1 | #!/usr/bin/perl -w
|
---|
2 |
|
---|
3 | ###########################################################################
|
---|
4 | #
|
---|
5 | # gsConvert.pl -- convert documents to HTML or TEXT format
|
---|
6 | #
|
---|
7 | # A component of the Greenstone digital library software
|
---|
8 | # from the New Zealand Digital Library Project at the
|
---|
9 | # University of Waikato, New Zealand.
|
---|
10 | #
|
---|
11 | # Copyright (C) 1999-2002 New Zealand Digital Library Project
|
---|
12 | #
|
---|
13 | # This program is free software; you can redistribute it and/or modify
|
---|
14 | # it under the terms of the GNU General Public License as published by
|
---|
15 | # the Free Software Foundation; either version 2 of the License, or
|
---|
16 | # (at your option) any later version.
|
---|
17 | #
|
---|
18 | # This program is distributed in the hope that it will be useful,
|
---|
19 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
20 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
21 | # GNU General Public License for more details.
|
---|
22 | #
|
---|
23 | # You should have received a copy of the GNU General Public License
|
---|
24 | # along with this program; if not, write to the Free Software
|
---|
25 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
26 | #
|
---|
27 | ###########################################################################
|
---|
28 |
|
---|
29 | # gsConvert.pl converts documents in a range of formats to HTML or TEXT
|
---|
30 | # by exploiting third-party programs. The sources of these are usually found
|
---|
31 | # in the $GSDLHOME/packages directory, and the executables should live in
|
---|
32 | # $GSDLHOME/bin/$GSDLOS (which is on the search path).
|
---|
33 | #
|
---|
34 | # Currently, we can convert the following formats by using external
|
---|
35 | # conversion utilities:
|
---|
36 | # Microsoft Word (versions 2,6,7 [==95?], 8[==97?], 9[==2000?]), RTF,
|
---|
37 | # Adobe PDF, PostScript, MS PowerPoint (95 and 97), and MS Excel (95 and 97).
|
---|
38 | #
|
---|
39 | # We can try to convert any file to text with a perl implementation of the
|
---|
40 | # UNIX strings command.
|
---|
41 | #
|
---|
42 | # We try to convert Postscript files to text using "gs" which is often on
|
---|
43 | # *nix machines. We fall back to performing weak text extraction by using
|
---|
44 | # regular expressions.
|
---|
45 |
|
---|
46 | BEGIN {
|
---|
47 | die "GSDLHOME not set\n" unless defined $ENV{'GSDLHOME'};
|
---|
48 | unshift (@INC, "$ENV{'GSDLHOME'}/perllib");
|
---|
49 | }
|
---|
50 |
|
---|
51 | use parsargv;
|
---|
52 | use util;
|
---|
53 | use Cwd;
|
---|
54 | use File::Basename;
|
---|
55 |
|
---|
56 | # Are we running on WinNT or Win2000 (or later)?
|
---|
57 | my $is_winnt_2000=eval {require Win32; return (Win32::IsWinNT()); return 0;};
|
---|
58 | if (!defined($is_winnt_2000)) {$is_winnt_2000=0;}
|
---|
59 |
|
---|
60 | sub print_usage
|
---|
61 | {
|
---|
62 | print STDERR "\n";
|
---|
63 | print STDERR "gsConvert.pl: Converts documents in a range of formats to html\n";
|
---|
64 | print STDERR " or text using third-party programs.\n\n";
|
---|
65 | print STDERR " usage: $0 [options] filename\n";
|
---|
66 | print STDERR " options:\n\t-type\tdoc|pdf|ps|ppt|rtf|xls\t(input file type)\n";
|
---|
67 | print STDERR "\t-errlog\t<filename>\t(append err messages)\n";
|
---|
68 | print STDERR "\t-output\thtml|text\n";
|
---|
69 | print STDERR "\t-timeout\t<max cpu seconds>\t(ulimit on unix systems)\n";
|
---|
70 | exit(1);
|
---|
71 | }
|
---|
72 |
|
---|
73 | my $faillogfile="";
|
---|
74 |
|
---|
75 | sub main
|
---|
76 | {
|
---|
77 | my (@ARGV) = @_;
|
---|
78 | my ($input_type,$output_type,$verbose,$timeout);
|
---|
79 |
|
---|
80 | $timeout = 0;
|
---|
81 | # read command-line arguments
|
---|
82 | if (!parsargv::parse(\@ARGV,
|
---|
83 | 'type/(doc|pdf|ps|ppt|rtf|xls)/', \$input_type,
|
---|
84 | '/errlog/.*/', \$faillogfile,
|
---|
85 | 'output/(html|text)/', \$output_type,
|
---|
86 | 'timeout/\d+/0',\$timeout,
|
---|
87 | 'verbose/\d+/0', \$verbose))
|
---|
88 | {
|
---|
89 | print_usage();
|
---|
90 | }
|
---|
91 |
|
---|
92 | # Make sure the input file exists and can be opened for reading
|
---|
93 | if (scalar(@ARGV!=1)) {
|
---|
94 | print_usage();
|
---|
95 | }
|
---|
96 |
|
---|
97 | my $input_filename = $ARGV[0];
|
---|
98 | if (!-r $input_filename) {
|
---|
99 | print STDERR "Error: unable to open $input_filename for reading\n";
|
---|
100 | exit(1);
|
---|
101 | }
|
---|
102 |
|
---|
103 | # Deduce filenames
|
