source: main/trunk/greenstone2/bin/script/gsConvert.pl@ 32224

Last change on this file since 32224 was 32224, checked in by ak19, 6 years ago

Adding PDF to text support for Windows using Xpdf's pdftotext tool. Previously PDFPlugin would override the output mode to HTML on Windows if txt output mode was selected.
  • Property svn:executable set to *
  • Property svn:keywords set to Author Date Id Revision
File size: 43.6 KB
Line 
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
46BEGIN {
47 die "GSDLHOME not set\n" unless defined $ENV{'GSDLHOME'};
48 unshift (@INC, "$ENV{'GSDLHOME'}/perllib");
49}
50
51use strict;
52
53use parsargv;
54use util;
55use FileUtils;
56use Cwd;
57
58# Are we running on WinNT or Win2000 (or later)?
59my $is_winnt_2000=eval {require Win32; return (Win32::IsWinNT()); return 0;};
60if (!defined($is_winnt_2000)) {$is_winnt_2000=0;}
61
62my $use_strings;
63my $pdf_complex;
64my $pdf_nohidden;
65my $pdf_zoom;
66my $pdf_ignore_images;
67my $pdf_allow_images_only;
68my $windows_scripting;
69my $enc;
70
71sub print_usage
72{
73 print STDERR "\n";
74 print STDERR "gsConvert.pl: Converts documents in a range of formats to html\n";
75 print STDERR " or text using third-party programs.\n\n";
76 print STDERR " usage: $0 [options] filename\n";
77 print STDERR " options:\n\t-type\tdoc|dot|pdf|ps|ppt|rtf|xls\t(input file type)\n";
78 print STDERR "\t-errlog\t<filename>\t(append err messages)\n";
79 print STDERR "\t-output\tauto|html|text|pagedimg_jpg|pagedimg_gif|pagedimg_png\t(output file type)\n";
80 print STDERR "\t-timeout\t<max cpu seconds>\t(ulimit on unix systems)\n";
81 print STDERR "\t-use_strings\tuse strings to extract text if conversion fails\n";
82 print STDERR "\t-windows_scripting\tuse windows VB script (if available) to convert Microsoft Word and PPT documents\n";
83 print STDERR "\t-pdf_complex\tuse complex output when converting PDF to HTML\n";
84 print STDERR "\t-pdf_nohidden\tDon't attempt to extract hidden text from PDF files\n";
85 print STDERR "\t-pdf_ignore_images\tdon't attempt to extract images when\n";
86 print STDERR "\t\tconverting PDF to HTML\n";
87 print STDERR "\t-pdf_allow_images_only\tallow images only (continue even if no text is present when converting to HTML)\n";
88 print STDERR "\t-pdf_zoom\tfactor by which to zoom PDF (only useful if\n";
89 print STDERR "\t\t-pdf_complex is set\n";
90 exit(1);
91}
92
93my $faillogfile="";
94my $timeout=0;
95my $verbosity=0;
96
97sub main
98{
99 my (@ARGV) = @_;
100 my ($input_type,$output_type,$verbose);
101
102 # Dynamically figure out what the --type option can support, based on whether -windows_scripting
103 # is in use or not
104 my $default_type_re = "(doc|dot|pdf|ps|ppt|rtf|xls)";
105 #my $enhanced_type_re = "(docx?|dot|pdf|ps|pptx?|rtf|xlsx?)";
106 #my $enhanced_type_re = "(docx?|dot|pdf|ps|pptx?|rtf|xlsx?)";
107 # Currently only have VBA for Word and PPT(but no XLS)
108 my $enhanced_type_re = "(docx?|dot|pdf|ps|pptx?|rtf|xls)";
109
110 my $type_re = $default_type_re;
111
112 foreach my $a (@ARGV) {
113 if ($a =~ m/^windows_scripting$/i) {
114 $type_re = $enhanced_type_re;
115 }
116 }
117
118 # read command-line arguments
119 if (!parsargv::parse(\@ARGV,
120 "type/$type_re/", \$input_type,
121 '/errlog/.*/', \$faillogfile,
122 'output/(auto|html|text|pagedimg).*/', \$output_type,
123 'timeout/\d+/0',\$timeout,
124 'verbose/\d+/0', \$verbose,
125 'windows_scripting',\$windows_scripting,
126 'use_strings', \$use_strings,
127 'pdf_complex', \$pdf_complex,
128 'pdf_ignore_images', \$pdf_ignore_images,
129 'pdf_allow_images_only', \$pdf_allow_images_only,
130 'pdf_nohidden', \$pdf_nohidden,
131 'pdf_zoom/\d+/2', \$pdf_zoom
132 ))
133 {
134 print_usage();
135 }
136
137 $verbosity=$verbose if defined $verbose;
138
139 # Make sure the input file exists and can be opened for reading
140 if (scalar(@ARGV!=1)) {
141 print_usage();
142 }
143
144 my $input_filename = $ARGV[0];
145 if (!-r $input_filename) {
146 print STDERR "Error: unable to open $input_filename for reading\n";
147 exit(1);
148 }
149
150 # Deduce filenames
151 my ($tailname,$dirname,$suffix)
152 = File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");
153 my $output_filestem = &FileUtils::filenameConcatenate($dirname, "$tailname");
154
155 if ($input_type eq "")
156 {
157 $input_type = lc (substr($suffix,1,length($suffix)-1));
158 }
159
160 # Change to temporary working directory
161 my $stored_dir = cwd();
162 chdir ($dirname) || die "Unable to change to directory $dirname";
163
164 # Select convert utility
165 if (!defined $input_type) {
166 print STDERR "Error: No filename extension or input type defined\n";
167 exit(1);
168 }
169 elsif ($input_type =~ m/^docx?$/ || $input_type eq "dot") {
170 print &convertDOC($input_filename, $output_filestem, $output_type);
171 print "\n";
172 }
173 elsif ($input_type eq "rtf") {
174 print &convertRTF($input_filename, $output_filestem, $output_type);
175 print "\n";
176 }
177 elsif ($input_type eq "pdf") {
178 print &convertPDF($dirname, $input_filename, $output_filestem, $output_type);
179 print "\n";
180 }
181 elsif ($input_type eq "ps") {
182 print &convertPS($dirname, $input_filename, $output_filestem, $output_type);
183 print "\n";
184 }
185 elsif ($input_type =~ m/pptx?$/) {
186 print &convertPPT($input_filename, $output_filestem, $output_type);
187 print "\n";
188 }
189 elsif ($input_type =~ m/xlsx?$/) {
190 print &convertXLS($input_filename, $output_filestem, $output_type);
191 print "\n";
192 }
193 else {
194 print STDERR "Error: Unable to convert type '$input_type'\n";
195 exit(1);
196 }
197
198 # restore to original working directory
199 chdir ($stored_dir) || die "Unable to return to directory $stored_dir";
200
201}
202
203&main(@ARGV);
204
205
206
207# Document-type conversion functions
208#
209# The following functions attempt to convert documents from their
210# input type to the specified output type. If no output type was
211# given, then they first attempt HTML, and then TEXT.
