Initiate new release.

[youtubedl] / youtube_dl / YoutubeDL.py

diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py
old mode 100644 (file)
new mode 100755 (executable)
index 2a078ad..74e4261
--- a/youtube_dl/YoutubeDL.py
+++ b/youtube_dl/YoutubeDL.py
@@ -1,12 +1,16 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 

-from __future__ import absolute_import
+from __future__ import absolute_import, unicode_literals

 
 import collections
 
 import collections

+import datetime

 import errno
 import io
 import errno
 import io

+import itertools

 import json
 import json

+import locale
+import operator

 import os
 import platform
 import re
 import os
 import platform
 import re


@@ -20,15 +24,22 @@ import traceback
 if os.name == 'nt':
     import ctypes
 
 if os.name == 'nt':
     import ctypes
 

-from .utils import (
+from .compat import (
+    compat_basestring,

     compat_cookiejar,
     compat_cookiejar,

+    compat_expanduser,

     compat_http_client,
     compat_http_client,

+    compat_kwargs,

     compat_str,
     compat_urllib_error,
     compat_urllib_request,
     compat_str,
     compat_urllib_error,
     compat_urllib_request,

+)
+from .utils import (
+    escape_url,

     ContentTooShortError,
     date_from_str,
     DateRange,
     ContentTooShortError,
     date_from_str,
     DateRange,

+    DEFAULT_OUTTMPL,

     determine_ext,
     DownloadError,
     encodeFilename,
     determine_ext,
     DownloadError,
     encodeFilename,


@@ -39,21 +50,38 @@ from .utils import (
     locked_file,
     make_HTTPS_handler,
     MaxDownloadsReached,
     locked_file,
     make_HTTPS_handler,
     MaxDownloadsReached,

+    PagedList,
+    parse_filesize,

     PostProcessingError,
     platform_name,
     preferredencoding,
     PostProcessingError,
     platform_name,
     preferredencoding,

+    render_table,

     SameFileError,
     sanitize_filename,
     SameFileError,
     sanitize_filename,

+    std_headers,

     subtitles_filename,
     takewhile_inclusive,
     UnavailableVideoError,
     url_basename,
     subtitles_filename,
     takewhile_inclusive,
     UnavailableVideoError,
     url_basename,

+    version_tuple,

     write_json_file,
     write_string,
     YoutubeDLHandler,
     write_json_file,
     write_string,
     YoutubeDLHandler,

+    prepend_extension,
+    args_to_str,
+    age_restricted,

 )
 )

+from .cache import Cache

 from .extractor import get_info_extractor, gen_extractors
 from .extractor import get_info_extractor, gen_extractors

-from .FileDownloader import FileDownloader
+from .downloader import get_suitable_downloader
+from .downloader.rtmp import rtmpdump_version
+from .postprocessor import (
+    FFmpegFixupM4aPP,
+    FFmpegFixupStretchedPP,
+    FFmpegMergerPP,
+    FFmpegPostProcessor,
+    get_postprocessor,
+)

 from .version import __version__
 
 
 from .version import __version__
 
 


@@ -90,6 +118,7 @@ class YoutubeDL(object):
     usenetrc:          Use netrc for authentication instead.
     verbose:           Print additional info to stdout.
     quiet:             Do not print messages to stdout.
     usenetrc:          Use netrc for authentication instead.
     verbose:           Print additional info to stdout.
     quiet:             Do not print messages to stdout.

+    no_warnings:       Do not print out anything for warnings.

     forceurl:          Force printing final URL.
     forcetitle:        Force printing title.
     forceid:           Force printing ID.
     forceurl:          Force printing final URL.
     forcetitle:        Force printing title.
     forceid:           Force printing ID.


@@ -98,8 +127,10 @@ class YoutubeDL(object):
     forcefilename:     Force printing final filename.
     forceduration:     Force printing duration.
     forcejson:         Force printing info_dict as JSON.
     forcefilename:     Force printing final filename.
     forceduration:     Force printing duration.
     forcejson:         Force printing info_dict as JSON.

+    dump_single_json:  Force printing the info_dict of the whole playlist
+                       (or video) as a single JSON line.

     simulate:          Do not download the video files.
     simulate:          Do not download the video files.

-    format:            Video format code.
+    format:            Video format code. See options.py for more information.

     format_limit:      Highest quality format to try.
     outtmpl:           Template for output names.
     restrictfilenames: Do not allow "&" and spaces in file names
     format_limit:      Highest quality format to try.
     outtmpl:           Template for output names.
     restrictfilenames: Do not allow "&" and spaces in file names


@@ -107,6 +138,8 @@ class YoutubeDL(object):
     nooverwrites:      Prevent overwriting files.
     playliststart:     Playlist item to start at.
     playlistend:       Playlist item to end at.
     nooverwrites:      Prevent overwriting files.
     playliststart:     Playlist item to start at.
     playlistend:       Playlist item to end at.

+    playlist_items:    Specific indices of playlist to download.
+    playlistreverse:   Download playlist items in reverse order.

     matchtitle:        Download only matching titles.
     rejecttitle:       Reject downloads for matching titles.
     logger:            Log messages to a logging.Logger instance.
     matchtitle:        Download only matching titles.
     rejecttitle:       Reject downloads for matching titles.
     logger:            Log messages to a logging.Logger instance.


@@ -115,18 +148,19 @@ class YoutubeDL(object):
     writeinfojson:     Write the video description to a .info.json file
     writeannotations:  Write the video annotations to a .annotations.xml file
     writethumbnail:    Write the thumbnail image to a file
     writeinfojson:     Write the video description to a .info.json file
     writeannotations:  Write the video annotations to a .annotations.xml file
     writethumbnail:    Write the thumbnail image to a file

+    write_all_thumbnails:  Write all thumbnail formats to files

     writesubtitles:    Write the video subtitles to a file
     writeautomaticsub: Write the automatic subtitles to a file
     allsubtitles:      Downloads all the subtitles of the video
                        (requires writesubtitles or writeautomaticsub)
     listsubtitles:     Lists all available subtitles for the video
     writesubtitles:    Write the video subtitles to a file
     writeautomaticsub: Write the automatic subtitles to a file
     allsubtitles:      Downloads all the subtitles of the video
                        (requires writesubtitles or writeautomaticsub)
     listsubtitles:     Lists all available subtitles for the video

-    subtitlesformat:   Subtitle format [srt/sbv/vtt] (default=srt)
+    subtitlesformat:   The format code for subtitles

     subtitleslangs:    List of languages of the subtitles to download
     keepvideo:         Keep the video file after post-processing
     daterange:         A DateRange object, download only if the upload_date is in the range.
     skip_download:     Skip the actual download of the video file
     cachedir:          Location of the cache files in the filesystem.
     subtitleslangs:    List of languages of the subtitles to download
     keepvideo:         Keep the video file after post-processing
     daterange:         A DateRange object, download only if the upload_date is in the range.
     skip_download:     Skip the actual download of the video file
     cachedir:          Location of the cache files in the filesystem.

-                       None to disable filesystem cache.
+                       False to disable filesystem cache.

     noplaylist:        Download single video instead of a playlist if in doubt.
     age_limit:         An integer representing the user's age in years.
                        Unsuitable videos for the given age are skipped.
     noplaylist:        Download single video instead of a playlist if in doubt.
     age_limit:         An integer representing the user's age in years.
                        Unsuitable videos for the given age are skipped.


