Changeset 29013 for main/trunk

Timestamp:

2014-05-01T18:25:31+12:00 (10 years ago)

Author:

ak19

Message:

Following Dr Bainbridge's suggestion on determining a document's title which forms the anchor text for its RSS link.

Location:

main/trunk/greenstone2/perllib

Files:

• inexport.pm (modified) (1 diff)
• plugouts/BasePlugout.pm (modified) (2 diffs)
• strings.properties (modified) (1 diff)

main/trunk/greenstone2/perllib/inexport.pm

@@ -633 +633 @@
         # If a plugout was specified in the collect.cfg file, assume it is sensible
         # We can't check the name because it could be anything, if it is a custom plugout
-        print STDERR "Using plugout specified in collect.cfg: $collectcfg->{'plugout'}\n";
+        print STDERR "Using plugout specified in collect.cfg: ".join(' ', @{$collectcfg->{'plugout'}})."\n";
         $plugout = $collectcfg->{'plugout'};
     }

main/trunk/greenstone2/perllib/plugouts/BasePlugout.pm

@@ -94 +94 @@
          'reqd' => 'no',
          'hiddengli' => 'yes'},
+       { 'name' => 'rss_title',
+         'desc' => "{BasPlugout.rss_title}",
+         'type' => 'string',
+     'deft' => 'dc.Title',
+         'reqd' => 'no',
+         'hiddengli' => 'yes'},
     { 'name' => "no_auxiliary_databases",
       'desc' => "{BasPlugout.no_auxiliary_databases}",
@@ -908 +914 @@
       if (($oid_files->{'index-status'} eq "I") || ($oid_files->{'index-status'} eq "R")) {
     my $top_section = $doc_obj->get_top_section();
-
-    my $titles = $doc_obj->get_metadata($top_section,"dc.Title");
-
-    # if there is no dc.Title for this document, then use ex.Title as fallback
-    if(scalar(@$titles) == 0) {
-        $titles = $doc_obj->get_metadata($top_section,"Title");
+    
+    # rss_title can be set in collect.cfg as follows: 
+    #      plugout GreenstoneXMLPlugout -rss_title "dc.Title; ex.Title"
+    # rss_title is a semi-colon or comma-separated list of the metadata field names that should
+    # be consulted in order to obtain a Title (anchor text) for the RSS document link.   
+    # If not specified, rss_title will default to dc.Title, and fall back on Untitled
+    my $metafieldnames = $self->{'rss_title'};
+    my @metafieldarray = split(/[,;] ?/,$metafieldnames); # , or ; separator can be followed by an optional space
+    my $titles;
+    #@$titles=(); # at worst @$titles will be (), as get_metadata(dc.Titles) may return ()
+    foreach my $metafieldname (@metafieldarray) {
+        $metafieldname =~ s@^ex\.@@; # if ex.Title, need to get_metadata() on metafieldname=Title
+        $titles = $doc_obj->get_metadata($top_section,$metafieldname);
+
+        if(scalar(@$titles) != 0) { # found at least one title for one metafieldname
+           last; # break out of the loop
+        }
+    }
+    
+    # if ex.Title was listed in the metafieldnames, then we'll surely have a value for title for this doc
+    # otherwise, if we have no titles at this point, add in a default of Untitled as this doc's title
+    if(scalar(@$titles) == 0) { #&& $metafieldnames !~ [email protected]@) {
+        push(@$titles, "Untitled");
+    }
+    
+    # encode basic html entities like <>"& in the title(s), since the & char can break RSS links
+    for (my $i = 0; $i < scalar(@$titles); $i++) {
+        &ghtml::htmlsafe(@$titles[$i]);
     }
 

main/trunk/greenstone2/perllib/strings.properties

@@ -1353 +1353 @@
 BasPlugout.subdir_split_length:The maximum number of characters before spliting an archives subdirectory.
 BasPlugout.no_rss:Suppress the automatic generation of RSS feed file.
+BasPlugout.rss_title:Comma separated list of metadata fields listed in order of preference from which the title for a document's RSS link is to be obtained
 
 DSpacePlugout.desc:DSpace Archive format.