---|
104 | my ($tailname,$dirname,$suffix)
|
---|
105 | = File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");
|
---|
106 | my $output_filestem = &util::filename_cat($dirname, "$tailname");
|
---|
107 |
|
---|
108 | if ($input_type eq "")
|
---|
109 | {
|
---|
110 | $input_type = lc (substr($suffix,1,length($suffix)-1));
|
---|
111 | }
|
---|
112 |
|
---|
113 | # Change to temporary working directory
|
---|
114 | my $stored_dir = cwd();
|
---|
115 | chdir ($dirname) || die "Unable to change to directory $dirname";
|
---|
116 |
|
---|
117 | # Select convert utility
|
---|
118 | if (!defined $input_type) {
|
---|
119 | print STDERR "Error: No filename extension or input type defined\n";
|
---|
120 | exit(1);
|
---|
121 | }
|
---|
122 | elsif ($input_type eq "doc") {
|
---|
123 | print &convertDOC($input_filename, $output_filestem, $output_type);
|
---|
124 | print "\n";
|
---|
125 | }
|
---|
126 | elsif ($input_type eq "rtf") {
|
---|
127 | print &convertRTF($input_filename, $output_filestem, $output_type);
|
---|
128 | print "\n";
|
---|
129 | }
|
---|
130 | elsif ($input_type eq "pdf") {
|
---|
131 | print &convertPDF($dirname, $input_filename, $output_filestem, $output_type);
|
---|
132 | print "\n";
|
---|
133 | }
|
---|
134 | elsif ($input_type eq "ps") {
|
---|
135 | print &convertPS($input_filename, $output_filestem, $output_type);
|
---|
136 | print "\n";
|
---|
137 | }
|
---|
138 | elsif ($input_type eq "ppt") {
|
---|
139 | print &convertPPT($input_filename, $output_filestem, $output_type);
|
---|
140 | print "\n";
|
---|
141 | }
|
---|
142 | elsif ($input_type eq "xls") {
|
---|
143 | print &convertXLS($input_filename, $output_filestem, $output_type);
|
---|
144 | print "\n";
|
---|
145 | }
|
---|
146 | else {
|
---|
147 | print STDERR "Error: Unable to convert type '$input_type'\n";
|
---|
148 | exit(1);
|
---|
149 | }
|
---|
150 |
|
---|
151 | # restore to original working directory
|
---|
152 | chdir ($stored_dir) || die "Unable to return to directory $stored_dir";
|
---|
153 |
|
---|
154 | }
|
---|
155 |
|
---|
156 | &main(@ARGV);
|
---|
157 |
|
---|
158 |
|
---|
159 |
|
---|
160 | # Document-type conversion functions
|
---|
161 | #
|
---|
162 | # The following functions attempt to convert documents from their
|
---|
163 | # input type to the specified output type. If no output type was
|
---|
164 | # given, then they first attempt HTML, and then TEXT.
|
---|
165 | #
|
---|
166 | # Each returns the output type ("html" or "text") or "fail" if no
|
---|
167 | # conversion is possible.
|
---|
168 |
|
---|
169 | # Convert a Microsoft word document
|
---|
170 |
|
---|
171 | sub convertDOC {
|
---|
172 | ($input_filename, $output_filestem, $output_type) = @_;
|
---|
173 |
|
---|
174 | # Many .doc files are not in fact word documents!
|
---|
175 | my $realtype = &find_docfile_type($input_filename);
|
---|
176 |
|
---|
177 | if ($realtype eq "word6" || $realtype eq "word7" || $realtype eq "word8") {
|
---|
178 | return &convertWord678($input_filename, $output_filestem, $output_type);
|
---|
179 | } elsif ($realtype eq "rtf") {
|
---|
180 | return &convertRTF($input_filename, $output_filestem, $output_type);
|
---|
181 | } else {
|
---|
182 | return &convertAnything($input_filename, $output_filestem, $output_type);
|
---|
183 | }
|
---|
184 | }
|
---|
185 |
|
---|
186 | # Convert a Microsoft word 6/7/8 document
|
---|
187 |
|
---|
188 | sub convertWord678 {
|
---|
189 | ($input_filename, $output_filestem, $output_type) = @_;
|
---|
190 |
|
---|
191 | my $success = 0;
|
---|
192 |
|
---|
193 | # Attempt specialised conversion to HTML
|
---|
194 | if (!$output_type || ($output_type =~ /html/i)) {
|
---|
195 | $success = &doc_to_html($input_filename, $output_filestem);
|
---|
196 | if ($success) {
|
---|
197 | return "html";
|
---|
198 | }
|
---|
199 | }
|
---|
200 |
|
---|
201 | return &convertAnything($input_filename, $output_filestem, $output_type);
|
---|
202 | }
|
---|
203 |
|
---|
204 |
|
---|
205 | # Convert a Rich Text Format (RTF) file
|
---|
206 |
|
---|
207 | sub convertRTF {
|
---|
208 | ($input_filename, $output_filestem, $output_type) = @_;
|
---|
209 |
|
---|
210 | my $success = 0;
|
---|
211 |
|
---|
212 | # Attempt specialised conversion to HTML
|
---|
213 | if (!$output_type || ($output_type =~ /html/i)) {
|
---|
214 | $success = &rtf_to_html($input_filename, $output_filestem);
|
---|
215 | if ($success) {
|
---|
216 | return "html";
|
---|
217 | }
|
---|
218 | }
|
---|
219 |
|
---|
220 | # rtf is so ugly that's it's not worth running strings over.
