref=shift;
|
---|
690 |
|
---|
691 | my $outhandle=$self->{'outhandle'};
|
---|
692 | my $text=$$textref;
|
---|
693 |
|
---|
694 | # remove space (if any) between \ and letter to accent (eg {\' a})
|
---|
695 | $text =~ s!(\\[`'="^~\.])\s(\w)\b!$1$2!g; # for emacs indenting... `]);
|
---|
696 |
|
---|
697 | # remove {} around a single character (eg \'{e})
|
---|
698 | $text =~ s!(\\[`'="^~\.]){(\w)}!{$1$2}!g; # for emacs indenting... `]);
|
---|
699 |
|
---|
700 | ## only in bibtex... not in latex proper?!
|
---|
701 | ### \, is another way of doing cedilla \c
|
---|
702 | ##$text =~ s~\\,(.)~\\c $1~g;
|
---|
703 |
|
---|
704 | # remove {} around a single character for special 1 letter commands -
|
---|
705 | # need to insert a space. Eg \v{s} -> {\v s}
|
---|
706 | $text =~ s~(\\[uvcH]){(\w)}~{$1 $2}~g;
|
---|
707 |
|
---|
708 | # only do if the text contains a '\' character.
|
---|
709 | if ($text =~ m|\\|) {
|
---|
710 | # "normal" accents - ie non-alpha latex tag
|
---|
711 | # xxx used to have ([\w]\b)@ (for word boundary)
|
---|
712 | while ($text =~ m/\\([`'="^~\.])([\w])/) { # for emacs `])){
|
---|
713 | my $tex="$1$2"; my $char="$2";
|
---|
714 | my $replacement=$utf8_chars{$tex};
|
---|
715 | if (!defined($replacement)) {
|
---|
716 | $text =~ m~(.{20}\\\Q$tex\E.{20})~s;
|
---|
717 | print $outhandle . $self->{'plugin_type'} .
|
---|
718 | ": Warning: unknown latex accent \"$tex\""
|
---|
719 | . " in \"$1\"\n";
|
---|
720 | $replacement=$char;
|
---|
721 | }
|
---|
722 | $text =~ s/\\\Q$tex/$replacement/g;
|
---|
723 | }
|
---|
724 |
|
---|
725 | # where the following letter matters (eg "sm\o rrebr\o d", \ss{})
|
---|
726 | # only do the change if immediately followed by a space, }, {, or \
|
---|
727 | # one letter accents ( + ss / aa)
|
---|
728 | while ($text =~ m~\\([DdhiLlOoTt]|ss|aa)[{}\s\"\\]~) {
|
---|
729 | my $tex=$1;
|
---|
730 | my $replacement=$special_utf8_chars{$tex};
|
---|
731 | if (!defined($replacement)) {
|
---|
732 | $text =~ m~(.{20}\\\Q$tex\E.{20})~s;
|
---|
733 | print $outhandle $self->{'plugin_type'} .
|
---|
734 | ": Warning: unknown latex accent \"$tex\""
|
---|
735 | . " in \"$1\"\n";
|
---|
736 | $replacement=$tex;
|
---|
737 | }
|
---|
738 | ($text =~ s/{\\$tex}/$replacement/g) or
|
---|
739 | $text =~ s/\\$tex([{}\s\"\\])/$replacement$1/g;
|
---|
740 |
|
---|
741 | }
|
---|
742 |
|
---|
743 | # one letter latex accent commands that affect following letter
|
---|
744 | while ($text =~ m~\\([uvcH]) ([\w])~) {
|
---|
745 | my $tex="$1 $2"; my $char="$2";
|
---|
746 | my $replacement=$special_utf8_chars{$tex};
|
---|
747 | if (!defined($replacement)) {
|
---|
748 | $text =~ m~(.{20}\\\Q$tex\E.{20})~s;
|
---|
749 | print $outhandle $self->{'plugin_type'} .
|
---|
750 | ": Warning: unknown latex accent \"$tex\""
|
---|
751 | . " in \"$1\"\n";
|
---|
752 | $replacement=$char;
|
---|
753 | }
|
---|
754 | $text =~ s/\\$tex/$replacement/g;
|
---|
755 | }
|
---|
756 | }
|
---|
757 | $textref=\$text;
|
---|
758 | }
|
---|
759 |
|
---|
760 |
|
---|
761 | # modules must return true
|
---|
762 | 1;
|
---|