@@ -144,15 +178,84 @@ class YoutubeDL(object):
                        again.
     cookiefile:        File name where cookies should be read from and dumped to.
     nocheckcertificate:Do not verify SSL certificates
                        again.
     cookiefile:        File name where cookies should be read from and dumped to.
     nocheckcertificate:Do not verify SSL certificates

+    prefer_insecure:   Use HTTP instead of HTTPS to retrieve information.
+                       At the moment, this is only supported by YouTube.

     proxy:             URL of the proxy server to use
     socket_timeout:    Time to wait for unresponsive hosts, in seconds
     bidi_workaround:   Work around buggy terminals without bidirectional text
                        support, using fridibi
     proxy:             URL of the proxy server to use
     socket_timeout:    Time to wait for unresponsive hosts, in seconds
     bidi_workaround:   Work around buggy terminals without bidirectional text
                        support, using fridibi

+    debug_printtraffic:Print out sent and received HTTP traffic
+    include_ads:       Download ads as well
+    default_search:    Prepend this string if an input url is not valid.
+                       'auto' for elaborate guessing
+    encoding:          Use this encoding instead of the system-specified.
+    extract_flat:      Do not resolve URLs, return the immediate result.
+                       Pass in 'in_playlist' to only show this behavior for
+                       playlist items.
+    postprocessors:    A list of dictionaries, each with an entry
+                       * key:  The name of the postprocessor. See
+                               youtube_dl/postprocessor/__init__.py for a list.
+                       as well as any further keyword arguments for the
+                       postprocessor.
+    progress_hooks:    A list of functions that get called on download
+                       progress, with a dictionary with the entries
+                       * status: One of "downloading", "error", or "finished".
+                                 Check this first and ignore unknown values.
+
+                       If status is one of "downloading", or "finished", the
+                       following properties may also be present:
+                       * filename: The final filename (always present)
+                       * tmpfilename: The filename we're currently writing to
+                       * downloaded_bytes: Bytes on disk
+                       * total_bytes: Size of the whole file, None if unknown
+                       * total_bytes_estimate: Guess of the eventual file size,
+                                               None if unavailable.
+                       * elapsed: The number of seconds since download started.
+                       * eta: The estimated time in seconds, None if unknown
+                       * speed: The download speed in bytes/second, None if
+                                unknown
+                       * fragment_index: The counter of the currently
+                                         downloaded video fragment.
+                       * fragment_count: The number of fragments (= individual
+                                         files that will be merged)
+
+                       Progress hooks are guaranteed to be called at least once
+                       (with status "finished") if the download is successful.
+    merge_output_format: Extension to use when merging formats.
+    fixup:             Automatically correct known faults of the file.
+                       One of:
+                       - "never": do nothing
+                       - "warn": only emit a warning
+                       - "detect_or_warn": check whether we can do anything
+                                           about it, warn otherwise (default)
+    source_address:    (Experimental) Client-side IP address to bind to.
+    call_home:         Boolean, true iff we are allowed to contact the
+                       youtube-dl servers for debugging.
+    sleep_interval:    Number of seconds to sleep before each download.
+    listformats:       Print an overview of available video formats and exit.
+    list_thumbnails:   Print a table of all thumbnails and exit.
+    match_filter:      A function that gets called with the info_dict of
+                       every video.
+                       If it returns a message, the video is ignored.
+                       If it returns None, the video is downloaded.
+                       match_filter_func in utils.py is one example for this.
+    no_color:          Do not emit color codes in output.
+
+    The following options determine which downloader is picked:
+    external_downloader: Executable of the external downloader to call.
+                       None or unset for standard (built-in) downloader.
+    hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.

 
     The following parameters are not used by YoutubeDL itself, they are used by
     the FileDownloader:
     nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
 
     The following parameters are not used by YoutubeDL itself, they are used by
     the FileDownloader:
     nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,

-    noresizebuffer, retries, continuedl, noprogress, consoletitle
+    noresizebuffer, retries, continuedl, noprogress, consoletitle,
+    xattr_set_filesize.
+
+    The following options are used by the post processors:
+    prefer_ffmpeg:     If True, use ffmpeg instead of avconv if both are available,
+                       otherwise prefer avconv.
+    exec_cmd:          Arbitrary command to run after downloading

     """
 
     params = None
     """
 
     params = None


@@ -162,8 +265,10 @@ class YoutubeDL(object):
     _num_downloads = None
     _screen_file = None
 
     _num_downloads = None
     _screen_file = None
 

-    def __init__(self, params=None):
+    def __init__(self, params=None, auto_init=True):

         """Create a FileDownloader object with the given options."""
         """Create a FileDownloader object with the given options."""

+        if params is None:
+            params = {}

         self._ies = []
         self._ies_instances = {}
         self._pps = []
         self._ies = []
         self._ies_instances = {}
         self._pps = []


@@ -172,7 +277,8 @@ class YoutubeDL(object):
         self._num_downloads = 0
         self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
         self._err_file = sys.stderr
         self._num_downloads = 0
         self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
         self._err_file = sys.stderr

-        self.params = {} if params is None else params
+        self.params = params
+        self.cache = Cache(self)

 
         if params.get('bidi_workaround', False):
             try:
 
         if params.get('bidi_workaround', False):
             try:


@@ -183,35 +289,69 @@ class YoutubeDL(object):
                     width_args = []
                 else:
                     width_args = ['-w', str(width)]
                     width_args = []
                 else:
                     width_args = ['-w', str(width)]

-                self._fribidi = subprocess.Popen(
-                    ['fribidi', '-c', 'UTF-8'] + width_args,
+                sp_kwargs = dict(

                     stdin=subprocess.PIPE,
                     stdout=slave,
                     stderr=self._err_file)
                     stdin=subprocess.PIPE,
                     stdout=slave,
                     stderr=self._err_file)

-                self._fribidi_channel = os.fdopen(master, 'rb')
+                try:
+                    self._output_process = subprocess.Popen(
+                        ['bidiv'] + width_args, **sp_kwargs
+                    )
+                except OSError:
+                    self._output_process = subprocess.Popen(
+                        ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
+                self._output_channel = os.fdopen(master, 'rb')

             except OSError as ose:
                 if ose.errno == 2:
             except OSError as ose:
                 if ose.errno == 2:

-                    self.report_warning(u'Could not find fribidi executable, ignoring --bidi-workaround . Make sure that  fribidi  is an executable file in one of the directories in your $PATH.')
+                    self.report_warning('Could not find fribidi executable, ignoring --bidi-workaround . Make sure that  fribidi  is an executable file in one of the directories in your $PATH.')

                 else:
                     raise
 
         if (sys.version_info >= (3,) and sys.platform != 'win32' and
                 else:
                     raise
 
         if (sys.version_info >= (3,) and sys.platform != 'win32' and

-                sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
-                and not params['restrictfilenames']):
+                sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
+                not params.get('restrictfilenames', False)):

             # On Python 3, the Unicode filesystem API will throw errors (#1474)
             self.report_warning(
             # On Python 3, the Unicode filesystem API will throw errors (#1474)
             self.report_warning(

-                u'Assuming --restrict-filenames since file system encoding '
-                u'cannot encode all charactes. '
-                u'Set the LC_ALL environment variable to fix this.')
+                'Assuming --restrict-filenames since file system encoding '
+                'cannot encode all characters. '
+                'Set the LC_ALL environment variable to fix this.')

             self.params['restrictfilenames'] = True
 
             self.params['restrictfilenames'] = True
 

-        self.fd = FileDownloader(self, self.params)
-

         if '%(stitle)s' in self.params.get('outtmpl', ''):
         if '%(stitle)s' in self.params.get('outtmpl', ''):

-            self.report_warning(u'%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
+            self.report_warning('%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')

 
         self._setup_opener()
 
 
         self._setup_opener()
 

+        if auto_init:
+            self.print_debug_header()
+            self.add_default_info_extractors()
+
+        for pp_def_raw in self.params.get('postprocessors', []):
+            pp_class = get_postprocessor(pp_def_raw['key'])
+            pp_def = dict(pp_def_raw)
+            del pp_def['key']
+            pp = pp_class(self, **compat_kwargs(pp_def))
+            self.add_post_processor(pp)
+
+        for ph in self.params.get('progress_hooks', []):
+            self.add_progress_hook(ph)
+
+    def warn_if_short_id(self, argv):
+        # short YouTube ID starting with dash?
+        idxs = [
+            i for i, a in enumerate(argv)
+            if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
+        if idxs:
+            correct_argv = (
+                ['youtube-dl'] +
+                [a for i, a in enumerate(argv) if i not in idxs] +
+                ['--'] + [argv[i] for i in idxs]
+            )
+            self.report_warning(
+                'Long argument string detected. '
+                'Use -- to separate parameters and URLs, like this:\n%s\n' %
+                args_to_str(correct_argv))
+

     def add_info_extractor(self, ie):
         """Add an InfoExtractor object to the end of the list."""
         self._ies.append(ie)
     def add_info_extractor(self, ie):
         """Add an InfoExtractor object to the end of the list."""
         self._ies.append(ie)


@@ -242,42 +382,50 @@ class YoutubeDL(object):
         self._pps.append(pp)
         pp.set_downloader(self)
 
         self._pps.append(pp)
         pp.set_downloader(self)
 

+    def add_progress_hook(self, ph):
+        """Add the progress hook (currently only for the file downloader)"""
+        self._progress_hooks.append(ph)
+

     def _bidi_workaround(self, message):
     def _bidi_workaround(self, message):