|
---|
221 | # One day I'll write some quick'n'dirty regexps to try to extract text - jrm21
|
---|
222 | # return &convertAnything($input_filename, $output_filestem, $output_type);
|
---|
223 | return "fail";
|
---|
224 | }
|
---|
225 |
|
---|
226 |
|
---|
227 | # Convert an unidentified file
|
---|
228 |
|
---|
229 | sub convertAnything {
|
---|
230 | ($input_filename, $output_filestem, $output_type) = @_;
|
---|
231 |
|
---|
232 | my $success = 0;
|
---|
233 |
|
---|
234 | # Attempt simple conversion to HTML
|
---|
235 | if (!$output_type || ($output_type =~ /html/i)) {
|
---|
236 | $success = &any_to_html($input_filename, $output_filestem);
|
---|
237 | if ($success) {
|
---|
238 | return "html";
|
---|
239 | }
|
---|
240 | }
|
---|
241 |
|
---|
242 | # Convert to text
|
---|
243 | if (!$output_type || ($output_type =~ /text/i)) {
|
---|
244 | $success = &any_to_text($input_filename, $output_filestem);
|
---|
245 | if ($success) {
|
---|
246 | return "text";
|
---|
247 | }
|
---|
248 | }
|
---|
249 | return "fail";
|
---|
250 | }
|
---|
251 |
|
---|
252 |
|
---|
253 |
|
---|
254 | # Convert an Adobe PDF document
|
---|
255 |
|
---|
256 | sub convertPDF {
|
---|
257 | my ($dirname, $input_filename, $output_filestem, $output_type) = @_;
|
---|
258 |
|
---|
259 | my $success = 0;
|
---|
260 |
|
---|
261 | # Attempt conversion to HTML
|
---|
262 | if (!$output_type || ($output_type =~ /html/i)) {
|
---|
263 | $success = &pdf_to_html($dirname, $input_filename, $output_filestem);
|
---|
264 | if ($success) {
|
---|
265 | return "html";
|
---|
266 | }
|
---|
267 | }
|
---|
268 |
|
---|
269 | # Attempt conversion to TEXT
|
---|
270 | if (!$output_type || ($output_type =~ /text/i)) {
|
---|
271 | $success = &pdf_to_text($dirname, $input_filename, $output_filestem);
|
---|
272 | if ($success) {
|
---|
273 | return "text";
|
---|
274 | }
|
---|
275 | }
|
---|
276 |
|
---|
277 | return "fail";
|
---|
278 |
|
---|
279 | }
|
---|
280 |
|
---|
281 |
|
---|
282 | # Convert an Adobe PostScript document
|
---|
283 |
|
---|
284 | sub convertPS {
|
---|
285 | ($input_filename, $output_filestem, $output_type) = @_;
|
---|
286 |
|
---|
287 | my $success = 0;
|
---|
288 |
|
---|
289 | # Attempt conversion to TEXT
|
---|
290 | if (!$output_type || ($output_type =~ /text/i)) {
|
---|
291 | $success = &ps_to_text($input_filename, $output_filestem);
|
---|
292 | if ($success) {
|
---|
293 | return "text";
|
---|
294 | }
|
---|
295 | }
|
---|
296 |
|
---|
297 | return "fail";
|
---|
298 |
|
---|
299 | }
|
---|
300 |
|
---|
301 |
|
---|
302 | sub convertPPT {
|
---|
303 | my ($input_filename, $output_filestem, $output_type) = @_;
|
---|
304 |
|
---|
305 | my $success = 0;
|
---|
306 |
|
---|
307 | # Attempt conversion to HTML
|
---|
308 | if (!$output_type || ($output_type =~ /html/i)) {
|
---|
309 | # formulate the command
|
---|
310 | $cmd = "";
|
---|
311 | $cmd .= "perl -S ppttohtml.pl ";
|
---|
312 | $cmd .= " \"$input_filename\" \"$output_filestem.html\"";
|
---|
313 | $cmd .= " 2>\"$output_filestem.err\""
|
---|
314 | if ($ENV{'GSDLOS'} !~ /^windows$/i || $is_winnt_2000);
|
---|
315 |
|
---|
316 |
|
---|
317 | # execute the command
|
---|
318 | $!=0;
|
---|
319 | if (system($cmd)!=0)
|
---|
320 | {
|
---|
321 | print STDERR "Powerpoint 95/97 converter failed $!\n";
|
---|
322 | } else {
|
---|
323 | return "html";
|
---|
324 | }
|
---|
325 | }
|
---|
326 |
|
---|
327 | $success = &any_to_text($input_filename, $output_filestem);
|
---|
328 | if ($success) {
|
---|
329 | return "text";
|
---|
330 | }
|
---|
331 |
|
---|
332 | return "fail";
|
---|
333 | }
|
---|
334 |
|
---|
335 |
|
---|
336 | sub convertXLS {
|
---|
337 | my ($input_filename, $output_filestem, $output_type) = @_;
|
---|
338 |
|
---|
339 | my $success = 0;
|
---|
340 |
|
---|
341 | # Attempt conversion to HTML
|
---|
342 | if (!$output_type || ($output_type =~ /html/i)) {
|
---|
343 | # formulate the command
|
---|
344 | $cmd = "";
|
---|
345 | $cmd .= "perl -S xlstohtml.pl ";
|
---|
346 | $cmd .= " \"$input_filename\" \"$output_filestem.html\"";
|
---|
347 | $cmd .= " 2>\"$output_filestem.err\""
|
---|
348 | if ($ENV{'GSDLOS'} !~ /^windows$/i || $is_winnt_2000);
|
---|
349 |
|
---|
350 |
|
---|
351 | # execute the command
|
---|
352 | $!=0;
|
---|
353 | if (system($cmd)!=0)
|
---|
354 | {
|
---|
355 | print STDERR "Excel 95/97 converter failed $!\n";
|
---|
356 | } else {
|
---|
357 | return "html";
|
---|
358 | }
|
---|
359 | }
|
---|
360 |
|
---|
361 | $success = &any_to_text($input_filename, $output_filestem);
|
---|
362 | if ($success) {
|
---|
363 | return "text";
|
---|
364 | }
|
---|
365 |
|
---|
366 | return "fail";
|
---|
367 | }
|
---|
368 |
|
---|
369 |
|
---|
370 |
|
---|
371 |
|
---|
372 |
|
---|
373 | # Find the real type of a .doc file
|
---|
374 | #
|
---|
375 | # We seem to have a lot of files with a .doc extension that are .rtf
|
---|
376 | # files or Word 5 files. This function attempts to tell the difference.
|
---|
377 |
|
---|
378 | sub find_docfile_type {
|
---|
379 | ($input_filename) = @_;
|
---|
380 |
|
---|
381 | open(CHK, "<$input_filename");
|
---|
382 | binmode(CHK);
|
---|
383 | my $line = "";
|
---|
384 | my $first = 1;
|
---|
385 |
|
---|
386 | while (<CHK>) {
|
---|
387 |
|
---|
388 | $line = $_;
|
---|
389 |
|
---|
390 | if ($first) {
|
---|
391 | # check to see if this is an rtf file
|
---|
392 | if ($line =~ /^\{\\rtf/) {
|
---|
393 | close(CHK);
|
---|
394 | return "rtf";
|
---|
395 | }
|
---|
396 | $first = 0;
|
---|
397 | }
|
---|
398 |
|
---|
399 | # is this is a word 6/7/8 document?