212#
213# Each returns the output type ("html" or "text") or "fail" if no
214# conversion is possible.
215
216# Convert a Microsoft word document
217
218sub convertDOC {
219 my ($input_filename, $output_filestem, $output_type) = @_;
220
221 # Many .doc files are not in fact word documents!
222 my $realtype = &find_docfile_type($input_filename);
223
224 if ($realtype eq "word6" || $realtype eq "word7"
225 || $realtype eq "word8" || $realtype eq "docx") {
226 return &convertWord678($input_filename, $output_filestem, $output_type);
227 } elsif ($realtype eq "rtf") {
228 return &convertRTF($input_filename, $output_filestem, $output_type);
229 } else {
230 return &convertAnything($input_filename, $output_filestem, $output_type);
231 }
232}
233
234# Convert a Microsoft word 6/7/8 document
235
236sub convertWord678 {
237 my ($input_filename, $output_filestem, $output_type) = @_;
238
239 my $success = 0;
240 if (!$output_type || ($output_type =~ m/html/i)){
241 if ($windows_scripting) {
242 $success = &native_doc_to_html($input_filename, $output_filestem);
243 }
244 else {
245 $success = &doc_to_html($input_filename, $output_filestem);
246 }
247 if ($success) {
248 return "html";
249 }
250 }
251 return &convertAnything($input_filename, $output_filestem, $output_type);
252}
253
254
255# Convert a Rich Text Format (RTF) file
256
257sub convertRTF {
258 my ($input_filename, $output_filestem, $output_type) = @_;
259
260 my $success = 0;
261
262 # Attempt specialised conversion to HTML
263 if (!$output_type || ($output_type =~ m/html/i)) {
264
265 if ($windows_scripting) {
266 $success = &native_doc_to_html($input_filename, $output_filestem);
267 }
268 else {
269 $success = &rtf_to_html($input_filename, $output_filestem);
270 }
271 if ($success) {
272 return "html";
273 }
274 }
275
276# rtf is so ugly that's it's not worth running strings over.
277# One day I'll write some quick'n'dirty regexps to try to extract text - jrm21
278# return &convertAnything($input_filename, $output_filestem, $output_type);
279 return "fail";
280}
281
282
283# Convert an unidentified file
284
285sub convertAnything {
286 my ($input_filename, $output_filestem, $output_type) = @_;
287
288 my $success = 0;
289
290 # Attempt simple conversion to HTML
291 if (!$output_type || ($output_type =~ m/html/i)) {
292 $success = &any_to_html($input_filename, $output_filestem);
293 if ($success) {
294 return "html";
295 }
296 }
297
298 # Convert to text
299 if (!$output_type || ($output_type =~ m/text/i)) {
300 $success = &any_to_text($input_filename, $output_filestem);
301 if ($success) {
302 return "text";
303 }
304 }
305 return "fail";
306}
307
308
309
310# Convert an Adobe PDF document
311
312sub convertPDF {
313 my ($dirname, $input_filename, $output_filestem, $output_type) = @_;
314
315 my $success = 0;
316 $output_type =~ s/.*\-(.*)/$1/i;
317 # Attempt coversion to Image
318 if ($output_type =~ m/jp?g|gif|png/i) {
319 $success = &pdfps_to_img($dirname, $input_filename, $output_filestem, $output_type);
320 if ($success){
321 return "item";
322 }
323 }
324
325 # Attempt conversion to HTML
326 # Uses the old pdftohtml that doesn't work for newer PDF versions
327 if ($output_type =~ m/^html/i) {
328 #if (!$output_type || ($output_type =~ m/^html/i)) {
329 $success = &pdf_to_html($dirname, $input_filename, $output_filestem);
330 if ($success) {
331 return "html";
332 }
333 }
334
335 # Attempt conversion to (paged) HTML using the newer pdftohtml of Xpdftools. This
336 # will be the new default for PDFs when output_type for PDF docs is not specified
337 # (once our use of xpdftools' pdftohtml has been implemented on win and mac).
338 #if ($output_type =~ m/paged_html/i) {
339 if (!$output_type || ($output_type =~ m/paged_html/i)) {
340 $success = &xpdf_to_html($dirname, $input_filename, $output_filestem);
341 if ($success) {
342 return "paged_html";
343 }
344 }
345
346 # Attempt conversion to TEXT
347 if (!$output_type || ($output_type =~ m/text/i)) {
348 if ($ENV{'GSDLOS'} =~ m/^windows$/i) { # we now have pdf to text support for windows by using xpdf tools
349 $success = &xpdf_to_text($dirname, $input_filename, $output_filestem);
350 } else {
351 $success = &pdf_to_text($dirname, $input_filename, $output_filestem);
352 }
353 if ($success) {
354 return "text";
355 }
356 }
357
358 return "fail";
359
360}
361
362
363# Convert an Adobe PostScript document
364
365sub convertPS {
366 my ($dirname,$input_filename, $output_filestem, $output_type) = @_;
367
368 my $success = 0;
369 $output_type =~ s/.*\-(.*)/$1/i;
370 # Attempt coversion to Image
371 if ($output_type =~ m/jp?g|gif|png/i) {
372 $success = &pdfps_to_img($dirname, $input_filename, $output_filestem, $output_type);
373 if ($success){
374 return "item";
375 }
376 }
377
378 # Attempt conversion to TEXT
379 if (!$output_type || ($output_type =~ m/text/i)) {
380 $success = &ps_to_text($input_filename, $output_filestem);
381 if ($success) {
382 return "text";
383 }
384 }
385 return "fail";
386}
387
388
389sub convertPPT {
390 my ($input_filename, $output_filestem, $output_type) = @_;
391 my $success = 0;
392
393 my $ppt_convert_type = "";
394
395 #if (!$output_type || $windows_scripting || ($output_type !~ m/html/i) || ($output_type !~ m/text/i)){
396 if ($windows_scripting && ($output_type !~ m/html/i) && ($output_type !~ m/text/i)){
397 if ($output_type =~ m/gif/i) {
398 $ppt_convert_type = "-g";
399 } elsif ($output_type =~ m/jp?g/i){
400 $ppt_convert_type = "-j";
401 } elsif ($output_type =~ m/png/i){
402 $ppt_convert_type = "-p";
403 }
404 my $vbScript = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'}, "bin",
405 $ENV{'GSDLOS'}, "pptextract");
406 $vbScript = "CScript //Nologo \"".$vbScript.".vbs\"" if ($ENV{'GSDLOS'} =~ m/^windows$/i); # now we use the .vbs VBScript
407 # $vbScript = "pptextract" if ($ENV{'GSDLOS'} =~ m/^windows$/i); # back when the pptextract.exe VB executable was used
408
409 my $cmd = "";
410 if ($timeout) {$cmd = "ulimit -t $timeout;";}
411 # if the converting directory already exists
412 if (-d $output_filestem) {
413 print STDERR "**The conversion directory already exists\n";
414 return "item";
415 } else {
416 $cmd .= "$vbScript $ppt_convert_type \"$input_filename\" \"$output_filestem\"";
417 $cmd .= " 2>\"$output_filestem.err\""
418 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000);
419
420 if (system($cmd) !=0) {
421 print STDERR "Powerpoint VB Scripting convert failed\n";