-        if not hasattr(self, '_fribidi_channel'):
+        if not hasattr(self, '_output_channel'):

             return message
 
             return message
 

-        assert type(message) == type(u'')
-        line_count = message.count(u'\n') + 1
-        self._fribidi.stdin.write((message + u'\n').encode('utf-8'))
-        self._fribidi.stdin.flush()
-        res = u''.join(self._fribidi_channel.readline().decode('utf-8')
-                       for _ in range(line_count))
-        return res[:-len(u'\n')]
+        assert hasattr(self, '_output_process')
+        assert isinstance(message, compat_str)
+        line_count = message.count('\n') + 1
+        self._output_process.stdin.write((message + '\n').encode('utf-8'))
+        self._output_process.stdin.flush()
+        res = ''.join(self._output_channel.readline().decode('utf-8')
+                      for _ in range(line_count))
+        return res[:-len('\n')]

 
     def to_screen(self, message, skip_eol=False):
         """Print message to stdout if not in quiet mode."""
         return self.to_stdout(message, skip_eol, check_quiet=True)
 
 
     def to_screen(self, message, skip_eol=False):
         """Print message to stdout if not in quiet mode."""
         return self.to_stdout(message, skip_eol, check_quiet=True)
 

+    def _write_string(self, s, out=None):
+        write_string(s, out=out, encoding=self.params.get('encoding'))
+

     def to_stdout(self, message, skip_eol=False, check_quiet=False):
         """Print message to stdout if not in quiet mode."""
         if self.params.get('logger'):
             self.params['logger'].debug(message)
         elif not check_quiet or not self.params.get('quiet', False):
             message = self._bidi_workaround(message)
     def to_stdout(self, message, skip_eol=False, check_quiet=False):
         """Print message to stdout if not in quiet mode."""
         if self.params.get('logger'):
             self.params['logger'].debug(message)
         elif not check_quiet or not self.params.get('quiet', False):
             message = self._bidi_workaround(message)

-            terminator = [u'\n', u''][skip_eol]
+            terminator = ['\n', ''][skip_eol]

             output = message + terminator
 
             output = message + terminator
 

-            write_string(output, self._screen_file)
+            self._write_string(output, self._screen_file)

 
     def to_stderr(self, message):
         """Print message to stderr."""
 
     def to_stderr(self, message):
         """Print message to stderr."""

-        assert type(message) == type(u'')
+        assert isinstance(message, compat_str)

         if self.params.get('logger'):
             self.params['logger'].error(message)
         else:
             message = self._bidi_workaround(message)
         if self.params.get('logger'):
             self.params['logger'].error(message)
         else:
             message = self._bidi_workaround(message)

-            output = message + u'\n'
-            write_string(output, self._err_file)
+            output = message + '\n'
+            self._write_string(output, self._err_file)

 
     def to_console_title(self, message):
         if not self.params.get('consoletitle', False):
 
     def to_console_title(self, message):
         if not self.params.get('consoletitle', False):


@@ -287,21 +435,21 @@ class YoutubeDL(object):
             # already of type unicode()
             ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
         elif 'TERM' in os.environ:
             # already of type unicode()
             ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
         elif 'TERM' in os.environ:

-            write_string(u'\033]0;%s\007' % message, self._screen_file)
+            self._write_string('\033]0;%s\007' % message, self._screen_file)

 
     def save_console_title(self):
         if not self.params.get('consoletitle', False):
             return
         if 'TERM' in os.environ:
             # Save the title on stack
 
     def save_console_title(self):
         if not self.params.get('consoletitle', False):
             return
         if 'TERM' in os.environ:
             # Save the title on stack

-            write_string(u'\033[22;0t', self._screen_file)
+            self._write_string('\033[22;0t', self._screen_file)

 
     def restore_console_title(self):
         if not self.params.get('consoletitle', False):
             return
         if 'TERM' in os.environ:
             # Restore the title from stack
 
     def restore_console_title(self):
         if not self.params.get('consoletitle', False):
             return
         if 'TERM' in os.environ:
             # Restore the title from stack

-            write_string(u'\033[23;0t', self._screen_file)
+            self._write_string('\033[23;0t', self._screen_file)

 
     def __enter__(self):
         self.save_console_title()
 
     def __enter__(self):
         self.save_console_title()


@@ -309,7 +457,7 @@ class YoutubeDL(object):
 
     def __exit__(self, *args):
         self.restore_console_title()
 
     def __exit__(self, *args):
         self.restore_console_title()

-    
+

         if self.params.get('cookiefile') is not None:
             self.cookiejar.save()
 
         if self.params.get('cookiefile') is not None:
             self.cookiejar.save()
 


@@ -327,13 +475,13 @@ class YoutubeDL(object):
         if self.params.get('verbose'):
             if tb is None:
                 if sys.exc_info()[0]:  # if .trouble has been called from an except block
         if self.params.get('verbose'):
             if tb is None:
                 if sys.exc_info()[0]:  # if .trouble has been called from an except block

-                    tb = u''
+                    tb = ''

                     if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
                     if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:

-                        tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
+                        tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))

                     tb += compat_str(traceback.format_exc())
                 else:
                     tb_data = traceback.format_list(traceback.extract_stack())
                     tb += compat_str(traceback.format_exc())
                 else:
                     tb_data = traceback.format_list(traceback.extract_stack())

-                    tb = u''.join(tb_data)
+                    tb = ''.join(tb_data)

             self.to_stderr(tb)
         if not self.params.get('ignoreerrors', False):
             if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
             self.to_stderr(tb)
         if not self.params.get('ignoreerrors', False):
             if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:


@@ -348,35 +496,36 @@ class YoutubeDL(object):
         Print the message to stderr, it will be prefixed with 'WARNING:'
         If stderr is a tty file the 'WARNING:' will be colored
         '''
         Print the message to stderr, it will be prefixed with 'WARNING:'
         If stderr is a tty file the 'WARNING:' will be colored
         '''

-        if self._err_file.isatty() and os.name != 'nt':
-            _msg_header = u'\033[0;33mWARNING:\033[0m'
+        if self.params.get('logger') is not None:
+            self.params['logger'].warning(message)

         else:
         else:

-            _msg_header = u'WARNING:'
-        warning_message = u'%s %s' % (_msg_header, message)
-        self.to_stderr(warning_message)
+            if self.params.get('no_warnings'):
+                return
+            if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
+                _msg_header = '\033[0;33mWARNING:\033[0m'
+            else:
+                _msg_header = 'WARNING:'
+            warning_message = '%s %s' % (_msg_header, message)
+            self.to_stderr(warning_message)

 
     def report_error(self, message, tb=None):
         '''
         Do the same as trouble, but prefixes the message with 'ERROR:', colored
         in red if stderr is a tty file.
         '''
 
     def report_error(self, message, tb=None):
         '''
         Do the same as trouble, but prefixes the message with 'ERROR:', colored
         in red if stderr is a tty file.
         '''

-        if self._err_file.isatty() and os.name != 'nt':
-            _msg_header = u'\033[0;31mERROR:\033[0m'
+        if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
+            _msg_header = '\033[0;31mERROR:\033[0m'

         else:
         else:

-            _msg_header = u'ERROR:'
-        error_message = u'%s %s' % (_msg_header, message)
+            _msg_header = 'ERROR:'
+        error_message = '%s %s' % (_msg_header, message)

         self.trouble(error_message, tb)
 
     def report_file_already_downloaded(self, file_name):
         """Report file has already been fully downloaded."""
         try:
         self.trouble(error_message, tb)
 
     def report_file_already_downloaded(self, file_name):
         """Report file has already been fully downloaded."""
         try:

-            self.to_screen(u'[download] %s has already been downloaded' % file_name)
+            self.to_screen('[download] %s has already been downloaded' % file_name)

         except UnicodeEncodeError:
         except UnicodeEncodeError:

-            self.to_screen(u'[download] The file has already been downloaded')
-
-    def increment_downloads(self):
-        """Increment the ordinal that assigns a number to each file."""
-        self._num_downloads += 1
+            self.to_screen('[download] The file has already been downloaded')

 
     def prepare_filename(self, info_dict):
         """Generate the output filename."""
 
     def prepare_filename(self, info_dict):
         """Generate the output filename."""


