New optoin --restrict-filenames

README.md

@@ -47,6 +47,8 @@ which means you can modify it, redistribute it or use it however you like.
                              %(extractor)s for the provider (youtube, metacafe,
                              etc), %(id)s for the video id and %% for a literal
                              percent. Use - to output to stdout.
+    --restrict-filenames     Avoid some characters such as "&" and spaces in
+                             filenames
     -a, --batch-file FILE    file containing URLs to download ('-' for stdin)
     -w, --no-overwrites      do not overwrite files
     -c, --continue           resume partially downloaded files

test/test_utils.py

@@ -30,11 +30,34 @@ class TestUtil(unittest.TestCase):
 		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'AT&T'), u'AT&T')
 		self.assertEqual(sanitize_filename(u'ä'), u'ä')
 		self.assertEqual(sanitize_filename(u'кириллица'), u'кириллица')
 
-		for forbidden in u'"\0\\/':
+		forbidden = u'"\0\\/'
-			self.assertTrue(forbidden not in sanitize_filename(forbidden))
+		for fc in forbidden:
+			for fbc in forbidden:
+				self.assertTrue(fbc not in sanitize_filename(fc))
+
+	def test_sanitize_filename_restricted(self):
+		self.assertEqual(sanitize_filename(u'abc', restricted=True), u'abc')
+		self.assertEqual(sanitize_filename(u'abc_d-e', restricted=True), u'abc_d-e')
+
+		self.assertEqual(sanitize_filename(u'123', restricted=True), u'123')
+
+		self.assertEqual(u'abc-de', sanitize_filename(u'abc/de', restricted=True))
+		self.assertFalse(u'/' in sanitize_filename(u'abc/de///', restricted=True))
+
+		self.assertEqual(u'abc-de', sanitize_filename(u'abc/<>\\*|de', restricted=True))
+		self.assertEqual(u'xxx', sanitize_filename(u'xxx/<>\\*|', restricted=True))
+		self.assertEqual(u'yes_no', sanitize_filename(u'yes? no', restricted=True))
+		self.assertEqual(u'this_-_that', sanitize_filename(u'this: that', restricted=True))
+
+		forbidden = u'"\0\\/&: \'\t\n'
+		for fc in forbidden:
+			print('input: ' + fc + ', result: ' + repr(sanitize_filename(fc, restricted=True)))
+			for fbc in forbidden:
+				self.assertTrue(fbc not in sanitize_filename(fc, restricted=True))
 
 	def test_ordered_set(self):
 		self.assertEqual(orderedSet([1,1,2,3,4,4,5,6,7,3,5]), [1,2,3,4,5,6,7])

youtube-dl.1

@@ -59,6 +59,8 @@ redistribute it or use it however you like.
 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ %(extractor)s\ for\ the\ provider\ (youtube,\ metacafe,
 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ etc),\ %(id)s\ for\ the\ video\ id\ and\ %%\ for\ a\ literal
 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ percent.\ Use\ -\ to\ output\ to\ stdout.
+--restrict-filenames\ \ \ \ \ Avoid\ some\ characters\ such\ as\ "&"\ and\ spaces\ in
+\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ filenames
 -a,\ --batch-file\ FILE\ \ \ \ file\ containing\ URLs\ to\ download\ (\[aq]-\[aq]\ for\ stdin)
 -w,\ --no-overwrites\ \ \ \ \ \ do\ not\ overwrite\ files
 -c,\ --continue\ \ \ \ \ \ \ \ \ \ \ resume\ partially\ downloaded\ files
@@ -210,7 +212,7 @@ Please note that Python 2.5 is not supported anymore.
 .PP
 Since June 2012 (#342) youtube-dl is packed as an executable zipfile,
 simply unzip it (might need renaming to \f[C]youtube-dl.zip\f[] first on
-some systems) or clone the git repo to see the code.
+some systems) or clone the git repository, as laid out above.
 If you modify the code, you can run it by executing the
 \f[C]__main__.py\f[] file.
 To recompile the executable, run \f[C]make\ youtube-dl\f[].