|
---|
400 | if ($line =~ /Word\.Document\.([678])/) {
|
---|
401 | close(CHK);
|
---|
402 | return "word$1";
|
---|
403 | }
|
---|
404 |
|
---|
405 | }
|
---|
406 |
|
---|
407 | return "unknown";
|
---|
408 | }
|
---|
409 |
|
---|
410 |
|
---|
411 |
|
---|
412 | # Specific type-to-type conversions
|
---|
413 | #
|
---|
414 | # Each of the following functions attempts to convert a document from
|
---|
415 | # a specific format to another. If they succeed they return 1 and leave
|
---|
416 | # the output document(s) in the appropriate place; if they fail they
|
---|
417 | # return 0 and delete any working files.
|
---|
418 |
|
---|
419 |
|
---|
420 | # Attempt to convert a word document to html with the wv program
|
---|
421 |
|
---|
422 | sub doc_to_html {
|
---|
423 | ($input_filename, $output_filestem) = @_;
|
---|
424 |
|
---|
425 | my $wvWare = &util::filename_cat($ENV{'GSDLHOME'}, "bin",
|
---|
426 | $ENV{'GSDLOS'}, "wvWare");
|
---|
427 |
|
---|
428 | # don't include path on windows (to avoid having to play about
|
---|
429 | # with quoting when GSDLHOME might contain spaces) but assume
|
---|
430 | # that the PATH is set up correctly
|
---|
431 | $wvWare = "wvWare" if ($ENV{'GSDLOS'} =~ /^windows$/i);
|
---|
432 |
|
---|
433 | my $wv_conf = &util::filename_cat($ENV{'GSDLHOME'}, "etc",
|
---|
434 | "packages", "wv", "wvHtml.xml");
|
---|
435 |
|
---|
436 | my $cmd = "";
|
---|
437 | if ($timeout) {$cmd = "ulimit -t $timeout;";}
|
---|
438 | $cmd .= "$wvWare --charset utf-8 --config \"$wv_conf\"";
|
---|
439 | $cmd .= " \"$input_filename\" > \"$output_filestem.html\"";
|
---|
440 |
|
---|
441 | # redirecting STDERR is a bad idea on windows 95/98
|
---|
442 | $cmd .= " 2> \"$output_filestem.err\""
|
---|
443 | if ($ENV{'GSDLOS'} !~ /^windows$/i || $is_winnt_2000);
|
---|
444 |
|
---|
445 | # execute the command
|
---|
446 | $!=0;
|
---|
447 | if (system($cmd)!=0)
|
---|
448 | {
|
---|
449 | print STDERR "Error executing wv converter:$!\n";
|
---|
450 | if (-s "$output_filestem.err") {
|
---|
451 | open (ERRFILE, "<$output_filestem.err");
|
---|
452 |
|
---|
453 | my $write_to_fail_log=0;
|
---|
454 | if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
|
---|
455 | {$write_to_fail_log=1;}
|
---|
456 |
|
---|
457 | my $line;
|
---|
458 | while ($line=<ERRFILE>) {
|
---|
459 | if ($line =~ /\w/) {
|
---|
460 | print STDERR "$line";
|
---|
461 | print FAILLOG "$line" if ($write_to_fail_log);
|
---|
462 | }
|
---|
463 | if ($line !~ m/startup error/) {next;}
|
---|
464 | print STDERR " (given an invalid .DOC file?)\n";
|
---|
465 | print FAILLOG " (given an invalid .DOC file?)\n"
|
---|
466 | if ($write_to_fail_log);
|
---|
467 |
|
---|
468 | } # while ERRFILE
|
---|
469 | close FAILLOG if ($write_to_fail_log);
|
---|
470 | }
|
---|
471 | print STDERR "Continuing...\n";
|
---|
472 | return 0; # we can try any_to_text
|
---|
473 | }
|
---|
474 |
|
---|
475 | # Was the conversion successful?
|
---|
476 |
|
---|
477 | if (-s "$output_filestem.html") {
|
---|
478 | open(TMP, "$output_filestem.html");
|
---|
479 | $line = <TMP>;
|
---|
480 | close(TMP);
|
---|
481 | if ($line && $line =~ /DOCTYPE HTML/) {
|
---|
482 | &util::rm("$output_filestem.err") if -e "$output_filestem.err";
|
---|
483 | return 1;
|
---|
484 | }
|
---|
485 | }
|
---|
486 |
|
---|
487 | # If here, an error of some sort occurred
|
---|
488 | &util::rm("$output_filestem.html") if -e "$output_filestem.html";
|
---|
489 | if (-e "$output_filestem.err") {
|
---|
490 | if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile"))) {
|
---|
491 | open (ERRLOG,"$output_filestem.err");
|
---|
492 | while (<ERRLOG>) {print FAILLOG $_;}
|
---|
493 | close FAILLOG;
|
---|
494 | close ERRLOG;
|
---|
495 | }
|
---|
496 | &util::rm("$output_filestem.err");
|
---|
497 | }
|
---|
498 |
|
---|
499 | return 0;
|
---|
500 | }
|
---|
501 |
|
---|
502 |
|
---|
503 | # Attempt to convert an RTF document to html with rtftohtml
|
---|
504 |
|
---|
505 | sub rtf_to_html {
|
---|
506 | my ($input_filename, $output_filestem) = @_;
|
---|
507 |
|
---|
508 | # formulate the command
|
---|
509 | $cmd = "";
|
---|
510 | if ($timeout) {$cmd = "ulimit -t $timeout;";}
|
---|
511 | $cmd .= "rtftohtml";
|
---|
512 |
|
---|
513 | $cmd .= " -o \"$output_filestem.html\" \"$input_filename\"";
|
---|
514 |
|
---|
515 | $cmd .= " 2>\"$output_filestem.err\""
|
---|
516 | if ($ENV{'GSDLOS'} !~ /^windows$/i || $is_winnt_2000);
|
---|
517 |
|
---|
518 |
|
---|
519 | # execute the command
|
---|
520 | $!=0;
|
---|
521 | if (system($cmd)!=0)
|
---|
522 | {
|
---|
523 | print STDERR "Error executing rtf converter $!\n";
|
---|
524 | # don't currently bother printing out error log...