422 } else {
423 return "item";
424 }
425 }
426 } elsif (!$output_type || ($output_type =~ m/html/i)) {
427 # Attempt conversion to HTML
428 #if (!$output_type || ($output_type =~ m/html/i)) {
429 # formulate the command
430 my $cmd = "";
431 my $full_perl_path = &util::get_perl_exec();
432 $cmd .= "\"$full_perl_path\" -S ppttohtml.pl ";
433 $cmd .= " \"$input_filename\" \"$output_filestem.html\"";
434 $cmd .= " 2>\"$output_filestem.err\""
435 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000);
436
437 # execute the command
438 $!=0;
439 if (system($cmd)!=0)
440 {
441 print STDERR "Powerpoint 95/97 converter failed $!\n";
442 } else {
443 return "html";
444 }
445 }
446
447 $success = &any_to_text($input_filename, $output_filestem);
448 if ($success) {
449 return "text";
450 }
451
452 return "fail";
453}
454
455
456sub convertXLS {
457 my ($input_filename, $output_filestem, $output_type) = @_;
458
459 my $success = 0;
460
461 # Attempt conversion to HTML
462 if (!$output_type || ($output_type =~ m/html/i)) {
463 # formulate the command
464 my $cmd = "";
465 my $full_perl_path = &util::get_perl_exec();
466 $cmd .= "\"$full_perl_path\" -S xlstohtml.pl ";
467 $cmd .= " \"$input_filename\" \"$output_filestem.html\"";
468 $cmd .= " 2>\"$output_filestem.err\""
469 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000);
470
471
472 # execute the command
473 $!=0;
474 if (system($cmd)!=0)
475 {
476 print STDERR "Excel 95/97 converter failed $!\n";
477 } else {
478 return "html";
479 }
480 }
481
482 $success = &any_to_text($input_filename, $output_filestem);
483 if ($success) {
484 return "text";
485 }
486
487 return "fail";
488}
489
490
491
492# Find the real type of a .doc file
493#
494# We seem to have a lot of files with a .doc extension that are .rtf
495# files or Word 5 files. This function attempts to tell the difference.
496sub find_docfile_type {
497 my ($input_filename) = @_;
498
499 if (($windows_scripting) && ($input_filename =~ m/\.docx$/)) {
500 return "docx";
501 }
502
503 open(CHK, "<$input_filename");
504 binmode(CHK);
505 my $line = "";
506 my $first = 1;
507
508 while (<CHK>) {
509
510 $line = $_;
511
512 if ($first) {
513 # check to see if this is an rtf file
514 if ($line =~ m/^\{\\rtf/) {
515 close(CHK);
516 return "rtf";
517 }
518 $first = 0;
519 }
520
521 # is this is a word 6/7/8 document?
522 if ($line =~ m/Word\.Document\.([678])/) {
523 close(CHK);
524
525 return "word$1";
526 }
527
528 }
529
530 return "unknown";
531}
532
533
534# Specific type-to-type conversions
535#
536# Each of the following functions attempts to convert a document from
537# a specific format to another. If they succeed they return 1 and leave
538# the output document(s) in the appropriate place; if they fail they
539# return 0 and delete any working files.
540
541
542# Attempt to convert a word document to html with the wv program
543sub doc_to_html {
544 my ($input_filename, $output_filestem) = @_;
545
546 my $wvware_status = 0;
547
548 # need to ensure that the path to perl is quoted (in case there's spaces in it)
549 my $launch_cmd = "\"".&util::get_perl_exec()."\" -S wvware.pl \"$input_filename\" \"$output_filestem\" \"$faillogfile\" $verbosity $timeout";
550
551# print STDERR "***** wvware launch cmd = $launch_cmd\n";
552
553 $wvware_status = system($launch_cmd)/256;
554 return $wvware_status;
555}
556
557# Attempt to convert a word document to html with the word2html scripting program
558sub native_doc_to_html {
559 my ($input_filename, $output_filestem) = @_;
560
561 # build up the path to the doc-to-html conversion tool we're going to use
562 my $vbScript = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'}, "bin", $ENV{'GSDLOS'});
563
564 if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
565 # if windows scripting with docx input, use new VBscript to get the local Word install (if
566 # any) to do the conversion, since docX can't be processed by word2html's windows_scripting
567
568 if($input_filename =~ m/docx$/i) { # need to use full path to docx2html script,
569 # else script launch fails when there are error msgs
570 $vbScript = &FileUtils::filenameConcatenate($vbScript, "docx2html.vbs");
571 $vbScript = "CScript //Nologo \"$vbScript\""; # launch with CScript for error output in STDERR
572 # //Nologo flag avoids Microsoft's opening/logo msgs
573 print STDERR "About to use windows scripting to process docx file $input_filename.\n";
574 print STDERR " This may take some time. Please wait...\n";
575 }
576 else { # old doc versions. use the usual VB executable word2html for the
577 # conversion. Doesn't need full path, since bin\windows is on PATH
578 $vbScript = "word2html"; #$vbScript = "\"".&FileUtils::filenameConcatenate($vbScript, "word2html")."\"";
579 }
580 }
581 else { # not windows
582 $vbScript = "\"".&FileUtils::filenameConcatenate($vbScript, "word2html")."\"";
583 }
584
585 if (-e "$output_filestem.html") {
586 print STDERR " The conversion file:\n";
587 print STDERR " $output_filestem.html\n";
588 print STDERR " ... already exists. Skipping\n";
589 return 1;
590 }
591
592 my $cmd = "";
593 if ($timeout) {$cmd = "ulimit -t $timeout;";}
594 #$cmd .= "$vbScript \"$input_filename\" \"$output_filestem.html\"";
595 #$cmd .= "$vbScript $input_filename $output_filestem.html";
596 $cmd .= "$vbScript \"$input_filename\" \"$output_filestem.html\"";
597
598 # redirecting STDERR
599
600 $cmd .= " 2> \"$output_filestem.err\""
601 if ($ENV {'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000);
602 #print STDERR "@@@@@@@@@ cmd=$cmd\n";
603
604 # execute the command
605 $!=0;
606 if (system($cmd)!=0)
607 {
608 print STDERR "Error executing $vbScript converter:$!\n";
609 if (-s "$output_filestem.err") {
610 open (ERRFILE, "<$output_filestem.err");
611
612 my $write_to_fail_log=0;
613 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
614 {$write_to_fail_log=1;}
615
616 my $line;
617 while ($line=<ERRFILE>) {
618 if ($line =~ m/\w/) {
619 print STDERR "$line";
620 print FAILLOG "$line" if ($write_to_fail_log);
621 }
622 if ($line !~ m/startup error/) {next;}
623 print STDERR " (given an invalid .DOC file?)\n";
624 print FAILLOG " (given an invalid .DOC file?)\n"
625 if ($write_to_fail_log);
626
627 } # while ERRFILE
628 close FAILLOG if ($write_to_fail_log);
629 }
630 return 0; # we can try any_to_text
631 }
632
633 # Was the conversion successful?
