Changeset 24016 for main/trunk/greenstone2/perllib/classify/List.pm

Timestamp:

2011-05-12T17:37:36+12:00 (13 years ago)

Author:

ak19

Message:

Fixed the unicode-issue that Diego discovered. When using the List classifier, the bookshelf titles were displayed without any special/accented characters, whereas, when the same titles appeared in documents under such bookshelves, the document titles were correctly displayed. Turns out the problem was from utf8-to-unicode and unicode-to-utf8 calls that were not removed after the changes to GS perlcode for making string unicode aware in Perl. So such conversions are already handled elsewhere in the code and no longer belong in List, which fixed the problem.

File:

• main/trunk/greenstone2/perllib/classify/List.pm (modified) (3 diffs)

main/trunk/greenstone2/perllib/classify/List.pm

@@ -383 +383 @@
             # uppercase the metadata - makes the AZList nicer
             $metadata_value = uc($metadata_value);
-            # Convert the metadata value from a UTF-8 string to a Unicode string
-            # This means that length() and substr() work properly
-            # We need to be careful to convert classifier node title values back to UTF-8, however
-            my $metadata_value_unicode_string = $self->convert_utf8_string_to_unicode_string($metadata_value);
+            
+            # We are already working with unicode aware strings at this 
+            # stage, so we no longer need to convert from utf8 to unicode 
+            my $metadata_value_unicode_string = $metadata_value; # $self->convert_utf8_string_to_unicode_string($metadata_value);
 
             # Add the metadata value into the list for this combination of metadata group and section
@@ -762 +762 @@
 
     # Create an hlist partition
-    my %child_classifier_node = ( 'Title' => $self->convert_unicode_string_to_utf8_string($partitionname),
+    # Note that we don't need to convert from unicode-aware strings 
+    # to utf8 here, as that is handled elsewhere in the code
+    my %child_classifier_node = ( 'Title' => $partitionname, #'Title' => $self->convert_unicode_string_to_utf8_string($partitionname),
                   'childtype' => "VList",
                   'contains' => [] );
@@ -803 +805 @@
     # Otherwise create a sublist (bookshelf) for the metadata value
     else {
-        my %child_classifier_node = ( 'Title' => $self->convert_unicode_string_to_utf8_string($metadata_value),
+        # Note that we don't need to convert from unicode-aware strings 
+        # to utf8 here, as that is handled elsewhere in the code
+        my %child_classifier_node = ( 'Title' => $metadata_value, # 'Title' => $self->convert_unicode_string_to_utf8_string($metadata_value),
                       'childtype' => "VList",
                       'mdtype' => $metadata_group,