@@ -387,61 +536,80 @@ class YoutubeDL(object):
             autonumber_size = self.params.get('autonumber_size')
             if autonumber_size is None:
                 autonumber_size = 5
             autonumber_size = self.params.get('autonumber_size')
             if autonumber_size is None:
                 autonumber_size = 5

-            autonumber_templ = u'%0' + str(autonumber_size) + u'd'
+            autonumber_templ = '%0' + str(autonumber_size) + 'd'

             template_dict['autonumber'] = autonumber_templ % self._num_downloads
             if template_dict.get('playlist_index') is not None:
             template_dict['autonumber'] = autonumber_templ % self._num_downloads
             if template_dict.get('playlist_index') is not None:

-                template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
+                template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
+            if template_dict.get('resolution') is None:
+                if template_dict.get('width') and template_dict.get('height'):
+                    template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
+                elif template_dict.get('height'):
+                    template_dict['resolution'] = '%sp' % template_dict['height']
+                elif template_dict.get('width'):
+                    template_dict['resolution'] = '?x%d' % template_dict['width']

 
             sanitize = lambda k, v: sanitize_filename(
                 compat_str(v),
                 restricted=self.params.get('restrictfilenames'),
 
             sanitize = lambda k, v: sanitize_filename(
                 compat_str(v),
                 restricted=self.params.get('restrictfilenames'),

-                is_id=(k == u'id'))
+                is_id=(k == 'id'))

             template_dict = dict((k, sanitize(k, v))
                                  for k, v in template_dict.items()
                                  if v is not None)
             template_dict = dict((k, sanitize(k, v))
                                  for k, v in template_dict.items()
                                  if v is not None)

-            template_dict = collections.defaultdict(lambda: u'NA', template_dict)
+            template_dict = collections.defaultdict(lambda: 'NA', template_dict)

 
 

-            tmpl = os.path.expanduser(self.params['outtmpl'])
+            outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
+            tmpl = compat_expanduser(outtmpl)

             filename = tmpl % template_dict
             filename = tmpl % template_dict

+            # Temporary fix for #4787
+            # 'Treat' all problem characters by passing filename through preferredencoding
+            # to workaround encoding issues with subprocess on python2 @ Windows
+            if sys.version_info < (3, 0) and sys.platform == 'win32':
+                filename = encodeFilename(filename, True).decode(preferredencoding())

             return filename
         except ValueError as err:
             return filename
         except ValueError as err:

-            self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')')
+            self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')

             return None
 
             return None
 

-    def _match_entry(self, info_dict):
+    def _match_entry(self, info_dict, incomplete):

         """ Returns None iff the file should be downloaded """
 
         """ Returns None iff the file should be downloaded """
 

-        video_title = info_dict.get('title', info_dict.get('id', u'video'))
+        video_title = info_dict.get('title', info_dict.get('id', 'video'))

         if 'title' in info_dict:
             # This can happen when we're just evaluating the playlist
             title = info_dict['title']
             matchtitle = self.params.get('matchtitle', False)
             if matchtitle:
                 if not re.search(matchtitle, title, re.IGNORECASE):
         if 'title' in info_dict:
             # This can happen when we're just evaluating the playlist
             title = info_dict['title']
             matchtitle = self.params.get('matchtitle', False)
             if matchtitle:
                 if not re.search(matchtitle, title, re.IGNORECASE):

-                    return u'"' + title + '" title did not match pattern "' + matchtitle + '"'
+                    return '"' + title + '" title did not match pattern "' + matchtitle + '"'

             rejecttitle = self.params.get('rejecttitle', False)
             if rejecttitle:
                 if re.search(rejecttitle, title, re.IGNORECASE):
             rejecttitle = self.params.get('rejecttitle', False)
             if rejecttitle:
                 if re.search(rejecttitle, title, re.IGNORECASE):

-                    return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
+                    return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'

         date = info_dict.get('upload_date', None)
         if date is not None:
             dateRange = self.params.get('daterange', DateRange())
             if date not in dateRange:
         date = info_dict.get('upload_date', None)
         if date is not None:
             dateRange = self.params.get('daterange', DateRange())
             if date not in dateRange:

-                return u'%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
+                return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)

         view_count = info_dict.get('view_count', None)
         if view_count is not None:
             min_views = self.params.get('min_views')
             if min_views is not None and view_count < min_views:
         view_count = info_dict.get('view_count', None)
         if view_count is not None:
             min_views = self.params.get('min_views')
             if min_views is not None and view_count < min_views:

-                return u'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
+                return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)

             max_views = self.params.get('max_views')
             if max_views is not None and view_count > max_views:
             max_views = self.params.get('max_views')
             if max_views is not None and view_count > max_views:

-                return u'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
-        age_limit = self.params.get('age_limit')
-        if age_limit is not None:
-            if age_limit < info_dict.get('age_limit', 0):
-                return u'Skipping "' + title + '" because it is age restricted'
+                return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
+        if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
+            return 'Skipping "%s" because it is age restricted' % video_title

         if self.in_download_archive(info_dict):
         if self.in_download_archive(info_dict):

-            return u'%s has already been recorded in archive' % video_title
+            return '%s has already been recorded in archive' % video_title
+
+        if not incomplete:
+            match_filter = self.params.get('match_filter')
+            if match_filter is not None:
+                ret = match_filter(info_dict)
+                if ret is not None:
+                    return ret
+

         return None
 
     @staticmethod
         return None
 
     @staticmethod


@@ -468,12 +636,12 @@ class YoutubeDL(object):
                 continue
 
             if not ie.working():
                 continue
 
             if not ie.working():

-                self.report_warning(u'The program functionality for this site has been marked as broken, '
-                                    u'and will probably not work.')
+                self.report_warning('The program functionality for this site has been marked as broken, '
+                                    'and will probably not work.')

 
             try:
                 ie_result = ie.extract(url)
 
             try:
                 ie_result = ie.extract(url)

-                if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
+                if ie_result is None:  # Finished already (backwards compatibility; listformats and friends should be moved here)

                     break
                 if isinstance(ie_result, list):
                     # Backwards compatibility: old IE result format
                     break
                 if isinstance(ie_result, list):
                     # Backwards compatibility: old IE result format


@@ -481,20 +649,16 @@ class YoutubeDL(object):
                         '_type': 'compat_list',
                         'entries': ie_result,
                     }
                         '_type': 'compat_list',
                         'entries': ie_result,
                     }

-                self.add_extra_info(ie_result,
-                    {
-                        'extractor': ie.IE_NAME,
-                        'webpage_url': url,
-                        'webpage_url_basename': url_basename(url),
-                        'extractor_key': ie.ie_key(),
-                    })
+                self.add_default_extra_info(ie_result, ie, url)

                 if process:
                     return self.process_ie_result(ie_result, download, extra_info)
                 else:
                     return ie_result
                 if process:
                     return self.process_ie_result(ie_result, download, extra_info)
                 else:
                     return ie_result

-            except ExtractorError as de: # An error we somewhat expected
+            except ExtractorError as de:  # An error we somewhat expected

                 self.report_error(compat_str(de), de.format_traceback())
                 break
                 self.report_error(compat_str(de), de.format_traceback())
                 break

+            except MaxDownloadsReached:
+                raise

             except Exception as e:
                 if self.params.get('ignoreerrors', False):
                     self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
             except Exception as e:
                 if self.params.get('ignoreerrors', False):
                     self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))


@@ -502,7 +666,15 @@ class YoutubeDL(object):
                 else:
                     raise
         else:
                 else:
                     raise
         else:

-            self.report_error(u'no suitable InfoExtractor: %s' % url)
+            self.report_error('no suitable InfoExtractor for URL %s' % url)
+
+    def add_default_extra_info(self, ie_result, ie, url):
+        self.add_extra_info(ie_result, {
+            'extractor': ie.IE_NAME,
+            'webpage_url': url,
+            'webpage_url_basename': url_basename(url),
+            'extractor_key': ie.ie_key(),
+        })

 
     def process_ie_result(self, ie_result, download=True, extra_info={}):
         """
 
     def process_ie_result(self, ie_result, download=True, extra_info={}):
         """


@@ -513,7 +685,16 @@ class YoutubeDL(object):
         Returns the resolved ie_result.
         """
 
         Returns the resolved ie_result.
         """
 

-        result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
+        result_type = ie_result.get('_type', 'video')
+
+        if result_type in ('url', 'url_transparent'):
+            extract_flat = self.params.get('extract_flat', False)
+            if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
+                    extract_flat is True):
+                if self.params.get('forcejson', False):
+                    self.to_stdout(json.dumps(ie_result))
+                return ie_result
+

         if result_type == 'video':
             self.add_extra_info(ie_result, extra_info)
             return self.process_video_result(ie_result, download=download)
         if result_type == 'video':
             self.add_extra_info(ie_result, extra_info)
             return self.process_video_result(ie_result, download=download)


@@ -530,51 +711,91 @@ class YoutubeDL(object):
                 ie_result['url'], ie_key=ie_result.get('ie_key'),
                 extra_info=extra_info, download=False, process=False)
 
                 ie_result['url'], ie_key=ie_result.get('ie_key'),
                 extra_info=extra_info, download=False, process=False)
 

