root/gsdl/trunk/perllib/plugins/ImageConverter.pm @ 19054

Revision 19054, 16.6 KB (checked in by davidb, 11 years ago)

MaxImageWidth? and MaxImageHeight? set as metadata

  • Property svn:executable set to *
Line 
1###########################################################################
2#
3# ImageConverter - helper plugin that does image conversion using ImageMagick
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# Copyright (C) 2008 New Zealand Digital Library Project
10#
11# This program is free software; you can redistribute it and/or modify
12# it under the terms of the GNU General Public License as published by
13# the Free Software Foundation; either version 2 of the License, or
14# (at your option) any later version.
15#
16# This program is distributed in the hope that it will be useful,
17# but WITHOUT ANY WARRANTY; without even the implied warranty of
18# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
19# GNU General Public License for more details.
20#
21# You should have received a copy of the GNU General Public License
22# along with this program; if not, write to the Free Software
23# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
24#
25###########################################################################
26package ImageConverter;
27
28use BaseMediaConverter;
29
30
31use strict;
32no strict 'refs'; # allow filehandles to be variables and viceversa
33
34use gsprintf 'gsprintf';
35
36BEGIN {
37    @ImageConverter::ISA = ('BaseMediaConverter');
38}
39
40my $arguments = [
41      { 'name' => "create_thumbnail",
42    'desc' => "{ImageConverter.create_thumbnail}",
43    'type' => "enum",
44    'list' => [{'name' => "true", 'desc' => "{common.true}"},
45           {'name' => "false", 'desc' => "{common.false}"}],
46    'deft' => "true",
47    'reqd' => "no" },
48      { 'name' => "thumbnailsize",
49    'desc' => "{ImageConverter.thumbnailsize}",
50    'type' => "int",
51    'deft' => "100",
52    'range' => "1,",
53    'reqd' => "no" },
54      { 'name' => "thumbnailtype",
55    'desc' => "{ImageConverter.thumbnailtype}",
56    'type' => "string",
57    'deft' => "gif",
58    'reqd' => "no" },
59      { 'name' => "noscaleup",
60    'desc' => "{ImageConverter.noscaleup}",
61    'type' => "flag",
62    'reqd' => "no" },
63      { 'name' => "create_screenview",
64    'desc' => "{ImageConverter.create_screenview}",
65    'type' => "enum",
66    'list' => [{'name' => "true", 'desc' => "{common.true}"},
67           {'name' => "false", 'desc' => "{common.false}"}],
68    'deft' => "true",
69    'reqd' => "no" },
70      { 'name' => "screenviewsize",
71    'desc' => "{ImageConverter.screenviewsize}",
72    'type' => "int",
73    'deft' => "500",
74    'range' => "1,",
75    'reqd' => "no" },
76      { 'name' => "screenviewtype",
77    'desc' => "{ImageConverter.screenviewtype}",
78    'type' => "string",
79    'deft' => "jpg",
80    'reqd' => "no" },
81      { 'name' => "converttotype",
82    'desc' => "{ImageConverter.converttotype}",
83    'type' => "string",
84    'deft' => "",
85    'reqd' => "no" },
86      { 'name' => "minimumsize",
87    'desc' => "{ImageConverter.minimumsize}",
88    'type' => "int",
89    'deft' => "100",
90    'range' => "1,",
91    'reqd' => "no" }
92         ];
93
94my $options = { 'name' => "ImageConverter",
95        'desc' => "{ImageConverter.desc}",
96        'abstract' => "yes",
97        'inherits' => "yes",
98        'args' => $arguments };
99
100sub new {
101    my ($class) = shift (@_);
102    my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
103    push(@$pluginlist, $class);
104
105    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
106    push(@{$hashArgOptLists->{"OptList"}},$options);
107
108    my $self = new BaseMediaConverter($pluginlist, $inputargs, $hashArgOptLists, 1);
109
110
111    return bless $self, $class;
112
113}
114
115# needs to be called after BasePlugin init, so that outhandle is set up.