|
---|
525 | # keep going, in case it still created an HTML file...
|
---|
526 | }
|
---|
527 |
|
---|
528 | # Was the conversion successful?
|
---|
529 | my $was_successful=0;
|
---|
530 | if (-s "$output_filestem.html") {
|
---|
531 | # make sure we have some content other than header
|
---|
532 | open (HTML, "$output_filestem.html"); # what to do if fail?
|
---|
533 | my $line;
|
---|
534 | my $past_header=0;
|
---|
535 | while ($line=<HTML>) {
|
---|
536 |
|
---|
537 | if ($past_header == 0) {
|
---|
538 | if ($line =~ /<body>/) {$past_header=1;}
|
---|
539 | next;
|
---|
540 | }
|
---|
541 |
|
---|
542 | $line =~ s/<[^>]+>//g;
|
---|
543 | if ($line =~ /\w/ && $past_header) { # we found some content...
|
---|
544 | $was_successful=1;
|
---|
545 | last;
|
---|
546 | }
|
---|
547 | }
|
---|
548 | close HTML;
|
---|
549 | }
|
---|
550 |
|
---|
551 | if ($was_successful) {
|
---|
552 | &util::rm("$output_filestem.err")
|
---|
553 | if (-e "$output_filestem.err");
|
---|
554 | # insert the (modified) table of contents, if it exists.
|
---|
555 | if (-e "${output_filestem}_ToC.html") {
|
---|
556 | &util::mv("$output_filestem.html","$output_filestem.src");
|
---|
557 | my $open_failed=0;
|
---|
558 | open HTMLSRC, "$output_filestem.src" || ++$open_failed;
|
---|
559 | open TOC, "${output_filestem}_ToC.html" || ++$open_failed;
|
---|
560 | open HTML, ">$output_filestem.html" || ++$open_failed;
|
---|
561 |
|
---|
562 | if ($open_failed) {
|
---|
563 | close HTMLSRC;
|
---|
564 | close TOC;
|
---|
565 | close HTML;
|
---|
566 | &util::mv("$output_filestem.src","$output_filestem.html");
|
---|
567 | return 1;
|
---|
568 | }
|
---|
569 |
|
---|
570 | # print out header info from src html.
|
---|
571 | while (($_ = <HTMLSRC>) =~ /\w/) {
|
---|
572 | print HTML "$_";
|
---|
573 | }
|
---|
574 |
|
---|
575 | # print out table of contents, making links relative
|
---|
576 | <TOC>; <TOC>; # ignore first 2 lines
|
---|
577 | print HTML scalar(<TOC>); # line 3 = "<ol>\n"
|
---|
578 | my $line;
|
---|
579 | while ($line=<TOC>) {
|
---|
580 | $line =~ s@</body></html>$@@ ; # only last line has this
|
---|
581 | # make link relative
|
---|
582 | $line =~ s@href=\"[^\#]+@href=\"@;
|
---|
583 | print HTML $line;
|
---|
584 | }
|
---|
585 | close TOC;
|
---|
586 |
|
---|
587 | # rest of html src
|
---|
588 | while (<HTMLSRC>) {
|
---|
589 | print HTML $_;
|
---|
590 | }
|
---|
591 | close HTMLSRC;
|
---|
592 | close HTML;
|
---|
593 |
|
---|
594 | &util::rm("${output_filestem}_ToC.html");
|
---|
595 | &util::rm("${output_filestem}.src");
|
---|
596 | }
|
---|
597 | # we don't yet do anything with footnotes ($output_filestem_fn.html) :(
|
---|
598 | return 1; # success
|
---|
599 | }
|
---|
600 |
|
---|
601 | if (-e "$output_filestem.err") {
|
---|
602 | if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
|
---|
603 | {
|
---|
604 | print FAILLOG "Error - rtftohtml - couldn't extract text\n";
|
---|
605 | print FAILLOG " (rtf file might be too recent):\n";
|
---|
606 | open (ERRLOG, "$output_filestem.err");
|
---|
607 | while (<ERRLOG>) {print FAILLOG $_;}
|
---|
608 | close ERRLOG;
|
---|
609 | close FAILLOG;
|
---|
610 | }
|
---|
611 | &util::rm("$output_filestem.err");
|
---|
612 | }
|
---|
613 |
|
---|
614 | &util::rm("$output_filestem.html") if (-e "$output_filestem.html");
|
---|
615 |
|
---|
616 | return 0;
|
---|
617 | }
|
---|
618 |
|
---|
619 |
|
---|
620 | # Convert a pdf file to html with the pdftohtml command
|
---|
621 |
|
---|
622 | sub pdf_to_html {
|
---|
623 | my ($dirname, $input_filename, $output_filestem) = @_;
|
---|
624 |
|
---|
625 | $cmd = "";
|
---|
626 | if ($timeout) {$cmd = "ulimit -t $timeout;";}
|
---|
627 | $cmd .= "perl -S pdftohtml.pl ";
|
---|
628 | $cmd .= " \"$input_filename\" \"$output_filestem\"";
|
---|
629 |
|
---|
630 | if ($ENV{'GSDLOS'} !~ /^windows$/i || $is_winnt_2000) {
|
---|
631 | $cmd .= " > \"$output_filestem.out\" 2> \"$output_filestem.err\"";
|
---|
632 | } else {
|
---|
633 | $cmd .= " > \"$output_filestem.err\"";
|
---|
634 | }
|
---|
635 |
|
---|
636 | $!=0;
|
---|
637 |
|
---|
638 | my $retval=system($cmd);
|
---|
639 | if ($retval!=0)
|
---|
640 | {
|
---|
641 | print STDERR "Error executing pdftohtml.pl";
|
---|
642 | if ($!) {print STDERR ": $!";}
|
---|
643 | print STDERR "\n";
|
---|
644 | }
|
---|
645 |
|
---|
646 | # make sure the converter made something
|
---|
647 | if ($retval!=0 || ! -s "$output_filestem.html")
|
---|
648 | {
|
---|
649 | &util::rm("$output_filestem.out") if (-e "$output_filestem.out");
|
---|
650 | # print out the converter's std err, if any
|
---|