-            def make_result(embedded_info):
-                new_result = ie_result.copy()
-                for f in ('_type', 'url', 'ext', 'player_url', 'formats',
-                          'entries', 'urlhandle', 'ie_key', 'duration',
-                          'subtitles', 'annotations', 'format',
-                          'thumbnail', 'thumbnails'):
-                    if f in new_result:
-                        del new_result[f]
-                    if f in embedded_info:
-                        new_result[f] = embedded_info[f]
-                return new_result
-            new_result = make_result(info)
+            force_properties = dict(
+                (k, v) for k, v in ie_result.items() if v is not None)
+            for f in ('_type', 'url'):
+                if f in force_properties:
+                    del force_properties[f]
+            new_result = info.copy()
+            new_result.update(force_properties)

 
             assert new_result.get('_type') != 'url_transparent'
 
             assert new_result.get('_type') != 'url_transparent'

-            if new_result.get('_type') == 'compat_list':
-                new_result['entries'] = [
-                    make_result(e) for e in new_result['entries']]

 
             return self.process_ie_result(
                 new_result, download=download, extra_info=extra_info)
 
             return self.process_ie_result(
                 new_result, download=download, extra_info=extra_info)

-        elif result_type == 'playlist':
+        elif result_type == 'playlist' or result_type == 'multi_video':

             # We process each entry in the playlist
             playlist = ie_result.get('title', None) or ie_result.get('id', None)
             # We process each entry in the playlist
             playlist = ie_result.get('title', None) or ie_result.get('id', None)

-            self.to_screen(u'[download] Downloading playlist: %s' % playlist)
+            self.to_screen('[download] Downloading playlist: %s' % playlist)

 
             playlist_results = []
 
 
             playlist_results = []
 

-            n_all_entries = len(ie_result['entries'])

             playliststart = self.params.get('playliststart', 1) - 1
             playlistend = self.params.get('playlistend', None)
             # For backwards compatibility, interpret -1 as whole list
             if playlistend == -1:
                 playlistend = None
 
             playliststart = self.params.get('playliststart', 1) - 1
             playlistend = self.params.get('playlistend', None)
             # For backwards compatibility, interpret -1 as whole list
             if playlistend == -1:
                 playlistend = None
 

-            entries = ie_result['entries'][playliststart:playlistend]
-            n_entries = len(entries)
+            playlistitems_str = self.params.get('playlist_items', None)
+            playlistitems = None
+            if playlistitems_str is not None:
+                def iter_playlistitems(format):
+                    for string_segment in format.split(','):
+                        if '-' in string_segment:
+                            start, end = string_segment.split('-')
+                            for item in range(int(start), int(end) + 1):
+                                yield int(item)
+                        else:
+                            yield int(string_segment)
+                playlistitems = iter_playlistitems(playlistitems_str)
+
+            ie_entries = ie_result['entries']
+            if isinstance(ie_entries, list):
+                n_all_entries = len(ie_entries)
+                if playlistitems:
+                    entries = [ie_entries[i - 1] for i in playlistitems]
+                else:
+                    entries = ie_entries[playliststart:playlistend]
+                n_entries = len(entries)
+                self.to_screen(
+                    "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
+                    (ie_result['extractor'], playlist, n_all_entries, n_entries))
+            elif isinstance(ie_entries, PagedList):
+                if playlistitems:
+                    entries = []
+                    for item in playlistitems:
+                        entries.extend(ie_entries.getslice(
+                            item - 1, item
+                        ))
+                else:
+                    entries = ie_entries.getslice(
+                        playliststart, playlistend)
+                n_entries = len(entries)
+                self.to_screen(
+                    "[%s] playlist %s: Downloading %d videos" %
+                    (ie_result['extractor'], playlist, n_entries))
+            else:  # iterable
+                if playlistitems:
+                    entry_list = list(ie_entries)
+                    entries = [entry_list[i - 1] for i in playlistitems]
+                else:
+                    entries = list(itertools.islice(
+                        ie_entries, playliststart, playlistend))
+                n_entries = len(entries)
+                self.to_screen(
+                    "[%s] playlist %s: Downloading %d videos" %
+                    (ie_result['extractor'], playlist, n_entries))

 
 

-            self.to_screen(
-                u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
-                (ie_result['extractor'], playlist, n_all_entries, n_entries))
+            if self.params.get('playlistreverse', False):
+                entries = entries[::-1]

 
             for i, entry in enumerate(entries, 1):
 
             for i, entry in enumerate(entries, 1):

-                self.to_screen(u'[download] Downloading video #%s of %s' % (i, n_entries))
+                self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))

                 extra = {
                 extra = {

+                    'n_entries': n_entries,

                     'playlist': playlist,
                     'playlist': playlist,

+                    'playlist_id': ie_result.get('id'),
+                    'playlist_title': ie_result.get('title'),

                     'playlist_index': i + playliststart,
                     'extractor': ie_result['extractor'],
                     'webpage_url': ie_result['webpage_url'],
                     'playlist_index': i + playliststart,
                     'extractor': ie_result['extractor'],
                     'webpage_url': ie_result['webpage_url'],


@@ -582,9 +803,9 @@ class YoutubeDL(object):
                     'extractor_key': ie_result['extractor_key'],
                 }
 
                     'extractor_key': ie_result['extractor_key'],
                 }
 

-                reason = self._match_entry(entry)
+                reason = self._match_entry(entry, incomplete=True)

                 if reason is not None:
                 if reason is not None:

-                    self.to_screen(u'[download] ' + reason)
+                    self.to_screen('[download] ' + reason)

                     continue
 
                 entry_result = self.process_ie_result(entry,
                     continue
 
                 entry_result = self.process_ie_result(entry,


@@ -594,14 +815,20 @@ class YoutubeDL(object):
             ie_result['entries'] = playlist_results
             return ie_result
         elif result_type == 'compat_list':
             ie_result['entries'] = playlist_results
             return ie_result
         elif result_type == 'compat_list':

+            self.report_warning(
+                'Extractor %s returned a compat_list result. '
+                'It needs to be updated.' % ie_result.get('extractor'))
+

             def _fixup(r):
             def _fixup(r):

-                self.add_extra_info(r,
+                self.add_extra_info(
+                    r,

                     {
                         'extractor': ie_result['extractor'],
                         'webpage_url': ie_result['webpage_url'],
                         'webpage_url_basename': url_basename(ie_result['webpage_url']),
                         'extractor_key': ie_result['extractor_key'],
                     {
                         'extractor': ie_result['extractor'],
                         'webpage_url': ie_result['webpage_url'],
                         'webpage_url_basename': url_basename(ie_result['webpage_url']),
                         'extractor_key': ie_result['extractor_key'],

-                    })
+                    }
+                )

                 return r
             ie_result['entries'] = [
                 self.process_ie_result(_fixup(r), download, extra_info)
                 return r
             ie_result['entries'] = [
                 self.process_ie_result(_fixup(r), download, extra_info)


@@ -611,13 +838,106 @@ class YoutubeDL(object):
         else:
             raise Exception('Invalid result type: %s' % result_type)
 
         else:
             raise Exception('Invalid result type: %s' % result_type)
 

+    def _apply_format_filter(self, format_spec, available_formats):
+        " Returns a tuple of the remaining format_spec and filtered formats "
+
+        OPERATORS = {
+            '<': operator.lt,
+            '<=': operator.le,
+            '>': operator.gt,
+            '>=': operator.ge,
+            '=': operator.eq,
+            '!=': operator.ne,
+        }
+        operator_rex = re.compile(r'''(?x)\s*\[
+            (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
+            \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
+            (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
+            \]$
+            ''' % '|'.join(map(re.escape, OPERATORS.keys())))
+        m = operator_rex.search(format_spec)
+        if m:
+            try:
+                comparison_value = int(m.group('value'))
+            except ValueError:
+                comparison_value = parse_filesize(m.group('value'))
+                if comparison_value is None:
+                    comparison_value = parse_filesize(m.group('value') + 'B')
+                if comparison_value is None:
+                    raise ValueError(
+                        'Invalid value %r in format specification %r' % (
+                            m.group('value'), format_spec))
+            op = OPERATORS[m.group('op')]
+
+        if not m:
+            STR_OPERATORS = {
+                '=': operator.eq,
+                '!=': operator.ne,
+            }
+            str_operator_rex = re.compile(r'''(?x)\s*\[
+                \s*(?P<key>ext|acodec|vcodec|container|protocol)
+                \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
+                \s*(?P<value>[a-zA-Z0-9_-]+)
+                \s*\]$
+                ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
+            m = str_operator_rex.search(format_spec)
+            if m:
+                comparison_value = m.group('value')
+                op = STR_OPERATORS[m.group('op')]
+
+        if not m:
+            raise ValueError('Invalid format specification %r' % format_spec)
+
+        def _filter(f):
+            actual_value = f.get(m.group('key'))
+            if actual_value is None:
+                return m.group('none_inclusive')
+            return op(actual_value, comparison_value)
+        new_formats = [f for f in available_formats if _filter(f)]
+
+        new_format_spec = format_spec[:-len(m.group(0))]
+        if not new_format_spec:
+            new_format_spec = 'best'
+
+        return (new_format_spec, new_formats)
+

     def select_format(self, format_spec, available_formats):
     def select_format(self, format_spec, available_formats):

