modified filename escaping to a "smarter" one

diff --git a/test/test_utils.py b/test/test_utils.py
index eb0af703f502b7fa80761164fab2c5423d7f1035..87884012040bd71f33cc6274ad7423a6c635c2d7 100644 (file)
--- a/test/test_utils.py
+++ b/test/test_utils.py
@@ -22,13 +22,13 @@ class TestUtil(unittest.TestCase):
 
                self.assertEqual(sanitize_filename(u'123'), u'123')
 
-               self.assertEqual(u'abc_de', sanitize_filename(u'abc/de'))
-               self.assertTrue(u'de' in sanitize_filename(u'abc/de'))
+               self.assertEqual(u'abc-de', sanitize_filename(u'abc/de'))
                self.assertFalse(u'/' in sanitize_filename(u'abc/de///'))
 
-               self.assertEqual(u'abc_de', sanitize_filename(u'abc\\de'))
-               self.assertEqual(u'abc_de', sanitize_filename(u'abc\\de'))
-               self.assertTrue(u'de' in  sanitize_filename(u'abc\\de'))
+               self.assertEqual(u'abc-de', sanitize_filename(u'abc/<>\\*|de'))
+               self.assertEqual(u'xxx', sanitize_filename(u'xxx/<>\\*|'))
+               self.assertEqual(u'yes no', sanitize_filename(u'yes? no'))
+               self.assertEqual(u'this - that', sanitize_filename(u'this: that'))
 
                self.assertEqual(sanitize_filename(u'ä'), u'ä')
                self.assertEqual(sanitize_filename(u'кириллица'), u'кириллица')

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 839da17d0763c6f41373856762d28f198ae94821..a64937b4c12077fef92c95d0ddd36e30de0e8917 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -194,10 +194,20 @@ def timeconvert(timestr):
 def sanitize_filename(s):
        """Sanitizes a string so it could be used as part of a filename."""
        def replace_insane(char):
-               if char in u' .\\/|?*<>:"' or ord(char) < 32:
-                       return '_'
+               if char == '?' or ord(char) < 32 or ord(char) == 127:
+                       return ''
+               elif char == '"':
+                       return '\''
+               elif char == ':':
+                       return ' -'
+               elif char in '\\/|*<>':
+                       return '-'
                return char
-       return u''.join(map(replace_insane, s)).strip('_')
+
+       result = u''.join(map(replace_insane, s))
+       while '--' in result:
+               result = result.replace('--', '-')
+       return result.strip('-')
 
 def orderedSet(iterable):
        """ Remove all duplicates from the input iterable """