116sub init {
117    my $self = shift(@_);
118
119    $self->{'tmp_file_paths'} = ();
120
121    # Check that ImageMagick is installed and available on the path
122    my $image_conversion_available = 1;
123    my $no_image_conversion_reason = "";
124    # None of this works very well on Windows 95/98...
125    if ($ENV{'GSDLOS'} eq "windows" && !Win32::IsWinNT()) {
126    $image_conversion_available = 0;
127    $no_image_conversion_reason = "win95notsupported";
128    } else {
129    my $result = `identify -version 2>&1`;
130    if ($? == -1 || $? == 256) {  # Linux and Windows return different values for "program not found"
131        $image_conversion_available = 0;
132        $no_image_conversion_reason = "imagemagicknotinstalled";
133    }
134    }
135    $self->{'image_conversion_available'} = $image_conversion_available;
136    $self->{'no_image_conversion_reason'} = $no_image_conversion_reason;
137
138    if ($self->{'image_conversion_available'} == 0) {
139    my $outhandle = $self->{'outhandle'};
140    &gsprintf($outhandle, "ImageConverter: {ImageConverter.noconversionavailable} ({ImageConverter.".$self->{'no_image_conversion_reason'}."})\n");
141    }
142       
143}
144   
145
146# convert image to new type if converttotype is set
147# generate thumbnails if required
148# generate screenview if required
149# discover image metadata
150# filename_no_path must be in utf8 and url-encoded
151sub generate_images {
152    my $self = shift(@_);
153    my ($filename_full_path, $filename_no_path, $doc_obj, $section) = @_;
154
155    # check image magick status
156    return 0 if $self->{'image_conversion_available'} == 0;
157
158    # check the filenames
159    return 0 if ($filename_no_path eq "" || !-f $filename_full_path);
160
161    if ($self->{'enable_cache'}) {
162    $self->init_cache_for_file($filename_full_path);
163    }
164
165    my $verbosity = $self->{'verbosity'};
166    my $outhandle = $self->{'outhandle'};
167
168    # check the size of the image against minimum size if specified
169    my $minimumsize = $self->{'minimumsize'};
170    if (defined $minimumsize && (-s $filename_full_path < $minimumsize)) {
171        print $outhandle "ImageConverter: \"$filename_full_path\" too small, skipping\n"
172        if ($verbosity > 1);
173    return 0; # or is there a better return value??
174    }
175   
176    my $filehead = $filename_no_path;
177    $filehead =~ s/\.([^\.]*)$//; # filename with no extension
178    my $assocfilemeta = "[assocfilepath]";
179    if ($section ne $doc_obj->get_top_section()) {
180    $assocfilemeta = "[parent(Top):assocfilepath]";
181    }
182
183    # The images that will get generated may contain percent signs in their src filenames
184    # Encode those percent signs themselves so that urls to the imgs refer to them correctly
185    my $url_to_filehead = &unicode::filename_to_url($filehead);
186    my $url_to_filename_no_path = &unicode::filename_to_url($filename_no_path);
187   
188    # Convert the image to a new type (if required).
