Changeset 16578 for gsdl/trunk/perllib/doc.pm

Timestamp:

2008-07-29T13:07:59+12:00 (16 years ago)

Author:

ak19

Message:

1. Base64 encoded gsdlsourcefilename to preserve original filename. 2. Both subroutines set_source_filename and set_converted_filename now call set_utf8_metadata_element. 3. Subroutine add_utf8_metadata checks to see that the meta is utf8 (or else tries to make it utf8) before adding it in.

File:

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

gsdl/trunk/perllib/doc.pm

@@ -39 +39 @@
 use ghtml;
 use File::stat;
+use MIME::Base64;
 ##use hashdoc;
 use docprint;
@@ -145 +146 @@
     my ($source_filename) = @_;
 
-    $self->set_metadata_element ($self->get_top_section(), 
+    # Since the gsdlsourcefilename element goes into the doc.xml it has 
+    # to be utf8. However, it should also *represent* the source filename 
+    # (in the import directory) which may not be utf8 at all. 
+    # For instance, if this meta element (gsdlsourcefilename) will be 
+    # used by other applications that parse doc.xml in order to locate 
+    # gsdlsourcefilename. Therefore, the solution is to encode the real
+    # filename into Base64 encoding which is a binary-to-text encoding, 
+    # meaning that the resulting string is ASCII (utf8). See also 
+    # MIME#Content-Transfer-Encoding in Wiki.
+    
+    print STDERR "******Base64 encoding the non-utf8 gsdl_source_filename $source_filename ";
+    
+    # remove "import/" from path, base 64 encode just the source_filename 
+    # (remove trailing spaces and newlines) and prefix "import/" again. 
+    my $import_dir = "import";
+    my $srcfilename = $source_filename;
+    $srcfilename =~ s/^$import_dir(?:\\|\/)+(.*)$/$1/;
+    
+    print STDERR "-> $srcfilename -> ";
+    $srcfilename = &MIME::Base64::encode_base64($srcfilename);
+    $srcfilename =~ s/\s*//sg; # for some reason it adds spaces not just at end but also in middle
+        
+    $source_filename = &util::filename_cat($import_dir, $srcfilename);
+    print STDERR "$source_filename\n";
+    
+    $self->set_utf8_metadata_element ($self->get_top_section(), 
                  "gsdlsourcefilename", 
                  $source_filename);
@@ -154 +180 @@
     my ($converted_filename) = @_;
 
-    $self->set_metadata_element ($self->get_top_section(), 
+    # we know the converted filename is utf8
+    $self->set_utf8_metadata_element ($self->get_top_section(), 
                  "gsdlconvertedfilename", 
                  $converted_filename);
@@ -826 +853 @@
     #print STDERR "###$field=$value\n";
     # double check that the value is utf-8
-    if (unicode::ensure_utf8(\$value)) {
-    print STDERR "doc::add_utf8_metadata: warning: '$field' wasn't utf8\n";
+    if (!&unicode::check_is_utf8($value)) {
+    print STDERR "doc::add_utf8_metadata - warning: '$field''s value $value wasn't utf8. ";
+    &unicode::ensure_utf8(\$value);
+    print STDERR "Tried converting to utf8: $value\n";
     }