+        while format_spec.endswith(']'):
+            format_spec, available_formats = self._apply_format_filter(
+                format_spec, available_formats)
+        if not available_formats:
+            return None
+

         if format_spec == 'best' or format_spec is None:
             return available_formats[-1]
         elif format_spec == 'worst':
             return available_formats[0]
         if format_spec == 'best' or format_spec is None:
             return available_formats[-1]
         elif format_spec == 'worst':
             return available_formats[0]

+        elif format_spec == 'bestaudio':
+            audio_formats = [
+                f for f in available_formats
+                if f.get('vcodec') == 'none']
+            if audio_formats:
+                return audio_formats[-1]
+        elif format_spec == 'worstaudio':
+            audio_formats = [
+                f for f in available_formats
+                if f.get('vcodec') == 'none']
+            if audio_formats:
+                return audio_formats[0]
+        elif format_spec == 'bestvideo':
+            video_formats = [
+                f for f in available_formats
+                if f.get('acodec') == 'none']
+            if video_formats:
+                return video_formats[-1]
+        elif format_spec == 'worstvideo':
+            video_formats = [
+                f for f in available_formats
+                if f.get('acodec') == 'none']
+            if video_formats:
+                return video_formats[0]

         else:
         else:

-            extensions = [u'mp4', u'flv', u'webm', u'3gp']
+            extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']

             if format_spec in extensions:
                 filter_f = lambda f: f['ext'] == format_spec
             else:
             if format_spec in extensions:
                 filter_f = lambda f: f['ext'] == format_spec
             else:


@@ -627,16 +947,78 @@ class YoutubeDL(object):
                 return matches[-1]
         return None
 
                 return matches[-1]
         return None
 

+    def _calc_headers(self, info_dict):
+        res = std_headers.copy()
+
+        add_headers = info_dict.get('http_headers')
+        if add_headers:
+            res.update(add_headers)
+
+        cookies = self._calc_cookies(info_dict)
+        if cookies:
+            res['Cookie'] = cookies
+
+        return res
+
+    def _calc_cookies(self, info_dict):
+        pr = compat_urllib_request.Request(info_dict['url'])
+        self.cookiejar.add_cookie_header(pr)
+        return pr.get_header('Cookie')
+

     def process_video_result(self, info_dict, download=True):
         assert info_dict.get('_type', 'video') == 'video'
 
     def process_video_result(self, info_dict, download=True):
         assert info_dict.get('_type', 'video') == 'video'
 

+        if 'id' not in info_dict:
+            raise ExtractorError('Missing "id" field in extractor result')
+        if 'title' not in info_dict:
+            raise ExtractorError('Missing "title" field in extractor result')
+

         if 'playlist' not in info_dict:
             # It isn't part of a playlist
             info_dict['playlist'] = None
             info_dict['playlist_index'] = None
 
         if 'playlist' not in info_dict:
             # It isn't part of a playlist
             info_dict['playlist'] = None
             info_dict['playlist_index'] = None
 

+        thumbnails = info_dict.get('thumbnails')
+        if thumbnails is None:
+            thumbnail = info_dict.get('thumbnail')
+            if thumbnail:
+                info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
+        if thumbnails:
+            thumbnails.sort(key=lambda t: (
+                t.get('preference'), t.get('width'), t.get('height'),
+                t.get('id'), t.get('url')))
+            for i, t in enumerate(thumbnails):
+                if 'width' in t and 'height' in t:
+                    t['resolution'] = '%dx%d' % (t['width'], t['height'])
+                if t.get('id') is None:
+                    t['id'] = '%d' % i
+
+        if thumbnails and 'thumbnail' not in info_dict:
+            info_dict['thumbnail'] = thumbnails[-1]['url']
+
+        if 'display_id' not in info_dict and 'id' in info_dict:
+            info_dict['display_id'] = info_dict['id']
+
+        if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
+            # Working around negative timestamps in Windows
+            # (see http://bugs.python.org/issue1646728)
+            if info_dict['timestamp'] < 0 and os.name == 'nt':
+                info_dict['timestamp'] = 0
+            upload_date = datetime.datetime.utcfromtimestamp(
+                info_dict['timestamp'])
+            info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
+
+        if self.params.get('listsubtitles', False):
+            if 'automatic_captions' in info_dict:
+                self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
+            self.list_subtitles(info_dict['id'], info_dict.get('subtitles'), 'subtitles')
+            return
+        info_dict['requested_subtitles'] = self.process_subtitles(
+            info_dict['id'], info_dict.get('subtitles'),
+            info_dict.get('automatic_captions'))
+

         # This extractors handle format selection themselves
         # This extractors handle format selection themselves

-        if info_dict['extractor'] in [u'youtube', u'Youku']:
+        if info_dict['extractor'] in ['Youku']:

             if download:
                 self.process_info(info_dict)
             return info_dict
             if download:
                 self.process_info(info_dict)
             return info_dict


@@ -648,40 +1030,53 @@ class YoutubeDL(object):
         else:
             formats = info_dict['formats']
 
         else:
             formats = info_dict['formats']
 

+        if not formats:
+            raise ExtractorError('No video formats found!')
+

         # We check that all the formats have the format and format_id fields
         # We check that all the formats have the format and format_id fields

-        for (i, format) in enumerate(formats):
+        for i, format in enumerate(formats):
+            if 'url' not in format:
+                raise ExtractorError('Missing "url" key in result (index %d)' % i)
+

             if format.get('format_id') is None:
                 format['format_id'] = compat_str(i)
             if format.get('format') is None:
             if format.get('format_id') is None:
                 format['format_id'] = compat_str(i)
             if format.get('format') is None:

-                format['format'] = u'{id} - {res}{note}'.format(
+                format['format'] = '{id} - {res}{note}'.format(

                     id=format['format_id'],
                     res=self.format_resolution(format),
                     id=format['format_id'],
                     res=self.format_resolution(format),

-                    note=u' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
+                    note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',

                 )
             # Automatically determine file extension if missing
             if 'ext' not in format:
                 )
             # Automatically determine file extension if missing
             if 'ext' not in format:

-                format['ext'] = determine_ext(format['url'])
-
-        if self.params.get('listformats', None):
-            self.list_formats(info_dict)
-            return
+                format['ext'] = determine_ext(format['url']).lower()
+            # Add HTTP headers, so that external programs can use them from the
+            # json output
+            full_format_info = info_dict.copy()
+            full_format_info.update(format)
+            format['http_headers'] = self._calc_headers(full_format_info)

 
         format_limit = self.params.get('format_limit', None)
         if format_limit:
             formats = list(takewhile_inclusive(
                 lambda f: f['format_id'] != format_limit, formats
             ))
 
         format_limit = self.params.get('format_limit', None)
         if format_limit:
             formats = list(takewhile_inclusive(
                 lambda f: f['format_id'] != format_limit, formats
             ))

-        if self.params.get('prefer_free_formats'):
-            def _free_formats_key(f):
-                try:
-                    ext_ord = [u'flv', u'mp4', u'webm'].index(f['ext'])
-                except ValueError:
-                    ext_ord = -1
-                # We only compare the extension if they have the same height and width
-                return (f.get('height'), f.get('width'), ext_ord)
-            formats = sorted(formats, key=_free_formats_key)
-
-        req_format = self.params.get('format', 'best')
+
+        # TODO Central sorting goes here
+
+        if formats[0] is not info_dict:
+            # only set the 'formats' fields if the original info_dict list them
+            # otherwise we end up with a circular reference, the first (and unique)
+            # element in the 'formats' field in info_dict is info_dict itself,
+            # wich can't be exported to json
+            info_dict['formats'] = formats
+        if self.params.get('listformats'):
+            self.list_formats(info_dict)
+            return
+        if self.params.get('list_thumbnails'):
+            self.list_thumbnails(info_dict)
+            return
+
+        req_format = self.params.get('format')

         if req_format is None:
             req_format = 'best'
         formats_to_download = []
         if req_format is None:
             req_format = 'best'
         formats_to_download = []