youtube-dl.bash-completion

@@ -3,7 +3,7 @@ __youtube-dl()
     local cur prev opts
     COMPREPLY=()
     cur="${COMP_WORDS[COMP_CWORD]}"
-    opts="--all-formats --audio-format --audio-quality --auto-number --batch-file --console-title --continue --cookies --dump-user-agent --extract-audio --format --get-description --get-filename --get-format --get-thumbnail --get-title --get-url --help --id --ignore-errors --keep-video --list-extractors --list-formats --literal --match-title --max-downloads --max-quality --netrc --no-continue --no-mtime --no-overwrites --no-part --no-progress --output --password --playlist-end --playlist-start --prefer-free-formats --quiet --rate-limit --reject-title --retries --simulate --skip-download --srt-lang --title --update --user-agent --username --verbose --version --write-description --write-info-json --write-srt"
+    opts="--all-formats --audio-format --audio-quality --auto-number --batch-file --console-title --continue --cookies --dump-user-agent --extract-audio --format --get-description --get-filename --get-format --get-thumbnail --get-title --get-url --help --id --ignore-errors --keep-video --list-extractors --list-formats --literal --match-title --max-downloads --max-quality --netrc --no-continue --no-mtime --no-overwrites --no-part --no-progress --output --password --playlist-end --playlist-start --prefer-free-formats --quiet --rate-limit --reject-title --restrict-filenames --retries --simulate --skip-download --srt-lang --title --update --user-agent --username --verbose --version --write-description --write-info-json --write-srt"
 
     if [[ ${cur} == * ]] ; then
         COMPREPLY=( $(compgen -W "${opts}" -- ${cur}) )

youtube_dl/FileDownloader.py

@@ -44,37 +44,38 @@ class FileDownloader(object):
 
 	Available options:
 
-	username:         Username for authentication purposes.
+	username:          Username for authentication purposes.
-	password:         Password for authentication purposes.
+	password:          Password for authentication purposes.
-	usenetrc:         Use netrc for authentication instead.
+	usenetrc:          Use netrc for authentication instead.
-	quiet:            Do not print messages to stdout.
+	quiet:             Do not print messages to stdout.
-	forceurl:         Force printing final URL.
+	forceurl:          Force printing final URL.
-	forcetitle:       Force printing title.
+	forcetitle:        Force printing title.
-	forcethumbnail:   Force printing thumbnail URL.
+	forcethumbnail:    Force printing thumbnail URL.
-	forcedescription: Force printing description.
+	forcedescription:  Force printing description.
-	forcefilename:    Force printing final filename.
+	forcefilename:     Force printing final filename.
-	simulate:         Do not download the video files.
+	simulate:          Do not download the video files.
-	format:           Video format code.
+	format:            Video format code.
-	format_limit:     Highest quality format to try.
+	format_limit:      Highest quality format to try.
-	outtmpl:          Template for output names.
+	outtmpl:           Template for output names.
-	ignoreerrors:     Do not stop on download errors.
+	restrictfilenames: Do not allow "&" and spaces in file names
-	ratelimit:        Download speed limit, in bytes/sec.
+	ignoreerrors:      Do not stop on download errors.
-	nooverwrites:     Prevent overwriting files.
+	ratelimit:         Download speed limit, in bytes/sec.
-	retries:          Number of times to retry for HTTP error 5xx
+	nooverwrites:      Prevent overwriting files.
-	continuedl:       Try to continue downloads if possible.
+	retries:           Number of times to retry for HTTP error 5xx
-	noprogress:       Do not print the progress bar.
+	continuedl:        Try to continue downloads if possible.
-	playliststart:    Playlist item to start at.
+	noprogress:        Do not print the progress bar.
-	playlistend:      Playlist item to end at.
+	playliststart:     Playlist item to start at.
-	matchtitle:       Download only matching titles.
+	playlistend:       Playlist item to end at.
-	rejecttitle:      Reject downloads for matching titles.
+	matchtitle:        Download only matching titles.
-	logtostderr:      Log messages to stderr instead of stdout.
+	rejecttitle:       Reject downloads for matching titles.
-	consoletitle:     Display progress in console window's titlebar.
+	logtostderr:       Log messages to stderr instead of stdout.
-	nopart:           Do not use temporary .part files.
+	consoletitle:      Display progress in console window's titlebar.
-	updatetime:       Use the Last-modified header to set output file timestamps.
+	nopart:            Do not use temporary .part files.
-	writedescription: Write the video description to a .description file
+	updatetime:        Use the Last-modified header to set output file timestamps.
-	writeinfojson:    Write the video description to a .info.json file
+	writedescription:  Write the video description to a .description file
-	writesubtitles:   Write the video subtitles to a .srt file
+	writeinfojson:     Write the video description to a .info.json file
-	subtitleslang:    Language of the subtitles to download
+	writesubtitles:    Write the video subtitles to a .srt file
+	subtitleslang:     Language of the subtitles to download
 	"""
 
