Changeset 29794

Timestamp:

2015-03-19T14:07:11+13:00 (9 years ago)

Author:

kjdon

Message:

trying to get a proper unicode version of the filename. Need to convert to long windows filenames if on windows, decode any %XX url codes, need to change to composed form if on macs

File:

• main/trunk/greenstone2/perllib/util.pm (modified) (1 diff)

main/trunk/greenstone2/perllib/util.pm

@@ -164 +164 @@
 }
 
+sub raw_filename_to_unicode
+{
+    my ($directory, $raw_file) = @_;
+        
+    my $unicode_filename = &util::get_dirsep_tail(&util::upgrade_if_dos_filename(&FileUtils::filenameConcatenate($directory, $raw_file), 1));
+    $unicode_filename = &Encode::decode("utf8", $unicode_filename);
+    #$unicode_filename = &Encode::decode(locale => $unicode_filename);
+
+    #Does the filenaem have url encoded chars in it?
+        if (&unicode::is_url_encoded($unicode_filename)) {
+            $unicode_filename = &unicode::url_decode($unicode_filename);
+            }
+    print STDERR "after url decoding: $unicode_filename\n";
+    # Normalise the filename to canonical composition - on mac, filenames use decopmposed form for accented chars
+    if ($ENV{'GSDLOS'} =~ m/^darwin$/i) {
+       $unicode_filename = normalize('C', $unicode_filename); # Composed form 'C'
+       }
+    return $unicode_filename;
+
+}
 sub fd_exists {
   warnings::warnif("deprecated", "util::fd_exists() is deprecated, using FileUtils::fileTest() instead");