@@ -689,21 +1084,59 @@ class YoutubeDL(object):
         if req_format in ('-1', 'all'):
             formats_to_download = formats
         else:
         if req_format in ('-1', 'all'):
             formats_to_download = formats
         else:

-            # We can accept formats requestd in the format: 34/5/best, we pick
-            # the first that is available, starting from left
-            req_formats = req_format.split('/')
-            for rf in req_formats:
-                selected_format = self.select_format(rf, formats)
-                if selected_format is not None:
-                    formats_to_download = [selected_format]
-                    break
+            for rfstr in req_format.split(','):
+                # We can accept formats requested in the format: 34/5/best, we pick
+                # the first that is available, starting from left
+                req_formats = rfstr.split('/')
+                for rf in req_formats:
+                    if re.match(r'.+?\+.+?', rf) is not None:
+                        # Two formats have been requested like '137+139'
+                        format_1, format_2 = rf.split('+')
+                        formats_info = (self.select_format(format_1, formats),
+                                        self.select_format(format_2, formats))
+                        if all(formats_info):
+                            # The first format must contain the video and the
+                            # second the audio
+                            if formats_info[0].get('vcodec') == 'none':
+                                self.report_error('The first format must '
+                                                  'contain the video, try using '
+                                                  '"-f %s+%s"' % (format_2, format_1))
+                                return
+                            output_ext = (
+                                formats_info[0]['ext']
+                                if self.params.get('merge_output_format') is None
+                                else self.params['merge_output_format'])
+                            selected_format = {
+                                'requested_formats': formats_info,
+                                'format': '%s+%s' % (formats_info[0].get('format'),
+                                                     formats_info[1].get('format')),
+                                'format_id': '%s+%s' % (formats_info[0].get('format_id'),
+                                                        formats_info[1].get('format_id')),
+                                'width': formats_info[0].get('width'),
+                                'height': formats_info[0].get('height'),
+                                'resolution': formats_info[0].get('resolution'),
+                                'fps': formats_info[0].get('fps'),
+                                'vcodec': formats_info[0].get('vcodec'),
+                                'vbr': formats_info[0].get('vbr'),
+                                'stretched_ratio': formats_info[0].get('stretched_ratio'),
+                                'acodec': formats_info[1].get('acodec'),
+                                'abr': formats_info[1].get('abr'),
+                                'ext': output_ext,
+                            }
+                        else:
+                            selected_format = None
+                    else:
+                        selected_format = self.select_format(rf, formats)
+                    if selected_format is not None:
+                        formats_to_download.append(selected_format)
+                        break

         if not formats_to_download:
         if not formats_to_download:

-            raise ExtractorError(u'requested format not available',
+            raise ExtractorError('requested format not available',

                                  expected=True)
 
         if download:
             if len(formats_to_download) > 1:
                                  expected=True)
 
         if download:
             if len(formats_to_download) > 1:

-                self.to_screen(u'[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
+                self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))

             for format in formats_to_download:
                 new_info = dict(info_dict)
                 new_info.update(format)
             for format in formats_to_download:
                 new_info = dict(info_dict)
                 new_info.update(format)


@@ -712,34 +1145,83 @@ class YoutubeDL(object):
         info_dict.update(formats_to_download[-1])
         return info_dict
 
         info_dict.update(formats_to_download[-1])
         return info_dict
 

+    def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
+        """Select the requested subtitles and their format"""
+        available_subs = {}
+        if normal_subtitles and self.params.get('writesubtitles'):
+            available_subs.update(normal_subtitles)
+        if automatic_captions and self.params.get('writeautomaticsub'):
+            for lang, cap_info in automatic_captions.items():
+                if lang not in available_subs:
+                    available_subs[lang] = cap_info
+
+        if (not self.params.get('writesubtitles') and not
+                self.params.get('writeautomaticsub') or not
+                available_subs):
+            return None
+
+        if self.params.get('allsubtitles', False):
+            requested_langs = available_subs.keys()
+        else:
+            if self.params.get('subtitleslangs', False):
+                requested_langs = self.params.get('subtitleslangs')
+            elif 'en' in available_subs:
+                requested_langs = ['en']
+            else:
+                requested_langs = [list(available_subs.keys())[0]]
+
+        formats_query = self.params.get('subtitlesformat', 'best')
+        formats_preference = formats_query.split('/') if formats_query else []
+        subs = {}
+        for lang in requested_langs:
+            formats = available_subs.get(lang)
+            if formats is None:
+                self.report_warning('%s subtitles not available for %s' % (lang, video_id))
+                continue
+            for ext in formats_preference:
+                if ext == 'best':
+                    f = formats[-1]
+                    break
+                matches = list(filter(lambda f: f['ext'] == ext, formats))
+                if matches:
+                    f = matches[-1]
+                    break
+            else:
+                f = formats[-1]
+                self.report_warning(
+                    'No subtitle format found matching "%s" for language %s, '
+                    'using %s' % (formats_query, lang, f['ext']))
+            subs[lang] = f
+        return subs
+

     def process_info(self, info_dict):
         """Process a single resolved IE result."""
 
         assert info_dict.get('_type', 'video') == 'video'
     def process_info(self, info_dict):
         """Process a single resolved IE result."""
 
         assert info_dict.get('_type', 'video') == 'video'

-        #We increment the download the download count here to match the previous behaviour.
-        self.increment_downloads()
+
+        max_downloads = self.params.get('max_downloads')
+        if max_downloads is not None:
+            if self._num_downloads >= int(max_downloads):
+                raise MaxDownloadsReached()

 
         info_dict['fulltitle'] = info_dict['title']
         if len(info_dict['title']) > 200:
 
         info_dict['fulltitle'] = info_dict['title']
         if len(info_dict['title']) > 200:

-            info_dict['title'] = info_dict['title'][:197] + u'...'
+            info_dict['title'] = info_dict['title'][:197] + '...'

 
         # Keep for backwards compatibility
         info_dict['stitle'] = info_dict['title']
 
 
         # Keep for backwards compatibility
         info_dict['stitle'] = info_dict['title']
 

-        if not 'format' in info_dict:
+        if 'format' not in info_dict:

             info_dict['format'] = info_dict['ext']
 
             info_dict['format'] = info_dict['ext']
 

-        reason = self._match_entry(info_dict)
+        reason = self._match_entry(info_dict, incomplete=False)

         if reason is not None:
         if reason is not None:

-            self.to_screen(u'[download] ' + reason)
+            self.to_screen('[download] ' + reason)

             return
 
             return
 

-        max_downloads = self.params.get('max_downloads')
-        if max_downloads is not None:
-            if self._num_downloads > int(max_downloads):
-                raise MaxDownloadsReached()
+        self._num_downloads += 1

 
 

-        filename = self.prepare_filename(info_dict)
+        info_dict['_filename'] = filename = self.prepare_filename(info_dict)

 
         # Forced printings
         if self.params.get('forcetitle', False):
 
         # Forced printings
         if self.params.get('forcetitle', False):


@@ -747,8 +1229,12 @@ class YoutubeDL(object):
         if self.params.get('forceid', False):
             self.to_stdout(info_dict['id'])
         if self.params.get('forceurl', False):
         if self.params.get('forceid', False):
             self.to_stdout(info_dict['id'])
         if self.params.get('forceurl', False):

-            # For RTMP URLs, also include the playpath
-            self.to_stdout(info_dict['url'] + info_dict.get('play_path', u''))
+            if info_dict.get('requested_formats') is not None:
+                for f in info_dict['requested_formats']:
+                    self.to_stdout(f['url'] + f.get('play_path', ''))
+            else:
+                # For RTMP URLs, also include the playpath
+                self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))

         if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
             self.to_stdout(info_dict['thumbnail'])
         if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
         if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
             self.to_stdout(info_dict['thumbnail'])
         if self.params.get('forcedescription', False) and info_dict.get('description') is not None:


@@ -760,7 +1246,6 @@ class YoutubeDL(object):
         if self.params.get('forceformat', False):
             self.to_stdout(info_dict['format'])
         if self.params.get('forcejson', False):
         if self.params.get('forceformat', False):
             self.to_stdout(info_dict['format'])
         if self.params.get('forcejson', False):

-            info_dict['_filename'] = filename

             self.to_stdout(json.dumps(info_dict))
 
         # Do nothing else if in simulate mode
             self.to_stdout(json.dumps(info_dict))
 
         # Do nothing else if in simulate mode


@@ -772,139 +1257,197 @@ class YoutubeDL(object):
 
         try:
             dn = os.path.dirname(encodeFilename(filename))
 
         try:
             dn = os.path.dirname(encodeFilename(filename))