634 if (-s "$output_filestem.html") {
635 open(TMP, "$output_filestem.html");
636 my $line = <TMP>;
637 close(TMP);
638 if ($line && $line =~ m/html/i) {
639 &FileUtils::removeFiles("$output_filestem.err") if -e "$output_filestem.err";
640 return 1;
641 }
642 }
643
644 # If here, an error of some sort occurred
645 &FileUtils::removeFiles("$output_filestem.html") if -e "$output_filestem.html";
646 if (-e "$output_filestem.err") {
647 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile"))) {
648 open (ERRLOG,"$output_filestem.err");
649 while (<ERRLOG>) {print FAILLOG $_;}
650 close FAILLOG;
651 close ERRLOG;
652 }
653 &FileUtils::removeFiles("$output_filestem.err");
654 }
655 return 0;
656}
657
658# Attempt to convert an RTF document to html with rtftohtml
659sub rtf_to_html {
660 my ($input_filename, $output_filestem) = @_;
661
662 # formulate the command
663 my $cmd = "";
664 if ($timeout) {$cmd = "ulimit -t $timeout;";}
665 $cmd .= "rtftohtml";
666 #$cmd .= "rtf-converter";
667
668 $cmd .= " -o \"$output_filestem.html\" \"$input_filename\"";
669
670 $cmd .= " 2>\"$output_filestem.err\""
671 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000);
672
673
674 # execute the command
675 $!=0;
676 if (system($cmd)!=0)
677 {
678 print STDERR "Error executing rtf converter $!\n";
679 # don't currently bother printing out error log...
680 # keep going, in case it still created an HTML file...
681 }
682
683 # Was the conversion successful?
684 my $was_successful=0;
685 if (-s "$output_filestem.html") {
686 # make sure we have some content other than header
687 open (HTML, "$output_filestem.html"); # what to do if fail?
688 my $line;
689 my $past_header=0;
690 while ($line=<HTML>) {
691
692 if ($past_header == 0) {
693 if ($line =~ m/<body>/) {$past_header=1;}
694 next;
695 }
696
697 $line =~ s/<[^>]+>//g;
698 if ($line =~ m/\w/ && $past_header) { # we found some content...
699 $was_successful=1;
700 last;
701 }
702 }
703 close HTML;
704 }
705
706 if ($was_successful) {
707 &FileUtils::removeFiles("$output_filestem.err")
708 if (-e "$output_filestem.err");
709 # insert the (modified) table of contents, if it exists.
710 if (-e "${output_filestem}_ToC.html") {
711 &FileUtils::moveFiles("$output_filestem.html","$output_filestem.src");
712 my $open_failed=0;
713 open HTMLSRC, "$output_filestem.src" || ++$open_failed;
714 open TOC, "${output_filestem}_ToC.html" || ++$open_failed;
715 open HTML, ">$output_filestem.html" || ++$open_failed;
716
717 if ($open_failed) {
718 close HTMLSRC;
719 close TOC;
720 close HTML;
721 &FileUtils::moveFiles("$output_filestem.src","$output_filestem.html");
722 return 1;
723 }
724
725 # print out header info from src html.
726 while (defined($_ = <HTMLSRC>) && $_ =~ m/\w/) {
727 print HTML "$_";
728 }
729
730 # print out table of contents, making links relative
731 <TOC>; <TOC>; # ignore first 2 lines
732 print HTML scalar(<TOC>); # line 3 = "<ol>\n"
733 my $line;
734 while ($line=<TOC>) {
735 $line =~ s@</body></html>$@@i ; # only last line has this
736 # make link relative
737 $line =~ s@href=\"[^\#]+@href=\"@i;
738 print HTML $line;
739 }
740 close TOC;
741
742 # rest of html src
743 while (<HTMLSRC>) {
744 print HTML $_;
745 }
746 close HTMLSRC;
747 close HTML;
748
749 &FileUtils::removeFiles("${output_filestem}_ToC.html");
750 &FileUtils::removeFiles("${output_filestem}.src");
751 }
752 # we don't yet do anything with footnotes ($output_filestem_fn.html) :(
753 return 1; # success
754 }
755
756 if (-e "$output_filestem.err") {
757 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
758 {
759 print FAILLOG "Error - rtftohtml - couldn't extract text\n";
760 #print FAILLOG "Error - rtf-converter - couldn't extract text\n";
761 print FAILLOG " (rtf file might be too recent):\n";
762 open (ERRLOG, "$output_filestem.err");
763 while (<ERRLOG>) {print FAILLOG $_;}
764 close ERRLOG;
765 close FAILLOG;
766 }
767 &FileUtils::removeFiles("$output_filestem.err");
768 }
769
770 &FileUtils::removeFiles("$output_filestem.html") if (-e "$output_filestem.html");
771
772 return 0;
773}
774
775
776# Convert a pdf file to html with the old pdftohtml command
777# which only works for older PDF versions
778sub pdf_to_html {
779 my ($dirname, $input_filename, $output_filestem) = @_;
780
781 my $cmd = "";
782 if ($timeout) {$cmd = "ulimit -t $timeout;";}
783 my $full_perl_path = &util::get_perl_exec();
784 $cmd .= "\"$full_perl_path\" -S pdftohtml.pl -zoom $pdf_zoom";
785 $cmd .= " -c" if ($pdf_complex);
786 $cmd .= " -i" if ($pdf_ignore_images);
787 $cmd .= " -a" if ($pdf_allow_images_only);
788 $cmd .= " -hidden" unless ($pdf_nohidden);
789 $cmd .= " \"$input_filename\" \"$output_filestem\"";
790
791 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000) {
792 $cmd .= " > \"$output_filestem.out\" 2> \"$output_filestem.err\"";
793 } else {
794 $cmd .= " > \"$output_filestem.err\"";
795 }
796
797 $!=0;
798
799 my $retval=system($cmd);
800 if ($retval!=0)
801 {
802 print STDERR "Error executing pdftohtml.pl";
803 if ($!) {print STDERR ": $!";}
804 print STDERR "\n";
805 }
806
807 # make sure the converter made something
808 if ($retval!=0 || ! -s "$output_filestem.html")
809 {
810 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
811 # print out the converter's std err, if any
812 if (-s "$output_filestem.err") {
813 open (ERRLOG, "$output_filestem.err") || die "$!";
814 print STDERR "pdftohtml error log:\n";
815 while (<ERRLOG>) {
816 print STDERR "$_";
817 }
818 close ERRLOG;
819 }
820 #print STDERR "***********output filestem $output_filestem.html\n";
821 &FileUtils::removeFiles("$output_filestem.html") if (-e "$output_filestem.html");
822 if (-e "$output_filestem.err") {
823 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
824 {
825 open (ERRLOG, "$output_filestem.err");
826 while (<ERRLOG>) {print FAILLOG $_;}
827 close ERRLOG;
828 close FAILLOG;
829 }
830 &FileUtils::removeFiles("$output_filestem.err");
831 }
832 return 0;
833 }
834
835 &FileUtils::removeFiles("$output_filestem.err") if (-e "$output_filestem.err");
836 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
837 return 1;
838}
839
840
841# Convert a pdf file to html with the newer Xpdftools' pdftohtml
842# This generates "paged HTML" where extracted, selectable text is positioned
843# over screenshots of each page.
