Changeset 22880

Timestamp:

2010-09-08T12:58:08+12:00 (14 years ago)

Author:

kjdon

Message:

implemented the read method for when using open office to convert to html multi - the powerpoint gets converted to individual html files, two per slide. one for the image, one for the text. each one gets passed to HTMLPlugin for processing, so all the slides end up as individual documents, but the first page, back, continue etc links work to link them all together

File:

• main/trunk/greenstone2/perllib/plugins/PowerPointPlugin.pm (modified) (4 diffs)

main/trunk/greenstone2/perllib/plugins/PowerPointPlugin.pm

@@ -62 +62 @@
     [ { 'name' => "auto",
     'desc' => "{ConvertBinaryFile.convert_to.auto}" },
-      { 'name' => "html",
-    'desc' => "{PowerPointPlugin.convert_to.oo_html}" },
+      { 'name' => "html_multi",
+    'desc' => "{PowerPointPlugin.convert_to.html_multi}" },
       { 'name' => "text",
     'desc' => "{ConvertBinaryFile.convert_to.text}" },
@@ -274 +274 @@
 
 # override default read in some situations, as the conversion of ppt to html results in many files, and we want them all to be processed.
-sub read_XX {
+sub read {
     my $self = shift (@_);  
     my ($pluginfo, $base_dir, $file, $block_hash, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
@@ -283 +283 @@
     return undef unless $self->can_process_this_file($filename_full_path);
     
-    my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
-    
-    if ((defined $process_status) && ($process_status == 1)) {
+    # we are only doing something special for html_multi
+    if (!($self->{'openoffice_conversion'} && $self->{'convert_to'} eq "html_multi")) {
+    return $self->BasePlugin::read(@_);
+    }
+    my $outhandle = $self->{'outhandle'};
+    print STDERR "<Processing n='$file' p='$self->{'plugin_type'}'>\n" if ($gli);
+    print $outhandle "$self->{'plugin_type'} processing $file\n"
+        if $self->{'verbosity'} > 1;
+
+    my $conv_filename = $self->tmp_area_convert_file("html", $filename_full_path);
+    if ("$conv_filename" eq "") {return -1;} # had an error, will be passed down pipeline 
+    if (! -e "$conv_filename") {return -1;} 
+
+    my ($tailname, $html_dirname, $suffix)
+    = &File::Basename::fileparse($conv_filename, "\\.[^\\.]+\$");
+
+    my $collect_file = &util::filename_within_collection($filename_full_path);
+    my $dirname_within_collection = &util::filename_within_collection($html_dirname);
+    my $secondary_plugin = $self->{'secondary_plugins'}->{"HTMLPlugin"};
+
+    my @dir;
+    if (!opendir (DIR, $html_dirname)) {
+    print $outhandle "PowerPointPlugin: Couldn't read directory $html_dirname\n";
+    # just process the original file
+    @dir = ("$tailname.$suffix");
+    
+    } else {
+    @dir = readdir (DIR);
+    closedir (DIR);
+    }
+
+    foreach my $file (@dir) {
+    next unless $file =~ /\.html$/;
+    
+    my ($rv, $doc_obj) = 
+        $secondary_plugin->read_into_doc_obj ($pluginfo,"", &util::filename_cat($html_dirname,$file), $block_hash, {}, $processor, $maxdocs, $total_count, $gli);
+    if ((!defined $rv) || ($rv<1)) {
+        # wasn't processed
+        return $rv;
+    }
+
+    # next block copied from ConvertBinaryFile
+    # from here ...
+    # Override previous gsdlsourcefilename set by secondary plugin
+    
+    $doc_obj->set_source_filename ($collect_file, $self->{'file_rename_method'}); 
+    ## set_source_filename does not set the doc_obj source_path which is used in archives dbs for incremental
+    # build. so set it manually.
+    $doc_obj->{'source_path'} = $filename_full_path;
+    $doc_obj->set_converted_filename(&util::filename_cat($dirname_within_collection, $file));
+    
+    $self->set_Source_metadata($doc_obj, $filename_no_path);
+        
+    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
+    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FileSize", (-s $filename_full_path));
+
+    
+    my ($tailname, $dirname, $suffix)
+        = &File::Basename::fileparse($filename_full_path, "\\.[^\\.]+\$");
+    $doc_obj->set_utf8_metadata_element($doc_obj->get_top_section(), "FilenameRoot", $tailname);
+    
+
+    my $topsection = $doc_obj->get_top_section();
+    $self->add_associated_files($doc_obj, $filename_full_path);
+    
+    # extra_metadata is already called by sec plugin in process??
+    $self->extra_metadata($doc_obj, $topsection, $metadata); # do we need this here??
+    # do any automatic metadata extraction
+    $self->auto_extract_metadata ($doc_obj);
+    
+    # have we found a Title??
+    $self->title_fallback($doc_obj,$topsection,$filename_no_path);
+    
+    # use the one generated by HTMLPlugin, otherwise they all end up with same id.
+    #$self->add_OID($doc_obj);
+    # to here...
+
+    # process it
+    $processor->process($doc_obj);
+    undef $doc_obj;
+    }
+    $self->{'num_processed'} ++;
+
+#    my ($process_status,$doc_obj) = $self->read_into_doc_obj(@_);
+    
+#    if ((defined $process_status) && ($process_status == 1)) {
     
     # process the document
-    $processor->process($doc_obj);
-
-    $self->{'num_processed'} ++;
-    undef $doc_obj; 
-    }
+#   $processor->process($doc_obj);
+
+#   $self->{'num_processed'} ++;
+#   undef $doc_obj; 
+#    }
     # delete any temp files that we may have created
     $self->clean_up_after_doc_obj_processing();
@@ -298 +381 @@
 
     # if process_status == 1, then the file has been processed.
-    return $process_status;
+    return 1;
 
 }