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