Title-related changes, better sanitization (bug 1351)

src/gpodder/model.py

@@ -849,9 +849,18 @@ class PodcastChannel(PodcastModelObject):
         """
         return self.EpisodeClass.create_from_dict(d, self)
 
+    def _consume_updated_title(self, new_title):
+        # Replace multi-space and newlines with single space (Maemo bug 11173)
+        new_title = re.sub('\s+', ' ', new_title).strip()
+
+        # Only update the podcast-supplied title when we
+        # don't yet have a title, or if the title is the
+        # feed URL (e.g. we didn't find a title before).
+        if not self.title or self.title == self.url:
+            self.title = new_title
+
     def _consume_custom_feed(self, custom_feed, max_episodes=0):
-        if not self.title:
-            self.title = custom_feed.get_title()
+        self._consume_updated_title(custom_feed.get_title())
         self.link = custom_feed.get_link()
         self.description = custom_feed.get_description()
         self.cover_url = custom_feed.get_image()
@@ -874,9 +883,7 @@ class PodcastChannel(PodcastModelObject):
     def _consume_updated_feed(self, feed, max_episodes=0, mimetype_prefs=''):
         #self.parse_error = feed.get('bozo_exception', None)
 
-        # Replace multi-space and newlines with single space (Maemo bug 11173)
-        if not self.title:
-            self.title = re.sub('\s+', ' ', feed.feed.get('title', self.url))
+        self._consume_updated_title(feed.feed.get('title', self.url))
 
         self.link = feed.feed.get('link', self.link)
         self.description = feed.feed.get('subtitle', self.description)

src/gpodder/util.py

@@ -42,6 +42,7 @@ import stat
 import shlex
 import socket
 import sys
+import string
 
 import re
 import subprocess
@@ -94,6 +95,20 @@ if encoding is None:
         logger.info('Assuming encoding: ISO-8859-15 ($LANG not set).')
 
 
+# Filename / folder name sanitization
+def _sanitize_char(c):
+    if c in string.whitespace:
+        return ' '
+    elif c in ',-.()':
+        return c
+    elif c in string.punctuation or ord(c) <= 31:
+        return '_'
+
+    return c
+
+SANITIZATION_TABLE = ''.join(map(_sanitize_char, map(chr, range(256))))
+del _sanitize_char
+
 # Used by file_type_by_extension()
 _BUILTIN_FILE_TYPES = None
 
@@ -1131,12 +1146,6 @@ def sanitize_filename(filename, max_length=0, use_ascii=False):
     If use_ascii is True, don't encode in the native language,
     but use only characters from the ASCII character set.
     """
-    global encoding
-    if use_ascii:
-        e = 'ascii'
-    else:
-        e = encoding
-
     if not isinstance(filename, unicode):
         filename = filename.decode(encoding, 'ignore')
 
@@ -1145,7 +1154,11 @@ def sanitize_filename(filename, max_length=0, use_ascii=False):
                 filename, max_length)
         filename = filename[:max_length]
 
-    return re.sub('[/|?*<>:+\[\]\"\\\]', '_', filename.strip().encode(e, 'ignore'))
+    filename = filename.encode('ascii' if use_ascii else encoding, 'ignore')
+    filename = filename.translate(SANITIZATION_TABLE)
+    filename = filename.strip('.' + string.whitespace)
+
+    return filename
 
 
 def find_mount_point(directory):