189    my $converttotype = $self->{'converttotype'};
190    my $type = "unknown";
191
192    if ($converttotype ne "" && $filename_full_path !~ m/$converttotype$/) {
193    #    $doc_obj->add_utf8_metadata($section, "Image", $utf8_filename_meta);
194
195    my ($result,$filename_full_path)
196        = $self->convert($filename_full_path, $converttotype, "", "CONVERTTYPE");
197
198    $type = $converttotype;
199    $filename_no_path = "$filehead.$type";
200    $url_to_filename_no_path = "$url_to_filehead.$type";
201    }
202
203    # add Image metadata
204    $doc_obj->add_utf8_metadata($section, "Image", $url_to_filename_no_path); # url to generated image
205
206    # here we overwrite the original with the potentially converted one
207    $doc_obj->set_utf8_metadata_element($section, "Source", &unicode::url_decode($filename_no_path)); # displayname of generated image
208    $doc_obj->set_utf8_metadata_element($section, "SourceFile", $url_to_filename_no_path); # displayname of generated image
209
210    # use identify to get info about the (possibly converted) image
211    my ($image_type, $image_width, $image_height, $image_size)
212    = &identify($filename_full_path, $outhandle, $verbosity);
213
214    if ($image_type ne " ") {
215    $type = $self->correct_mime_type($image_type);
216    }
217
218    #overwrite the ones added in BasePlugin
219    $doc_obj->set_metadata_element ($section, "FileFormat", $type);
220    $doc_obj->set_metadata_element ($section, "FileSize",   $image_size);
221
222    $doc_obj->add_metadata ($section, "ImageType",   $image_type);
223    $doc_obj->add_metadata ($section, "ImageWidth",  $image_width);
224    $doc_obj->add_metadata ($section, "ImageHeight", $image_height);
225    $doc_obj->add_metadata ($section, "ImageSize",   $image_size);
226
227    if ((defined $self->{'MaxImageWidth'})
228    && ($image_width > $self->{'MaxImageWidth'})) {
229    $self->{'MaxImageWidth'} = $image_width;
230    }
231    if ((defined $self->{'MaxImageHeight'})
232    && ($image_height > $self->{'MaxImageHeight'})) {
233    $self->{'MaxImageHeight'} = $image_height;
234    }
235
236
237    $doc_obj->add_metadata ($section, "srclink", "<a href=\"_httpprefix_/collect/[collection]/index/assoc/$assocfilemeta/[Image]\">");
238    $doc_obj->add_metadata ($section, "/srclink", "</a>");
239    $doc_obj->add_metadata ($section, "srcicon", "<img src=\"_httpprefix_/collect/[collection]/index/assoc/$assocfilemeta/[Image]\" width=\"[ImageWidth]\" height=\"[ImageHeight]\">");
240
241    # Add the image as an associated file
242    $doc_obj->associate_file($filename_full_path, $filename_no_path, "image/$type", $section);
243
244    if ($self->{'create_thumbnail'} eq "true") {
245    $self->create_thumbnail($filename_full_path, $filehead, $doc_obj, $section, $assocfilemeta, $url_to_filehead);
246    }
247    if ($self->{'create_screenview'} eq "true") {
248    $self->create_screenview($filename_full_path, $filehead, $doc_obj, $section, $assocfilemeta, $url_to_filehead, $image_width, $image_height);
249    }
250}
251
252sub create_thumbnail {
253    my $self = shift(@_);
254    my ($original_file, $filehead, $doc_obj, $section, $assocfilemeta, $url_to_filehead) = @_;
255    $url_to_filehead = $filehead unless defined $url_to_filehead;
256
257    my $thumbnailsize = $self->{'thumbnailsize'};
258    my $thumbnailtype = $self->correct_mime_type($self->{'thumbnailtype'});
259
260    # Generate the thumbnail with convert
261    my ($result,$thumbnailfile)
262    = $self->convert($original_file, $thumbnailtype, "-geometry $thumbnailsize" . "x$thumbnailsize", "THUMB");
263   
264    # Add the thumbnail as an associated file ...