651 | if (-s "$output_filestem.err") {
|
---|
652 | open (ERRLOG, "$output_filestem.err") || die "$!";
|
---|
653 | print STDERR "pdftohtml error log:\n";
|
---|
654 | while (<ERRLOG>) {
|
---|
655 | print STDERR "$_";
|
---|
656 | }
|
---|
657 | close ERRLOG;
|
---|
658 | }
|
---|
659 | &util::rm("$output_filestem.html") if (-e "$output_filestem.html");
|
---|
660 | if (-e "$output_filestem.err") {
|
---|
661 | if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
|
---|
662 | {
|
---|
663 | open (ERRLOG, "$output_filestem.err");
|
---|
664 | while (<ERRLOG>) {print FAILLOG $_;}
|
---|
665 | close ERRLOG;
|
---|
666 | close FAILLOG;
|
---|
667 | }
|
---|
668 | &util::rm("$output_filestem.err");
|
---|
669 | }
|
---|
670 | return 0;
|
---|
671 | }
|
---|
672 |
|
---|
673 | &util::rm("$output_filestem.err") if (-e "$output_filestem.err");
|
---|
674 | &util::rm("$output_filestem.out") if (-e "$output_filestem.out");
|
---|
675 | return 1;
|
---|
676 | }
|
---|
677 |
|
---|
678 | # Convert a PDF file to text with the pdftotext command
|
---|
679 |
|
---|
680 | sub pdf_to_text {
|
---|
681 | my ($dirname, $input_filename, $output_filestem) = @_;
|
---|
682 |
|
---|
683 | my $cmd = "pdftotext \"$input_filename\" \"$output_filestem.text\"";
|
---|
684 |
|
---|
685 | if ($ENV{'GSDLOS'} !~ /^windows$/i) {
|
---|
686 | $cmd .= " > \"$output_filestem.out\" 2> \"$output_filestem.err\"";
|
---|
687 | } else {
|
---|
688 | $cmd .= " > \"$output_filestem.err\"";
|
---|
689 | }
|
---|
690 |
|
---|
691 | if (system($cmd)!=0)
|
---|
692 | {
|
---|
693 | print STDERR "Error executing $cmd: $!\n";
|
---|
694 | &util::rm("$output_filestem.text") if (-e "$output_filestem.text");
|
---|
695 | }
|
---|
696 |
|
---|
697 | # make sure there is some extracted text.
|
---|
698 | if (-e "$output_filestem.text") {
|
---|
699 | open (EXTR_TEXT, "$output_filestem.text") || warn "open: $!";
|
---|
700 | binmode(EXTR_TEXT); # just in case...
|
---|
701 | my $line="";
|
---|
702 | my $seen_text=0;
|
---|
703 | while (($seen_text==0) && ($line=<EXTR_TEXT>)) {
|
---|
704 | if ($line=~ /\w/) {$seen_text=1;}
|
---|
705 | }
|
---|
706 | close EXTR_TEXT;
|
---|
707 | if ($seen_text==0) { # no text was extracted
|
---|
708 | print STDERR "Error: pdftotext found no text\n";
|
---|
709 | &util::rm("$output_filestem.text");
|
---|
710 | }
|
---|
711 | }
|
---|
712 |
|
---|
713 | # make sure the converter made something
|
---|
714 | if (! -s "$output_filestem.text")
|
---|
715 | {
|
---|
716 | # print out the converters std err, if any
|
---|
717 | if (-s "$output_filestem.err") {
|
---|
718 | open (ERRLOG, "$output_filestem.err") || die "$!";
|
---|
719 | print STDERR "pdftotext error log:\n";
|
---|
720 | while (<ERRLOG>) {
|
---|
721 | print STDERR "$_";
|
---|
722 | }
|
---|
723 | close ERRLOG;
|
---|
724 | }
|
---|
725 | # does this converter create a .out file?
|
---|
726 | &util::rm("$output_filestem.out") if (-e "$output_filestem.out");
|
---|
727 | &util::rm("$output_filestem.text") if (-e "$output_filestem.text");
|
---|
728 | if (-e "$output_filestem.err") {
|
---|
729 | if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
|
---|
730 | {
|
---|
731 | open (ERRLOG,"$output_filestem.err");
|
---|
732 | while (<ERRLOG>) {print FAILLOG $_;}
|
---|
733 | close ERRLOG;
|
---|
734 | close FAILLOG;
|
---|
735 | }
|
---|
736 | &util::rm("$output_filestem.err");
|
---|
737 | }
|
---|
738 | return 0;
|
---|
739 | }
|
---|
740 | &util::rm("$output_filestem.err") if (-e "$output_filestem.err");
|
---|
741 | return 1;
|
---|
742 | }
|
---|
743 |
|
---|
744 | # Convert a PostScript document to text
|
---|
745 | # note - just using "ps2ascii" isn't good enough, as it
|
---|
746 | # returns 0 for a postscript interpreter error. ps2ascii is just
|
---|
747 | # a wrapper to "gs" anyway, so we use that cmd here.
|
---|
748 |
|
---|
749 | sub ps_to_text {
|
---|
750 | my ($input_filename, $output_filestem) = @_;
|
---|
751 |
|
---|
752 | my $error = "";
|
---|
753 |
|
---|
754 | # if we're on windows we'll fall straight through without attempting
|
---|
755 | # to use gs
|
---|
756 | if ($ENV{'GSDLOS'} =~ /^windows$/i) {
|
---|
757 | $error = "Windows does not support gs";
|
---|
758 |
|
---|
759 | } else {
|
---|
760 | my $cmd = "gs -q -dNODISPLAY -dNOBIND -dWRITESYSTEMDICT -dSIMPLE -c save ";
|
---|
761 | $cmd .= "-f ps2ascii.ps \"$input_filename\" -c quit > \"$output_filestem.text\"";
|
---|
762 | $cmd .= " 2> $output_filestem.err";
|
---|
763 | $!=0;
|
---|
764 |
|
---|
765 | my $retcode=system($cmd);
|
---|
766 | $retcode = $? >> 8; # see man perlfunc - system for this...