844# Since xpdf's pdftohtml fails if the output dir already exists and for easier
845# naming, the output files are created in a "pages" subdirectory of the tmp
846# location parent of $output_filestem instead
847sub xpdf_to_html {
848 my ($dirname, $input_filename, $output_filestem) = @_;
849
850 my $cmd = "";
851
852 # build up the path to the doc-to-html conversion tool we're going to use
853 my $xpdf_pdftohtml = &FileUtils::filenameConcatenate(_get_xpdftools_bindir(), "pdftohtml");
854
855 # We'll create the file by name $output_filestem during post-conversion processing.
856 # Note that Xpdf tools will only create its conversion products in a dir that does
857 # not yet exist. So we'll create this location as a subdir of the output_filestem's
858 # parent directory. The parent dir is the already generated tmp area for conversion. So:
859 # - tmpdir gs2build/tmp/<random-num> already exists at this stage
860 # - We'll create gs2build/tmp/<rand>/output_filestem.html later, during post-processing
861 # - For now, XPdftools will create gs2build/tmp/<rand>/pages and put its products in there.
862 my ($tailname, $tmp_dirname, $suffix)
863 = &File::Basename::fileparse($output_filestem, "\\.[^\\.]+\$");
864 $tmp_dirname = &FileUtils::filenameConcatenate($tmp_dirname, "pages");
865
866 # xpdf's pdftohtml tool also takes a zoom factor, where a zoom of 1 is 100%
867 $cmd .= "\"$xpdf_pdftohtml\"";
868 $cmd .= " -z $pdf_zoom" if ($pdf_zoom);
869# $cmd .= " -c" if ($pdf_complex);
870# $cmd .= " -i" if ($pdf_ignore_images);
871# $cmd .= " -a" if ($pdf_allow_images_only);
872# $cmd .= " -hidden" unless ($pdf_nohidden);
873 $cmd .= " \"$input_filename\" \"$tmp_dirname\"";
874 #$cmd .= " \"$input_filename\" \"$output_filestem\"";
875
876 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000) {
877 $cmd .= " > \"$output_filestem.out\" 2> \"$output_filestem.err\"";
878 } else {
879 $cmd .= " > \"$output_filestem.err\"";
880 }
881
882 #print STDERR "@@@@ Running command: $cmd\n";
883
884 $!=0;
885 my $retval=system($cmd);
886 if ($retval!=0)
887 {
888 print STDERR "Error executing xpdf's pdftohtml tool";
889 if ($!) {print STDERR ": $!";}
890 print STDERR "\n";
891 }
892
893 # make sure the converter made something
894 if ($retval!=0 || ! -s &FileUtils::filenameConcatenate($tmp_dirname,"index.html"))
895 {
896 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
897 # print out the converter's std err, if any
898 if (-s "$output_filestem.err") {
899 open (ERRLOG, "$output_filestem.err") || die "$!";
900 print STDERR "pdftohtml error log:\n";
901 while (<ERRLOG>) {
902 print STDERR "$_";
903 }
904 close ERRLOG;
905 }
906 #print STDERR "***********output filestem $output_filestem.html\n";
907 &FileUtils::removeFiles("$tmp_dirname") if (-d "$tmp_dirname");
908 if (-e "$output_filestem.err") {
909 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
910 {
911 open (ERRLOG, "$output_filestem.err");
912 while (<ERRLOG>) {print FAILLOG $_;}
913 close ERRLOG;
914 close FAILLOG;
915 }
916 &FileUtils::removeFiles("$output_filestem.err");
917 }
918 return 0;
919 }
920
921 &FileUtils::removeFiles("$output_filestem.err") if (-e "$output_filestem.err");
922 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
923 return 1;
924}
925
926# Returns the path to xpdf-tools's containing bin dir appropriate for this machine's OS and bitness
927sub _get_xpdftools_bindir {
928
929 # build up the path to the containing bin dir of the xpdf conversion tool we're going to use
930 my $xpdf_tools_bin = &FileUtils::filenameConcatenate($ENV{'GSDLHOME'}, "bin", $ENV{'GSDLOS'}, "xpdf-tools");
931
932 if($ENV{'GSDLOS'} =~ m/^windows$/i) { # For Windows, just use the 32 bit xpdf's pdftohtml as it works the same as the 64 bit
933 $xpdf_tools_bin = &FileUtils::filenameConcatenate($xpdf_tools_bin, "bin32");
934 } else { # unix (linux|darwin), use the bin32/bin64 folder depending on the BITNESS env var
935
936 # Don't use $ENV{'GSDLARCH'}, use the new $ENV{'BITNESS'}, since $ENV{'GSDLARCH'}
937 # isn't always set and has side-effects when it is set:
938 # $ENV{'GSDLARCH'} is only (meant to be) set when many other 32-bit or 64-bit
939 # specific subdirectories exist in a greenstone installation.