265    if (-e "$thumbnailfile") {
266    $doc_obj->associate_file("$thumbnailfile", $filehead."_thumb.$thumbnailtype",
267                 "image/$thumbnailtype",$section); # name of generated image
268    $doc_obj->add_metadata ($section, "ThumbType", $thumbnailtype);
269    $doc_obj->add_utf8_metadata ($section, "Thumb", $url_to_filehead."_thumb.$thumbnailtype"); # url to generated image
270   
271    $doc_obj->add_metadata ($section, "thumbicon", "<img src=\"_httpprefix_/collect/[collection]/index/assoc/$assocfilemeta/[Thumb]\" width=[ThumbWidth] height=[ThumbHeight]>");
272   
273   
274    # Extract Thumbnail metadata from convert output
275    if ($result =~ m/[0-9]+x[0-9]+=>([0-9]+)x([0-9]+)/) {
276        $doc_obj->add_metadata ($section, "ThumbWidth", $1);
277        $doc_obj->add_metadata ($section, "ThumbHeight", $2);
278    }
279    } else {
280    my $outhandle = $self->{'outhandle'};
281    print $outhandle "Couldn't find thumbnail $thumbnailfile\n";
282
283    }
284}
285
286sub create_screenview {
287   
288    my $self = shift(@_);
289    my ($original_file, $filehead, $doc_obj, $section, $assocfilemeta, $url_to_filehead, $image_width, $image_height) = @_;
290    $url_to_filehead = $filehead unless defined $url_to_filehead;
291
292    my $screenviewsize = $self->{'screenviewsize'};
293    my $screenviewtype = $self->correct_mime_type($self->{'screenviewtype'});
294
295    # Scale the image, unless the original image is smaller than the screenview size and -noscaleup is set
296    my $scale_option = "-geometry $screenviewsize" . "x$screenviewsize";
297    if ($self->{'noscaleup'} && $image_width < $screenviewsize && $image_height < $screenviewsize)
298    {
299    $scale_option = "";
300    }
301   
302    # make the screenview image
303    my ($result,$screenviewfilename)
304    = $self->convert($original_file, $screenviewtype, $scale_option, "SCREEN");   
305   
306    #add the screenview as an associated file ...
307    if (-e "$screenviewfilename") {
308    $doc_obj->associate_file("$screenviewfilename", $filehead."_screen.$screenviewtype", "image/$screenviewtype",$section); # name of generated image
309    $doc_obj->add_metadata ($section, "ScreenType", $screenviewtype);
310    $doc_obj->add_utf8_metadata ($section, "Screen", $url_to_filehead."_screen.$screenviewtype"); # url to generated image
311   
312    $doc_obj->add_metadata ($section, "screenicon", "<img src=\"_httpprefix_/collect/[collection]/index/assoc/$assocfilemeta/[Screen]\" width=[ScreenWidth] height=[ScreenHeight]>");
313
314    # get screenview dimensions, size and type
315    if ($result =~ m/[0-9]+x[0-9]+=>([0-9]+)x([0-9]+)/) {
316        $doc_obj->add_metadata ($section, "ScreenWidth", $1);
317        $doc_obj->add_metadata ($section, "ScreenHeight", $2);
318    } elsif ($result =~ m/([0-9]+)x([0-9]+)/) {
319        #if the image hasn't changed size, the previous regex doesn't match
320        $doc_obj->add_metadata ($section, "ScreenWidth", $1);
321        $doc_obj->add_metadata ($section, "ScreenHeight", $2);
322    }
323    } else {
324    my $outhandle = $self->{'outhandle'};
325    print $outhandle "Couldn't find screenview file $screenviewfilename\n";
326
327    }
328
329}
330
331
332
333sub convert {
334    my $self = shift(@_);
335    my $source_file_path = shift(@_);
336    my $target_file_type = shift(@_);
337    my $convert_options  = shift(@_) || "";
338    my $convert_id       = shift(@_) || "";
339
340    my $outhandle = $self->{'outhandle'};
341    my $verbosity = $self->{'verbosity'};
342
343    my $source_file_no_path = &File::Basename::basename($source_file_path);
344
345    # Determine the full name and path of the output file
346    my $target_file_path;
347    if ($self->{'enable_cache'}) {
348    my $cached_image_dir = $self->{'cached_dir'};
349    my $image_root = $self->{'cached_file_root'};
350    $image_root .= "_$convert_id" if ($convert_id ne "");
351    my $image_file = "$image_root.$target_file_type";
352    $target_file_path = &util::filename_cat($cached_image_dir,$image_file);
353    }
354    else {
355    $target_file_path = &util::get_tmp_filename($target_file_type);
356    push(@{$self->{'tmp_file_paths'}}, $target_file_path);
357
358    # Output filename used to be parsed from result line:
359    #   my ($ofilename) = ($result =~ m/=>(.*\.$target_file_type)/);
360    # by the function that called 'convert'
361    # but this is no longer needed, as output filename is now
362    # explicitly passed back
363    }
364
365    # Generate and run the convert command
366    my $convert_command = "convert -interlace plane -verbose $convert_options \"$source_file_path\" \"$target_file_path\"";
367
368    my $print_info = { 'message_prefix' => $convert_id,
369               'message' => "Converting image $source_file_no_path to: $convert_id $target_file_type" };
370 
371    my ($regenerated,$result,$had_error)
372    = $self->autorun_general_cmd($convert_command,$target_file_path,$print_info);
373
374    return ($result,$target_file_path);
375}
376
377
378# Discover the characteristics of an image file with the ImageMagick
379# "identify" command.