|
---|
767 | # if system returns -1 | 127 (couldn't start program), look at $! for message
|
---|
768 |
|
---|
769 | if ($retcode!=0) {if ($!) {$error=$!;} else {$error="couldn't run.\n";}}
|
---|
770 | elsif (! -e "$output_filestem.text") {
|
---|
771 | $error="did not create output file.\n";
|
---|
772 | }
|
---|
773 | else
|
---|
774 | { # make sure the interpreter didn't get an error. It is technically
|
---|
775 | # possible for the actual text to start with this, but....
|
---|
776 | open PSOUT, "$output_filestem.text";
|
---|
777 | if (<PSOUT> =~ /^Error: (.*)/) {
|
---|
778 | $error="interpreter error - \"$1\"";
|
---|
779 | }
|
---|
780 | close PSOUT;
|
---|
781 | }
|
---|
782 | }
|
---|
783 |
|
---|
784 | if ($error ne "")
|
---|
785 | {
|
---|
786 | print STDERR "Warning: Error executing gs: $error\n";
|
---|
787 | &util::rm("$output_filestem.text") if (-e "$output_filestem.text");
|
---|
788 |
|
---|
789 | if ("$faillogfile" ne "" && defined(open (FAILLOG, ">>$faillogfile")))
|
---|
790 | {
|
---|
791 | print FAILLOG "gs - $error\n";
|
---|
792 | if (-e "$output_filestem.err") {
|
---|
793 | open(ERRLOG, "$output_filestem.err");
|
---|
794 | while (<ERRLOG>) {print FAILLOG $_;}
|
---|
795 | close ERRLOG;
|
---|
796 | }
|
---|
797 | close FAILLOG;
|
---|
798 | }
|
---|
799 | &util::rm("$output_filestem.err") if (-e "$output_filestem.err");
|
---|
800 |
|
---|
801 |
|
---|
802 | # Fine then. We'll just do a lousy job by ourselves...
|
---|
803 | # Based on 5-line regexp sed script found at:
|
---|
804 | # http://snark.ptc.spbu.ru/mail-archives/lout/brown/msg00003.html
|
---|
805 | #
|
---|
806 | print STDERR "Stripping text from postscript\n";
|
---|
807 | my $errorcode=0;
|
---|
808 | open (IN, "$input_filename")
|
---|
809 | || ($errorcode=1, warn "Couldn't read file: $!");
|
---|
810 | open (OUT, ">$output_filestem.text")
|
---|
811 | || ($errorcode=1, warn "Couldn't write file: $!");
|
---|
812 | if ($errorcode) {print STDERR "errors\n";return 0;}
|
---|
813 |
|
---|
814 | my $text=""; # this is for whole .ps file...
|
---|
815 | $text = join('', <IN>); # see man perlport, under "System Resources"
|
---|
816 | close IN;
|
---|
817 |
|
---|
818 | # Make sure this is a ps file...
|
---|
819 | if ($text !~ /^%!/) {
|
---|
820 | print STDERR "Bad postscript header: not '%!'\n";
|
---|
821 | if ($faillogfile ne "" && defined(open(FAILLOG, ">>$faillogfile")))
|
---|
822 | {
|
---|
823 | print FAILLOG "Bad postscript header: not '%!'\n";
|
---|
824 | close FAILLOG;
|
---|
825 | }
|
---|
826 | return 0;
|
---|
827 | }
|
---|
828 |
|
---|
829 | # if ps has Page data, then use it to delete all stuff before it.
|
---|
830 | $text =~ s/^.*?%%Page:.*?\n//s; # treat string as single line
|
---|
831 |
|
---|
832 | # remove all leading non-data stuff
|
---|
833 | $text =~ s/^.*?\(//s;
|
---|
834 |
|
---|
835 | # remove all newline chars for easier processing
|
---|
836 | $text =~ s/\n//g;
|
---|
837 |
|
---|
838 | # Big assumption here - assume that if any co-ordinates are
|
---|
839 | # given, then we are at the end of a sentence.
|
---|
840 | $text =~ s/\)-?\d+\ -?\d+/\) \(\n\)/g;
|
---|
841 |
|
---|
842 | # special characters--
|
---|
843 | $text =~ s/\(\|\)/\(\ - \)/g; # j -> em-dash?
|
---|
844 |
|
---|
845 | # ? ps text formatting (eg italics?) ?
|
---|
846 | $text =~ s/Fn\(f\)/\(\{\)/g; # f -> {
|
---|
847 | $text =~ s/Fn\(g\)/\(\}\)/g; # g -> }
|
---|
848 | $text =~ s/Fn\(j\)/\(\|\)/g; # j -> |
|
---|
849 | # default - remove the rest
|
---|
850 | $text =~ s/\ ?F.\((.+?)\)/\($1\)/g;
|
---|
851 |
|
---|
852 | # attempt to add whitespace between words...
|
---|
853 | # this is based purely on observation, and may be completely wrong...
|
---|
854 | $text =~ s/([^F])[defghijkuy]\(/$1 \( /g;
|
---|
855 | # eg I notice "b(" is sometimes NOT a space if preceded by a
|
---|
856 | # negative number.
|
---|
857 | $text =~ s/\)\d+ ?b\(/\) \( /g;
|
---|
858 |
|
---|
859 | # change quoted braces to brackets
|
---|
860 | $text =~ s/([^\\])\\\(/$1\{/g;
|
---|
861 | $text =~ s/([^\\])\\\)/$1\}/g ;
|
---|
862 |
|
---|
863 | # remove everything that is not between braces
|
---|
864 | $text =~ s/\)([^\(\)])+?\(//sg ;
|
---|
865 |
|
---|
866 | # remove any Trailer eof stuff.
|
---|
867 | $text =~ s/\)[^\)]*$//sg;
|
---|
868 |
|
---|
869 | ### ligatures have special characters...