940 # None of those locations need exist when xpdf-tools is installed with GS.
941 # So don't depend on GSDLARCH as forcing that to be exported has side-effects
942 if($ENV{'BITNESS'}) {
943 $xpdf_tools_bin = &FileUtils::filenameConcatenate($xpdf_tools_bin, "bin".$ENV{'BITNESS'});
944 } else { # what if $ENV{'BITNESS'} undefined, fallback on bin32? or 64?
945 $xpdf_tools_bin = &FileUtils::filenameConcatenate($xpdf_tools_bin, "bin32");
946 }
947 }
948
949 return $xpdf_tools_bin;
950}
951
952# Convert a pdf file to various types of image with the convert command
953
954sub pdfps_to_img {
955 my ($dirname, $input_filename, $output_filestem, $output_type) = @_;
956
957 # Check that ImageMagick is installed and available on the path (except for Windows 95/98)
958 if (!($ENV{'GSDLOS'} eq "windows" && !Win32::IsWinNT())) {
959 my $imagick_cmd = "\"".&util::get_perl_exec()."\" -S gs-magick.pl";
960 $imagick_cmd = $imagick_cmd." --verbosity=$verbosity" if defined $verbosity;
961 my $result = `$imagick_cmd identify 2>&1`;
962
963 # Linux and Windows return different values for "program not found".
964 # Linux returns -1 and Windows 256 for "program not found". But once they're
965 # converted to signed values, it will be -1 for Linux and 1 for Windows.
966 # Whenever we test for return values other than 0, shift by 8 and perform
967 # unsigned to signed status conversion on $? to get expected range of return vals
968 # Although gs-magick.pl already shifts its $? by 8, converts it to a signed value
969 # and then exits on that, by the time we get here, we need to do it again
970 my $status = $?;
971 $status >>= 8;
972 $status = (($status & 0x80) ? -(0x100 - ($status & 0xFF)) : $status);
973 if (($ENV{'GSDLOS'} ne "windows" && $status == -1) || ($ENV{'GSDLOS'} eq "windows" && $status == 1)) {
974 # if ($status == -1 || $status == 1) #if ($status == -1 || $status == 256) {
975 #ImageMagick is not installed, thus the convert utility is not available.
976 print STDERR "*** ImageMagick is not installed, the convert utility is not available. Unable to convert PDF/PS to images. Status: $status\n";
977 return 0;
978 }
979 }
980
981 my $cmd = "";
982 if ($timeout) {$cmd = "ulimit -t $timeout;";}
983 $output_type =~ s/.*\_(.*)/$1/i;
984 my $full_perl_path = &util::get_perl_exec();
985 $cmd .= "\"$full_perl_path\" -S pdfpstoimg.pl -convert_to $output_type \"$input_filename\" \"$output_filestem\"";
986 if ($ENV{'GSDLOS'} !~ m/^windows$/i || $is_winnt_2000) {
987 $cmd .= " > \"$output_filestem.out\" 2> \"$output_filestem.err\"";
988 } else {
989 $cmd .= " > \"$output_filestem.err\"";
990 }
991
992 # don't include path on windows (to avoid having to play about
993 # with quoting when GSDLHOME might contain spaces) but assume
994 # that the PATH is set up correctly
995 $!=0;
996 my $retval=system($cmd);
997 if ($retval!=0)
998 {
999 print STDERR "Error executing pdfpstoimg.pl";
1000 if ($!) {print STDERR ": $!";}
1001 print STDERR "\n";
1002 }
1003
1004 #make sure the converter made something
1005 #if ($retval !=0) || ! -s "$output_filestem")
1006 if ($retval !=0)
1007 {
1008 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
1009 #print out the converter's std err, if any
1010 if (-s "$output_filestem.err") {
1011 open (ERRLOG, "$output_filestem.err") || die "$!";
1012 print STDERR "pdfpstoimg error log:\n";
1013 while (<ERRLOG>) {
1014 print STDERR "$_";
1015 }
1016 close ERRLOG;
1017 }
1018 #&FileUtils::removeFiles("$output_filestem.html") if (-e "$output_filestem.html");
1019 if (-e "$output_filestem.err") {
1020 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
1021 {
1022 open (ERRLOG, "$output_filestem.err");
1023 while (<ERRLOG>) {print FAILLOG $_;}
1024 close ERRLOG;
1025 close FAILLOG;
1026 }
1027 &FileUtils::removeFiles("$output_filestem.err");
1028 }
1029 return 0;
1030 }
1031 &FileUtils::removeFiles("$output_filestem.err") if (-e "$output_filestem.err");
1032 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
1033 return 1;
1034}
1035
1036# Convert a PDF file to text with xpdftools' pdftotext command
1037# Works for Windows too, whereas the old pdftotxt didn't
1038sub xpdf_to_text {
1039 my ($dirname, $input_filename, $output_filestem) = @_;
1040
1041 my $cmd = "";
1042
1043 # build up the path to the doc-to-txt conversion tool we're going to use
1044 my $xpdf_pdftotxt = &FileUtils::filenameConcatenate(_get_xpdftools_bindir(), "pdftotext");
1045
1046 # For xpdf's pdftotxt options, see https://www.xpdfreader.com/pdftotext-man.html
1047 $cmd .= "\"$xpdf_pdftotxt\"";
1048 if($enc) {
1049 $cmd .= " -enc $enc"; # decode the bytes in the file using the designated encoding scheme
1050 } else {
1051 # as per https://www.xpdfreader.com/pdftotext-man.html
1052 # xpdf's pdftotxt defaults to using Latin-1 encoding, should we default to UTF-8?
1053 $cmd .= " -enc UTF-8"; # see https://www.xpdfreader.com/xpdfrc-man.html
1054 }
1055 $cmd .= " -nopgbrk";
1056 # Avoid the silly solitary carriage returns (CR in Notepad) at the end
1057 # of lines that ends up as \n appended to the doc title
1058 # by setting the end of line marker to unix style solitary newline (LF or \n),
1059 # which doesn't end up in the doc title
1060 $cmd .= " -eol unix";
1061 $cmd .= " \"$input_filename\" \"$output_filestem.text\"";
1062
1063 print STDERR "@@@@ Running command: $cmd\n";
1064
1065 return _run_pdf_to_text_cmd($cmd, $output_filestem);
1066}
1067
1068# Convert a PDF file to text with the pdftotext command
1069
1070sub pdf_to_text {
1071 my ($dirname, $input_filename, $output_filestem) = @_;
1072
1073 my $cmd = "pdftotext \"$input_filename\" \"$output_filestem.text\"";
1074
1075 return _run_pdf_to_text_cmd($cmd, $output_filestem);
1076}
1077
1078sub _run_pdf_to_text_cmd {
1079 my ($cmd, $output_filestem) = @_;
1080
1081 if ($ENV{'GSDLOS'} !~ m/^windows$/i) {
1082 $cmd .= " > \"$output_filestem.out\" 2> \"$output_filestem.err\"";
1083 } else {
1084 $cmd .= " > \"$output_filestem.err\"";
1085 }
1086
1087 if (system($cmd)!=0)
1088 {
1089 print STDERR "Error executing $cmd: $!\n";
1090 &FileUtils::removeFiles("$output_filestem.text") if (-e "$output_filestem.text");
1091 }
1092
1093 # make sure there is some extracted text.
