Changeset 18320 for gsdl/trunk

Timestamp:

2009-01-06T19:03:58+13:00 (15 years ago)

Author:

ak19

Message:

Now plugins provide the option of base64 encoding or url encoding filenames that are to be renamed (when copied into the archives dir). Previously renamed files would always be url-encoded. URL-encoding is the default now for most plugins except MP3Plugin and OggVorbisPlugin, where the default is base64 encoding. Base64 encoding filenames upon renaming them was introduced so that more files that browsers try to open in external applications can open them, since url encoding does not seem to be implemented the same everywhere (for instance, windows media player is unable to handle url-encoded wmv filenames when such files are launched in it through the browser).

Location:

gsdl/trunk/perllib/plugins

Files:

• BasePlugin.pm (modified) (10 diffs)
• ConvertBinaryFile.pm (modified) (2 diffs)
• HTMLPlugin.pm (modified) (2 diffs)
• ImagePlugin.pm (modified) (1 diff)
• MP3Plugin.pm (modified) (2 diffs)
• OggVorbisPlugin.pm (modified) (2 diffs)
• ReadTextFile.pm (modified) (1 diff)

gsdl/trunk/perllib/plugins/BasePlugin.pm

@@ -44 +44 @@
     @BasePlugin::ISA = ( 'PrintInfo' );
 }
+
+# the different methods that can be applied when renaming
+# imported documents and their associated files
+our $file_rename_method_list = 
+    [ { 'name' => "url",
+    'desc' => "{BasePlugin.rename_method.url}" },
+      { 'name' => "base64",
+    'desc' => "{BasePlugin.rename_method.base64}" } ];
 
 our $encoding_list =
@@ -126 +134 @@
     'deft' => "dc.Identifier",
     'reqd' => "no",
-    'modegli' => "2" },
-
+    'modegli' => "2" },      
 #      { 'name' => "use_as_doc_identifier",
 #   'desc' => "{BasePlugin.use_as_doc_identifier}",
@@ -133 +140 @@
 #   'reqd' => "no" ,
 #   'deft' => "" } ,
-     { 'name' => "no_cover_image",
+      { 'name' => "no_cover_image",
     'desc' => "{BasePlugin.no_cover_image}",
     'type' => "flag",
@@ -147 +154 @@
         'type' => "flag",
         'reqd' => "no",
-    'hiddengli' => "yes" } # deprecated, but leave in for old collections
-
+    'hiddengli' => "yes" }, # deprecated, but leave in for old collections
+      { 'name' => "file_rename_method",
+    'desc' => "{BasePlugin.file_rename_method}",
+    'type' => "enum",
+    'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
+    'list' => $file_rename_method_list,
+    'reqd' => "no"
+      }
       
       ];
@@ -158 +171 @@
         'inherits' => "no",
         'args'     => $arguments };
-
 
 sub new {
@@ -346 +358 @@
 
     return "";
+}
+
+# rename imported files and assoc files using URL encoding by default
+# as this will work for most plugins and give more legible filenames
+sub get_default_file_rename_method() {
+    my $self = shift (@_);
+    return "url";
 }
 
@@ -566 +585 @@
 
 # gets the filename with no path, converts to utf8, and then dm safes it.
-#filename_encoding set by user
+# filename_encoding set by user
 sub filename_to_utf8_metadata
 {
@@ -732 +751 @@
     my ($doc_obj, $filename_no_path, $file_encoding) = @_;
 
-
     my $top_section = $doc_obj->get_top_section();
     
     # UTF-8 version of filename
     my $filemeta = $self->filename_to_utf8_metadata($filename_no_path, $file_encoding);
-
+    
     # Source is the UTF8 display name - not necessarily the name of the file on the system
     $doc_obj->set_utf8_metadata_element($top_section, "Source", $filemeta); 
 
-    # SourceFile is the url-reference to url-encoded filemeta: 
-    # it's a url that refers to the actual file on the system
-    $filemeta = &unicode::url_encode($filemeta);
-    $filemeta = &unicode::filename_to_url($filemeta);
+    $filemeta = &util::rename_file($filemeta, $self->{'file_rename_method'});
+    # If using URL encoding, then SourceFile is the url-reference to url-encoded
+    # filemeta: it's a url that refers to the actual file on the system
+    $filemeta = &unicode::filename_to_url($filemeta, $self->{'file_rename_method'});
+
     $doc_obj->set_utf8_metadata_element($top_section, "SourceFile", $filemeta);
 }
@@ -797 +816 @@
     my ($filename_full_path, $filename_no_path) = &util::get_full_filenames($base_dir, $file);
     # create a new document
-    my $doc_obj = new doc ($filename_full_path, "indexed_doc");
+    my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
     my $top_section = $doc_obj->get_top_section();
 