 	params = None
@@ -349,7 +350,7 @@ class FileDownloader(object):
 	def process_info(self, info_dict):
 		"""Process a single dictionary returned by an InfoExtractor."""
 
-		info_dict['stitle'] = sanitize_filename(info_dict['title'])
+		info_dict['stitle'] = sanitize_filename(info_dict['title'], self.params.get('restrictfilenames'))
 
 		reason = self._match_entry(info_dict)
 		if reason is not None:

youtube_dl/__init__.py

@@ -272,6 +272,9 @@ def parseOpts():
 			help='number downloaded files starting from 00000', default=False)
 	filesystem.add_option('-o', '--output',
 			dest='outtmpl', metavar='TEMPLATE', help='output filename template. Use %(title)s to get the title, %(uploader)s for the uploader name, %(autonumber)s to get an automatically incremented number, %(ext)s for the filename extension, %(upload_date)s for the upload date (YYYYMMDD), %(extractor)s for the provider (youtube, metacafe, etc), %(id)s for the video id and %% for a literal percent. Use - to output to stdout.')
+	filesystem.add_option('--restrict-filenames',
+			action='store_true', dest='restrictfilenames',
+			help='Avoid some characters such as "&" and spaces in filenames', default=False)
 	filesystem.add_option('-a', '--batch-file',
 			dest='batchfile', metavar='FILE', help='file containing URLs to download (\'-\' for stdin)')
 	filesystem.add_option('-w', '--no-overwrites',
@@ -485,6 +488,7 @@ def _real_main():
 			or (opts.useid and u'%(id)s.%(ext)s')
 			or (opts.autonumber and u'%(autonumber)s-%(id)s.%(ext)s')
 			or u'%(id)s.%(ext)s'),
+		'restrictfilenames': opts.restrictfilenames,
 		'ignoreerrors': opts.ignoreerrors,
 		'ratelimit': opts.ratelimit,
 		'nooverwrites': opts.nooverwrites,

youtube_dl/utils.py

@@ -195,17 +195,21 @@ def timeconvert(timestr):
 		timestamp = email.utils.mktime_tz(timetuple)
 	return timestamp
 
-def sanitize_filename(s):
+def sanitize_filename(s, restricted=False):
-	"""Sanitizes a string so it could be used as part of a filename."""
+	"""Sanitizes a string so it could be used as part of a filename.
+	If restricted is set, use a stricter subset of allowed characters.
+	"""
 	def replace_insane(char):
 		if char == '?' or ord(char) < 32 or ord(char) == 127:
 			return ''
 		elif char == '"':
-			return '\''
+			return '' if restricted else 'FOO\''
 		elif char == ':':
-			return ' -'
+			return '_-' if restricted else ' -'
 		elif char in '\\/|*<>':
 			return '-'
+		if restricted and (char in '&\'' or char.isspace()):
+			return '_'
 		return char
 
 	result = u''.join(map(replace_insane, s))