1094 if (-e "$output_filestem.text") {
1095 open (EXTR_TEXT, "$output_filestem.text") || warn "open: $!";
1096 binmode(EXTR_TEXT); # just in case...
1097 my $line="";
1098 my $seen_text=0;
1099 while (($seen_text==0) && ($line=<EXTR_TEXT>)) {
1100 if ($line=~ m/\w/) {$seen_text=1;}
1101 }
1102 close EXTR_TEXT;
1103 if ($seen_text==0) { # no text was extracted
1104 print STDERR "Error: pdftotext found no text\n";
1105 &FileUtils::removeFiles("$output_filestem.text");
1106 }
1107 }
1108
1109 # make sure the converter made something
1110 if (! -s "$output_filestem.text")
1111 {
1112 # print out the converters std err, if any
1113 if (-s "$output_filestem.err") {
1114 open (ERRLOG, "$output_filestem.err") || die "$!";
1115 print STDERR "pdftotext error log:\n";
1116 while (<ERRLOG>) {
1117 print STDERR "$_";
1118 }
1119 close ERRLOG;
1120 }
1121 # does this converter create a .out file?
1122 &FileUtils::removeFiles("$output_filestem.out") if (-e "$output_filestem.out");
1123 &FileUtils::removeFiles("$output_filestem.text") if (-e "$output_filestem.text");
1124 if (-e "$output_filestem.err") {
1125 if ($faillogfile ne "" && defined(open(FAILLOG,">>$faillogfile")))
1126 {
1127 open (ERRLOG,"$output_filestem.err");
1128 while (<ERRLOG>) {print FAILLOG $_;}
1129 close ERRLOG;
1130 close FAILLOG;
1131 }
1132 &FileUtils::removeFiles("$output_filestem.err");
1133 }
1134 return 0;
1135 }
1136 &FileUtils::removeFiles("$output_filestem.err") if (-e "$output_filestem.err");
1137 return 1;
1138}
1139
1140# Convert a PostScript document to text
1141# note - just using "ps2ascii" isn't good enough, as it
1142# returns 0 for a postscript interpreter error. ps2ascii is just
1143# a wrapper to "gs" anyway, so we use that cmd here.
1144
1145sub ps_to_text {
1146 my ($input_filename, $output_filestem) = @_;
1147
1148 my $error = "";
1149
1150 # if we're on windows we'll fall straight through without attempting
1151 # to use gs
1152 if ($ENV{'GSDLOS'} =~ m/^windows$/i) {
1153 $error = "Windows does not support gs";
1154
1155 } else {
1156 my $cmd = "";
1157 if ($timeout) {$cmd = "ulimit -t $timeout; ";}
1158 $cmd .= "gs -q -dNODISPLAY -dNOBIND -dWRITESYSTEMDICT -dSIMPLE -c save ";
1159 $cmd .= "-f ps2ascii.ps \"$input_filename\" -c quit > \"$output_filestem.text\"";
1160 #$cmd .= "pstotext -output \"$output_filestem.text\" $input_filename\"";
1161 $cmd .= " 2> $output_filestem.err";
1162 $!=0;
1163
1164 my $retcode=system($cmd);
1165 $retcode = $? >> 8; # see man perlfunc - system for this...
1166 # if system returns -1 | 127 (couldn't start program), look at $! for message
1167
1168 if ($retcode!=0) {if ($!) {$error=$!;} else {$error="couldn't run.\n";}}
1169 elsif (! -e "$output_filestem.text") {
1170 $error="did not create output file.\n";
1171 }
1172 else
1173 { # make sure the interpreter didn't get an error. It is technically
1174 # possible for the actual text to start with this, but....
1175 open PSOUT, "$output_filestem.text";
1176 if (<PSOUT> =~ m/^Error: (.*)/) {
1177 $error="interpreter error - \"$1\"";
1178 }
1179 close PSOUT;
1180 }
1181 }
1182
1183 if ($error ne "")
1184 {
1185 print STDERR "Warning: Error executing gs: $error\n";
1186 print STDERR "Resorting to Perl regular expressions to extract text from PostScript...\n";
1187 &FileUtils::removeFiles("$output_filestem.text") if (-e "$output_filestem.text");
1188
1189 if ("$faillogfile" ne "" && defined(open (FAILLOG, ">>$faillogfile")))
1190 {
1191 print FAILLOG "gs - $error\n";
1192 if (-e "$output_filestem.err") {
1193 open(ERRLOG, "$output_filestem.err");
1194 while (<ERRLOG>) {print FAILLOG $_;}
1195 close ERRLOG;
1196 }
1197 close FAILLOG;
1198 }
1199 &FileUtils::removeFiles("$output_filestem.err") if (-e "$output_filestem.err");
1200
1201
1202 # Fine then. We'll just do a lousy job by ourselves...
1203 # Based on 5-line regexp sed script found at:
1204 # http://snark.ptc.spbu.ru/mail-archives/lout/brown/msg00003.html
1205 #
1206 print STDERR "Stripping text from postscript\n";
1207 my $errorcode=0;
1208 open (IN, "$input_filename")
1209 || ($errorcode=1, warn "Couldn't read file: $!");
1210 open (OUT, ">$output_filestem.text")
1211 || ($errorcode=1, warn "Couldn't write file: $!");
1212 if ($errorcode) {print STDERR "errors\n";return 0;}
1213
1214 my $text=""; # this is for whole .ps file...
1215 $text = join('', <IN>); # see man perlport, under "System Resources"
1216 close IN;
1217
1218 # Make sure this is a ps file...
1219 if ($text !~ m/^%!/) {
1220 print STDERR "Bad postscript header: not '%!'\n";
1221 if ($faillogfile ne "" && defined(open(FAILLOG, ">>$faillogfile")))
1222 {
1223 print FAILLOG "Bad postscript header: not '%!'\n";
1224 close FAILLOG;
1225 }
1226 return 0;
1227 }
1228
1229 # if ps has Page data, then use it to delete all stuff before it.