380
381sub identify {
382    my ($image, $outhandle, $verbosity) = @_;
383
384    # Use the ImageMagick "identify" command to get the file specs
385    my $command = "identify \"$image\" 2>&1";
386    print $outhandle "$command\n" if ($verbosity > 2);
387    my $result = '';
388    $result = `$command`;
389    print $outhandle "$result\n" if ($verbosity > 3);
390
391    # Read the type, width, and height
392    my $type =   'unknown';
393    my $width =  'unknown';
394    my $height = 'unknown';
395
396    my $image_safe = quotemeta $image;
397    if ($result =~ /^$image_safe (\w+) (\d+)x(\d+)/) {
398    $type = $1;
399    $width = $2;
400    $height = $3;
401    }
402
403    # Read the size
404    my $size = "unknown";
405    if ($result =~ m/^.* ([0-9]+)b/) {
406    $size = $1;
407    }
408    elsif ($result =~ m/^.* ([0-9]+)(\.([0-9]+))?kb?/) {
409    $size = 1024 * $1;
410    if (defined($2)) {
411        $size = $size + (1024 * $2);
412        # Truncate size (it isn't going to be very accurate anyway)
413        $size = int($size);
414    }
415    }
416    elsif ($result =~ m/^.* (([0-9]+)(\.([0-9]+))?e\+([0-9]+))(kb|b)?/) {
417    # Deals with file sizes on Linux of type "3.4e+02kb" where e+02 is 1*10^2.
418    # 3.4e+02 therefore evaluates to 3.4 x 1 x 10^2 = 340kb.
419    # Programming languages including Perl know how that 3.4e+02 is a number,
420    # so we don't need to do any calculations.
421    $size = $1*1; # turn the string into a number by multiplying it by 1
422           #if we did $size = $1; $size would be merely the string "3.4e+02"
423    $size = int($size); # truncate size
424    }
425    print $outhandle "file: $image:\t $type, $width, $height, $size\n"
426    if ($verbosity > 2);
427
428    # Return the specs
429    return ($type, $width, $height, $size);
430}
431
432sub clean_up_temporary_files {
433    my $self = shift(@_);
434
435    foreach my $tmp_file_path (@{$self->{'tmp_file_paths'}}) {
436    if (-e $tmp_file_path) {
437        &util::rm($tmp_file_path);
438    }
439    }
440   
441}
442
443# image/jpg is not a valid mime-type, it ought to be image/jpeg.
444# Sometimes JPEG is passed in also, want to keep things lowercase just in case.
445sub correct_mime_type {
446    my $self = shift(@_);
447    my ($file_extension) = @_;
448   
449    $file_extension = lc($file_extension);
450    $file_extension =~ s/jpg/jpeg/s;
451
452    return $file_extension;
453}
454
4551; 
Note: See TracBrowser for help on using the browser.