|
---|
870 | $text =~ s/\\013/ff/g;
|
---|
871 | $text =~ s/\\014/fi/g;
|
---|
872 | $text =~ s/\\015/fl/g;
|
---|
873 | $text =~ s/\\016/ffi/g;
|
---|
874 | $text =~ s/\\214/fi/g;
|
---|
875 | $text =~ s/\\215/fl/g;
|
---|
876 | $text =~ s/\\017/\n\* /g; # asterisk?
|
---|
877 | $text =~ s/\\023/\023/g; # e acute ('e)
|
---|
878 | $text =~ s/\\177/\252/g; # u"
|
---|
879 | # $text =~ s/ ?? /\344/g; # a"
|
---|
880 |
|
---|
881 | print OUT "$text";
|
---|
882 | close OUT;
|
---|
883 | }
|
---|
884 | # wrap the text - use a minimum length. ie, first space after this length.
|
---|
885 | my $wrap_length=72;
|
---|
886 | &util::mv("$output_filestem.text", "$output_filestem.text.tmp");
|
---|
887 | open INFILE, "$output_filestem.text.tmp" ||
|
---|
888 | die "Couldn't open file: $!";
|
---|
889 | open OUTFILE, ">$output_filestem.text" ||
|
---|
890 | die "Couldn't open file for writing: $!";
|
---|
891 | my $line="";
|
---|
892 | while ($line=<INFILE>) {
|
---|
893 | while (length($line)>0) {
|
---|
894 | if (length($line)>$wrap_length) {
|
---|
895 | $line =~ s/^(.{$wrap_length}[^\s]*)\s*//;
|
---|
896 | print OUTFILE "$1\n";
|
---|
897 | } else {
|
---|
898 | print OUTFILE "$line";
|
---|
899 | $line="";
|
---|
900 | }
|
---|
901 | }
|
---|
902 | }
|
---|
903 | close INFILE;
|
---|
904 | close OUTFILE;
|
---|
905 | &util::rm("$output_filestem.text.tmp");
|
---|
906 |
|
---|
907 | &util::rm("$output_filestem.err") if (-e "$output_filestem.err");
|
---|
908 | return 1;
|
---|
909 | }
|
---|
910 |
|
---|
911 |
|
---|
912 | # Convert any file to HTML with a crude perl implementation of the
|
---|
913 | # UNIX strings command.
|
---|
914 |
|
---|
915 | sub any_to_html {
|
---|
916 | ($input_filename, $output_filestem) = @_;
|
---|
917 |
|
---|
918 | # First generate a text file
|
---|
919 | return 0 unless (&any_to_text($input_filename, $output_filestem));
|
---|
920 |
|
---|
921 | # create an HTML file from the text file
|
---|
922 | open(TEXT, "<$output_filestem.text");
|
---|
923 | open(HTML, ">$output_filestem.html");
|
---|
924 |
|
---|
925 | print HTML "<html><head>\n";
|
---|
926 | print HTML "<META HTTP-EQUIV=\"Content-Type\" CONTENT=\"text/html\">\n";
|
---|
927 | print HTML "<META NAME=\"GENERATOR\" CONTENT=\"Greenstone any_to_html\">\n";
|
---|
928 | print HTML "</head><body>\n\n";
|
---|
929 |
|
---|
930 | my $line;
|
---|
931 | while ($line=<TEXT>) {
|
---|
932 | $line =~ s/</</g;
|
---|
933 | $line =~ s/>/>/g;
|
---|
934 | if ($line =~ /^\s*$/) {
|
---|
935 | print HTML "<p>";
|
---|
936 | } else {
|
---|
937 | print HTML "<br> ", $line;
|
---|
938 | }
|
---|
939 | }
|
---|
940 | print HTML "\n</body></html>\n";
|
---|
941 |
|
---|
942 | close HTML;
|
---|
943 | close TEXT;
|
---|
944 |
|
---|
945 | &util::rm("$output_filestem.text") if (-e "$output_filestem.text");
|
---|
946 | return 1;
|
---|
947 | }
|
---|
948 |
|
---|
949 | # Convert any file to TEXT with a crude perl implementation of the
|
---|
950 | # UNIX strings command.
|
---|
951 | # Note - this assumes ascii charsets :( (jrm21)
|
---|
952 |
|
---|
953 | sub any_to_text {
|
---|
954 | ($input_filename, $output_filestem) = @_;
|
---|
955 |
|
---|
956 | open(IN, "<$input_filename") || return 0;
|
---|
957 | binmode(IN);
|
---|
958 | open(OUT, ">$output_filestem.text") || return 0;
|
---|
959 |
|
---|
960 | my ($line);
|
---|
961 | my $output_line_count = 0;
|
---|
962 | while (<IN>) {
|
---|
963 | $line = $_;
|
---|
964 |
|
---|
965 | # delete anything that isn't a printable character
|
---|
966 | $line =~ s/[^\040-\176]+/\n/sg;
|
---|
967 |
|
---|
968 | # delete any string less than 10 characters long
|
---|
969 | $line =~ s/^.{0,9}$/\n/mg;
|
---|
970 | while ($line =~ /^.{1,9}$/m) {
|
---|
971 | $line =~ s/^.{0,9}$/\n/mg;
|
---|
972 | $line =~ s/\n+/\n/sg;
|
---|
973 | }
|
---|
974 |
|
---|
975 | # remove extraneous whitespace
|
---|
976 | $line =~ s/\n+/\n/gs;
|
---|
977 | $line =~ s/^\n//gs;
|
---|
978 |
|
---|
979 | # output whatever is left
|
---|
980 | if ($line =~ /[^\n ]/) {
|
---|
981 | print OUT $line;
|
---|
982 | ++$output_line_count;
|
---|
983 | }
|
---|
984 | }
|
---|
985 |
|
---|
986 | close OUT;
|
---|
987 | close IN;
|
---|
988 |
|
---|
989 | if ($output_line_count) { # try to protect against binary only formats
|
---|
990 | return 1;
|
---|
991 | }
|
---|
992 |
|
---|
993 | &util::rm("$output_filestem.text");
|
---|
994 | return 0;
|
---|
995 |
|
---|
996 | }
|
---|