1230 $text =~ s/^.*?%%Page:.*?\n//s; # treat string as single line
1231
1232 # remove all leading non-data stuff
1233 $text =~ s/^.*?\(//s;
1234
1235 # remove all newline chars for easier processing
1236 $text =~ s/\n//g;
1237
1238 # Big assumption here - assume that if any co-ordinates are
1239 # given, then we are at the end of a sentence.
1240 $text =~ s/\)-?\d+\ -?\d+/\) \(\n\)/g;
1241
1242 # special characters--
1243 $text =~ s/\(\|\)/\(\ - \)/g; # j -> em-dash?
1244
1245 # ? ps text formatting (eg italics?) ?
1246 $text =~ s/Fn\(f\)/\(\{\)/g; # f -> {
1247 $text =~ s/Fn\(g\)/\(\}\)/g; # g -> }
1248 $text =~ s/Fn\(j\)/\(\|\)/g; # j -> |
1249 # default - remove the rest
1250 $text =~ s/\ ?F.\((.+?)\)/\($1\)/g;
1251
1252 # attempt to add whitespace between words...
1253 # this is based purely on observation, and may be completely wrong...
1254 $text =~ s/([^F])[defghijkuy]\(/$1 \( /g;
1255 # eg I notice "b(" is sometimes NOT a space if preceded by a
1256 # negative number.
1257 $text =~ s/\)\d+ ?b\(/\) \( /g;
1258
1259 # change quoted braces to brackets
1260 $text =~ s/([^\\])\\\(/$1\{/g;
1261 $text =~ s/([^\\])\\\)/$1\}/g ;
1262
1263 # remove everything that is not between braces
1264 $text =~ s/\)([^\(\)])+?\(//sg ;
1265
1266 # remove any Trailer eof stuff.
1267 $text =~ s/\)[^\)]*$//sg;
1268
1269 ### ligatures have special characters...
1270 $text =~ s/\\013/ff/g;
1271 $text =~ s/\\014/fi/g;
1272 $text =~ s/\\015/fl/g;
1273 $text =~ s/\\016/ffi/g;
1274 $text =~ s/\\214/fi/g;
1275 $text =~ s/\\215/fl/g;
1276 $text =~ s/\\017/\n\* /g; # asterisk?
1277 $text =~ s/\\023/\023/g; # e acute ('e)
1278 $text =~ s/\\177/\252/g; # u"
1279# $text =~ s/ ?? /\344/g; # a"
1280
1281 print OUT "$text";
1282 close OUT;
1283 }
1284 # wrap the text - use a minimum length. ie, first space after this length.
1285 my $wrap_length=72;
1286 &FileUtils::moveFiles("$output_filestem.text", "$output_filestem.text.tmp");
1287 open INFILE, "$output_filestem.text.tmp" ||
1288 die "Couldn't open file: $!";
1289 open OUTFILE, ">$output_filestem.text" ||
1290 die "Couldn't open file for writing: $!";
1291 my $line="";
1292 while ($line=<INFILE>) {
1293 while (length($line)>0) {
1294 if (length($line)>$wrap_length) {
1295 $line =~ s/^(.{$wrap_length}[^\s]*)\s*//;
1296 print OUTFILE "$1\n";
1297 } else {
1298 print OUTFILE "$line";
1299 $line="";
1300 }
1301 }
1302 }
1303 close INFILE;
1304 close OUTFILE;
1305 &FileUtils::removeFiles("$output_filestem.text.tmp");
1306
1307 &FileUtils::removeFiles("$output_filestem.err") if (-e "$output_filestem.err");
1308 return 1;
1309}
1310
1311
1312# Convert any file to HTML with a crude perl implementation of the
1313# UNIX strings command.
1314
1315sub any_to_html {
1316 my ($input_filename, $output_filestem) = @_;
1317
1318 # First generate a text file
1319 return 0 unless (&any_to_text($input_filename, $output_filestem));
1320
1321 # create an HTML file from the text file
1322 open(TEXT, "<$output_filestem.text");
1323 open(HTML, ">$output_filestem.html");
1324
1325 print HTML "<html><head>\n";
1326 print HTML "<META HTTP-EQUIV=\"Content-Type\" CONTENT=\"text/html\">\n";
1327 print HTML "<META NAME=\"GENERATOR\" CONTENT=\"Greenstone any_to_html\">\n";
1328 print HTML "</head><body>\n\n";
1329
1330 my $line;
1331 while ($line=<TEXT>) {
1332 $line =~ s/</&lt;/g;
1333 $line =~ s/>/&gt;/g;
1334 if ($line =~ m/^\s*$/) {
1335 print HTML "<p>";
1336 } else {
1337 print HTML "<br> ", $line;
1338 }
1339 }
1340 print HTML "\n</body></html>\n";
1341
1342 close HTML;
1343 close TEXT;
1344
1345 &FileUtils::removeFiles("$output_filestem.text") if (-e "$output_filestem.text");
1346 return 1;
1347}
1348
1349# Convert any file to TEXT with a crude perl implementation of the
1350# UNIX strings command.
1351# Note - this assumes ascii charsets :( (jrm21)
1352
1353sub any_to_text {
1354 my ($input_filename, $output_filestem) = @_;
1355
1356 if (!$use_strings) {
1357 return 0;
1358 }
1359
1360 print STDERR "\n**** In any to text****\n\n";
1361 open(IN, "<$input_filename") || return 0;
1362 binmode(IN);
1363 open(OUT, ">$output_filestem.text") || return 0;
1364
1365 my ($line);
1366 my $output_line_count = 0;
1367 while (<IN>) {
1368 $line = $_;
1369
1370 # delete anything that isn't a printable character
1371 $line =~ s/[^\040-\176]+/\n/sg;
1372
1373 # delete any string less than 10 characters long
1374 $line =~ s/^.{0,9}$/\n/mg;
1375 while ($line =~ m/^.{1,9}$/m) {
1376 $line =~ s/^.{0,9}$/\n/mg;
1377 $line =~ s/\n+/\n/sg;
1378 }
1379
1380 # remove extraneous whitespace
1381 $line =~ s/\n+/\n/gs;
1382 $line =~ s/^\n//gs;
1383
1384 # output whatever is left
1385 if ($line =~ m/[^\n ]/) {
1386 print OUT $line;
1387 ++$output_line_count;
1388 }
1389 }
1390
1391 close OUT;
1392 close IN;
1393
1394 if ($output_line_count) { # try to protect against binary only formats
1395 return 1;
1396 }
1397
1398 &FileUtils::removeFiles("$output_filestem.text");
1399 return 0;
1400
1401}
Note: See TracBrowser for help on using the repository browser.