@@ -1004 +1023 @@
         
         # we need to make sure the filename is valid utf-8 - we do 
-        # this by url encoding it
+        # this by url or base64 encoding it
         # $tail_filename is the name that we store the file as
-        $tail_filename = &unicode::url_encode($tail_filename);
+        $tail_filename = &util::rename_file($tail_filename, $self->{'file_rename_method'});
         $doc_obj->associate_file($full_filename,$tail_filename,$mimetype);
-        # then we need to encode the % signs in the filename, so that 
-        # it works in a url
+        # If the filename is url_encoded, we need to encode the % signs 
+        # in the filename, so that it works in a url
         my $url_tail_filename = &unicode::filename_to_url($tail_filename);
         # work out extended tail extension (i.e. matching tail re)

gsdl/trunk/perllib/plugins/ConvertBinaryFile.pm

@@ -284 +284 @@
     # to work on Windows (IE or Firefox), as browsers are looking for filesystem-encoded
     # files on the filesystem.
-    $tailname = &unicode::url_encode($tailname);
+    $tailname = &util::rename_file($tailname, $self->{'file_rename_method'});
 
     $suffix = lc($suffix);
@@ -429 +429 @@
     my $collect_file = &util::filename_within_collection($filename_full_path);
     my $collect_conv_file = &util::filename_within_collection($conv_filename);
-    $doc_obj->set_source_filename ($collect_file); 
+    $doc_obj->set_source_filename ($collect_file, $self->{'file_rename_method'}); 
     $doc_obj->set_converted_filename($collect_conv_file);
 

gsdl/trunk/perllib/plugins/HTMLPlugin.pm

@@ -635 +635 @@
 # Given a filename in any encoding, will URL decode it to get back the original filename
 # in the original encoding. Because this method is intended to work out the *original* 
-# filename*, it not URL decode any filename if a file by the name of the *URL-encoded*
+# filename*, it does not URL decode any filename if a file by the name of the *URL-encoded*
 # string already exists in the local folder.
 # Return the original filename corresponding to the parameter URL-encoded filename, and
@@ -1140 +1140 @@
     }
 
-    # Make sure this name uses only ASCII characters 
-    # => use URL encoding, as it preserves original encoding
-    $newname = &unicode::url_encode($newname);
+    # Make sure this name uses only ASCII characters. 
+    # We use either base64 or URL encoding, as these preserve original encoding
+    $newname = &util::rename_file($newname, $self->{'file_rename_method'});
 
     $doc_obj->associate_file($filename, $newname, undef, $section);

gsdl/trunk/perllib/plugins/ImagePlugin.pm

@@ -114 +114 @@
     if ($self->{'image_conversion_available'} == 1) {
     my $utf8_filename_no_path = $self->filepath_to_utf8($filename_no_path);
-    my $url_encoded_filename = &unicode::url_encode($utf8_filename_no_path);
+    my $url_encoded_filename = &util::rename_file($utf8_filename_no_path, $self->{'file_rename_method'});
     $self->generate_images($filename_full_path, $url_encoded_filename, $doc_obj, $doc_obj->get_top_section()); # should we check the return value?
     } else {

gsdl/trunk/perllib/plugins/MP3Plugin.pm

@@ -60 +60 @@
     'desc' => "{MP3Plugin.metadata_fields}",
     'type' => "string",
-    'deft' => "Title,Artist,Genre" } ];
+    'deft' => "Title,Artist,Genre" },
+      { 'name' => "file_rename_method",
+    'desc' => "{BasePlugin.file_rename_method}",
+    'type' => "enum",
+    'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
+    'list' => $BasePlugin::file_rename_method_list,
+    'reqd' => "no"
+      } ];
 
 my $options = { 'name'     => "MP3Plugin",
@@ -83 +90 @@
 sub get_default_process_exp {
     return q^(?i)\.mp3$^;
+}
+
+# rename imported files and assoc files using base64 encoding by default
+# so that the urls generated will always work with media files even when
+# opened in external applications (when wmv file names contain spaces in
+# them and they get url-encoded, wmv player fails to open the doubly url-
+# encoded url reference to the url-encoded filename). 
+sub get_default_file_rename_method() {
+    return "base64";
 }
 

gsdl/trunk/perllib/plugins/OggVorbisPlugin.pm

@@ -51 +51 @@
     'desc' => "{OggVorbisPlugin.add_technical_metadata}",
     'type' => "flag",
-    'deft' => "" } ];
+    'deft' => "" }, 
+      { 'name' => "file_rename_method",
+    'desc' => "{BasePlugin.file_rename_method}",
+    'type' => "enum",
+    'deft' => &get_default_file_rename_method(), # by default rename imported files and assoc files using this encoding
+    'list' => $BasePlugin::file_rename_method_list,
+    'reqd' => "no"
+      } ];
 
 my $options = { 'name'     => "OggVorbisPlugin",
@@ -66 +73 @@
 }
 
+# rename imported media files using base64 encoding by default
+# so that the urls generated will always work with external apps
+sub get_default_file_rename_method() {
+    return "base64";
+}
 
 sub new

gsdl/trunk/perllib/plugins/ReadTextFile.pm

@@ -133 +133 @@
 
     # create a new document
-    my $doc_obj = new doc ($filename_full_path, "indexed_doc");
+    my $doc_obj = new doc ($filename_full_path, "indexed_doc", $self->{'file_rename_method'});
     my $top_section = $doc_obj->get_top_section();