Changeset 4791

Timestamp:

2003-06-25T09:44:33+12:00 (21 years ago)

Author:

davidb

Message:

Modified so -input_encoding flag used.

File:

• trunk/gsdl/perllib/plugins/MARCPlug.pm (modified) (4 diffs)

trunk/gsdl/perllib/plugins/MARCPlug.pm

@@ -159 +159 @@
 
 
+# The bulk of this function is based on read_line in multiread.pm
+# Unable to use read_line original because it expects to get its input
+# from a file.  Here the line to be converted is passed in as a string
+
+sub to_utf8
+{
+    my $self = shift (@_);
+    my ($encoding, $line) = @_;
+
+    if ($encoding eq "iso_8859_1") {
+    # we'll use ascii2utf8() for this as it's faster than going
+    # through convert2unicode()
+    return &unicode::ascii2utf8 (\$line);
+    }
+
+    # everything else uses unicode::convert2unicode
+    return &unicode::unicode2utf8 (&unicode::convert2unicode ($encoding, \$line));
+}
+
+
 sub read_file {
     my $self = shift (@_);
     my ($filename, $encoding, $language, $textref) = @_;
+
+    $self->{'readfile_encoding'} = $encoding;
 
     @marc_entries = ();
@@ -214 +236 @@
     $$textref =~ s/>/>/g;
 
+    my $encoding = $self->{'readfile_encoding'};
+    $$textref = $self->to_utf8($encoding,$$textref);
+
     print $outhandle "  Adding Marc Record:\n",substr($$textref,0,40), " ...\n"
     if $self->{'verbosity'} > 2;
@@ -231 +256 @@
     my ($marc, $metadata, $doc_obj, $section) = @_;
     my $outhandle = $self->{'outhandle'};
+
+    my $encoding = $self->{'readfile_encoding'};
 
     my $metadata_mapping = $self->{'metadata_mapping'};
@@ -245 +272 @@
         foreach $metavalue ( @metavalues )
         {
-        $doc_obj->add_utf8_metadata ($thissection, $metaname, $metavalue->as_string());
+        $metavalue_str = $self->to_utf8($encoding,$metavalue->as_string());
+        $doc_obj->add_utf8_metadata ($thissection, $metaname, $metavalue_str);
         }
     }