Changeset 29816 for main

Timestamp:

2015-03-30T13:42:14+13:00 (9 years ago)

Author:

kjdon

Message:

removing debug and commented out test lines

File:

• main/trunk/greenstone2/perllib/util.pm (modified) (2 diffs)

main/trunk/greenstone2/perllib/util.pm

@@ -154 +154 @@
     require Win32;
 
-    print STDERR "***** utf8 filename = $utf8_filename\n\n\n";
-
     my $unicode_filename = decode("utf8",$utf8_filename);
     $real_filename = Win32::GetShortPathName($unicode_filename);
@@ -171 +169 @@
         
     my $unicode_filename = $raw_file;
-    print STDERR "&&&&&&&&&&&&&& raw file = $unicode_filename ".&unicode::debug_unicode_string($unicode_filename)."\n";
     if (($ENV{'GSDLOS'} =~ m/^windows$/i) && ($^O ne "cygwin")) {
-        print STDERR "windows, upgrading...\n";
-        $unicode_filename = &util::get_dirsep_tail(&util::upgrade_if_dos_filename(&FileUtils::filenameConcatenate($directory, $raw_file), 0));
-            
+        # Try turning a short version to the long version
+        # If there are "funny" characters in the file name, that can't be represented in the ANSI code, then we will have a short weird version, eg E74~1.txt
+        $unicode_filename = &util::get_dirsep_tail(&util::upgrade_if_dos_filename(&FileUtils::filenameConcatenate($directory, $raw_file), 0));
+        
+        
+        if ($unicode_filename eq $raw_file) {
+        # This means the original filename *was* able to be encoded in the local ANSI file encoding (eg windows_1252), so now we turn it back to perl's unicode
         
-            if ($unicode_filename eq $raw_file) {
-                print STDERR "long = original\n";
-                $unicode_filename = &Encode::decode(locale_fs => $unicode_filename);
-                } else {
-                    print STDERR "long not equal original\n";
-                    #$unicode_filename = &Encode::decode("UTF-16le" => $unicode_filename);
-                    }
+        $unicode_filename = &Encode::decode(locale_fs => $unicode_filename);
+        }
+        # else This means we did have one of the funny filenames. the getLongPathName (used in upgrade_if_dos_filename) will return unicode, so we don't need to do anything more.
+        
                     
     } else {
-        print STDERR "not windows, not upgrading. decoding in utf8\n";
-        $unicode_filename = &Encode::decode("utf-8", $unicode_filename);
+        # we had a utf-8 string, turn it into perl internal unicode
+        $unicode_filename = &Encode::decode("utf-8", $unicode_filename);
     
         
-        }
-    #
-    #$unicode_filename = &Encode::decode("ascii", $unicode_filename);
-    #$unicode_filename = &Encode::decode(locale_fs => $unicode_filename);
-    print STDERR "after decoding = $unicode_filename ".&unicode::debug_unicode_string($unicode_filename)."\n";
-    #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: $unicode_filename\n";
+    }
+    #Does the filename have url encoded chars in it?
+    if (&unicode::is_url_encoded($unicode_filename)) {
+        $unicode_filename = &unicode::url_decode($unicode_filename);
+    }
+    
     # 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'
-       }
+        $unicode_filename = normalize('C', $unicode_filename); # Composed form 'C'
+    }
     return $unicode_filename;