-            if dn != '' and not os.path.exists(dn):
+            if dn and not os.path.exists(dn):

                 os.makedirs(dn)
         except (OSError, IOError) as err:
                 os.makedirs(dn)
         except (OSError, IOError) as err:

-            self.report_error(u'unable to create directory ' + compat_str(err))
+            self.report_error('unable to create directory ' + compat_str(err))

             return
 
         if self.params.get('writedescription', False):
             return
 
         if self.params.get('writedescription', False):

-            descfn = filename + u'.description'
+            descfn = filename + '.description'

             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):

-                self.to_screen(u'[info] Video description is already present')
+                self.to_screen('[info] Video description is already present')
+            elif info_dict.get('description') is None:
+                self.report_warning('There\'s no description to write.')

             else:
                 try:
             else:
                 try:

-                    self.to_screen(u'[info] Writing video description to: ' + descfn)
+                    self.to_screen('[info] Writing video description to: ' + descfn)

                     with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
                         descfile.write(info_dict['description'])
                     with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
                         descfile.write(info_dict['description'])

-                except (KeyError, TypeError):
-                    self.report_warning(u'There\'s no description to write.')

                 except (OSError, IOError):
                 except (OSError, IOError):

-                    self.report_error(u'Cannot write description file ' + descfn)
+                    self.report_error('Cannot write description file ' + descfn)

                     return
 
         if self.params.get('writeannotations', False):
                     return
 
         if self.params.get('writeannotations', False):

-            annofn = filename + u'.annotations.xml'
+            annofn = filename + '.annotations.xml'

             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):

-                self.to_screen(u'[info] Video annotations are already present')
+                self.to_screen('[info] Video annotations are already present')

             else:
                 try:
             else:
                 try:

-                    self.to_screen(u'[info] Writing video annotations to: ' + annofn)
+                    self.to_screen('[info] Writing video annotations to: ' + annofn)

                     with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
                         annofile.write(info_dict['annotations'])
                 except (KeyError, TypeError):
                     with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
                         annofile.write(info_dict['annotations'])
                 except (KeyError, TypeError):

-                    self.report_warning(u'There are no annotations to write.')
+                    self.report_warning('There are no annotations to write.')

                 except (OSError, IOError):
                 except (OSError, IOError):

-                    self.report_error(u'Cannot write annotations file: ' + annofn)
+                    self.report_error('Cannot write annotations file: ' + annofn)

                     return
 
         subtitles_are_requested = any([self.params.get('writesubtitles', False),
                                        self.params.get('writeautomaticsub')])
 
                     return
 
         subtitles_are_requested = any([self.params.get('writesubtitles', False),
                                        self.params.get('writeautomaticsub')])
 

-        if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
+        if subtitles_are_requested and info_dict.get('requested_subtitles'):

             # subtitles download errors are already managed as troubles in relevant IE
             # that way it will silently go on when used with unsupporting IE
             # subtitles download errors are already managed as troubles in relevant IE
             # that way it will silently go on when used with unsupporting IE

-            subtitles = info_dict['subtitles']
-            sub_format = self.params.get('subtitlesformat', 'srt')
-            for sub_lang in subtitles.keys():
-                sub = subtitles[sub_lang]
-                if sub is None:
-                    continue
+            subtitles = info_dict['requested_subtitles']
+            ie = self.get_info_extractor(info_dict['extractor_key'])
+            for sub_lang, sub_info in subtitles.items():
+                sub_format = sub_info['ext']
+                if sub_info.get('data') is not None:
+                    sub_data = sub_info['data']
+                else:
+                    try:
+                        sub_data = ie._download_webpage(
+                            sub_info['url'], info_dict['id'], note=False)
+                    except ExtractorError as err:
+                        self.report_warning('Unable to download subtitle for "%s": %s' %
+                                            (sub_lang, compat_str(err.cause)))
+                        continue

                 try:
                     sub_filename = subtitles_filename(filename, sub_lang, sub_format)
                     if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
                 try:
                     sub_filename = subtitles_filename(filename, sub_lang, sub_format)
                     if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):

-                        self.to_screen(u'[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
+                        self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))

                     else:
                     else:

-                        self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
+                        self.to_screen('[info] Writing video subtitles to: ' + sub_filename)

                         with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
                         with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:

-                                subfile.write(sub)
+                            subfile.write(sub_data)

                 except (OSError, IOError):
                 except (OSError, IOError):

-                    self.report_error(u'Cannot write subtitles file ' + descfn)
+                    self.report_error('Cannot write subtitles file ' + sub_filename)

                     return
 
         if self.params.get('writeinfojson', False):
                     return
 
         if self.params.get('writeinfojson', False):

-            infofn = os.path.splitext(filename)[0] + u'.info.json'
+            infofn = os.path.splitext(filename)[0] + '.info.json'

             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):

-                self.to_screen(u'[info] Video description metadata is already present')
+                self.to_screen('[info] Video description metadata is already present')

             else:
             else:

-                self.to_screen(u'[info] Writing video description metadata as JSON to: ' + infofn)
+                self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)

                 try:
                 try:

-                    json_info_dict = dict((k, v) for k, v in info_dict.items() if not k in ['urlhandle'])
-                    write_json_file(json_info_dict, encodeFilename(infofn))
+                    write_json_file(info_dict, infofn)

                 except (OSError, IOError):
                 except (OSError, IOError):

-                    self.report_error(u'Cannot write metadata to JSON file ' + infofn)
+                    self.report_error('Cannot write metadata to JSON file ' + infofn)

                     return
 
                     return
 

-        if self.params.get('writethumbnail', False):
-            if info_dict.get('thumbnail') is not None:
-                thumb_format = determine_ext(info_dict['thumbnail'], u'jpg')
-                thumb_filename = os.path.splitext(filename)[0] + u'.' + thumb_format
-                if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
-                    self.to_screen(u'[%s] %s: Thumbnail is already present' %
-                                   (info_dict['extractor'], info_dict['id']))
-                else:
-                    self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
-                                   (info_dict['extractor'], info_dict['id']))
-                    try:
-                        uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
-                        with open(thumb_filename, 'wb') as thumbf:
-                            shutil.copyfileobj(uf, thumbf)
-                        self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
-                            (info_dict['extractor'], info_dict['id'], thumb_filename))
-                    except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
-                        self.report_warning(u'Unable to download thumbnail "%s": %s' %
-                            (info_dict['thumbnail'], compat_str(err)))
+        self._write_thumbnails(info_dict, filename)

 
         if not self.params.get('skip_download', False):
 
         if not self.params.get('skip_download', False):

-            if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
-                success = True
-            else:
-                try:
-                    success = self.fd._do_download(filename, info_dict)
-                except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
-                    self.report_error(u'unable to download video data: %s' % str(err))
-                    return
-                except (OSError, IOError) as err:
-                    raise UnavailableVideoError(err)
-                except (ContentTooShortError, ) as err:
-                    self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
-                    return
+            try:
+                def dl(name, info):
+                    fd = get_suitable_downloader(info, self.params)(self, self.params)
+                    for ph in self._progress_hooks:
+                        fd.add_progress_hook(ph)
+                    if self.params.get('verbose'):
+                        self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
+                    return fd.download(name, info)
+
+                if info_dict.get('requested_formats') is not None:
+                    downloaded = []
+                    success = True
+                    merger = FFmpegMergerPP(self, not self.params.get('keepvideo'))
+                    if not merger.available:
+                        postprocessors = []
+                        self.report_warning('You have requested multiple '
+                                            'formats but ffmpeg or avconv are not installed.'
+                                            ' The formats won\'t be merged')
+                    else:
+                        postprocessors = [merger]
+                    for f in info_dict['requested_formats']:
+                        new_info = dict(info_dict)
+                        new_info.update(f)
+                        fname = self.prepare_filename(new_info)
+                        fname = prepend_extension(fname, 'f%s' % f['format_id'])
+                        downloaded.append(fname)
+                        partial_success = dl(fname, new_info)
+                        success = success and partial_success
+                    info_dict['__postprocessors'] = postprocessors
+                    info_dict['__files_to_merge'] = downloaded
+                else:
+                    # Just a single file
+                    success = dl(filename, info_dict)
+            except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
+                self.report_error('unable to download video data: %s' % str(err))
+                return
+            except (OSError, IOError) as err:
+                raise UnavailableVideoError(err)
+            except (ContentTooShortError, ) as err:
+                self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
+                return

 
             if success:
 
             if success:

+                # Fixup content
+                fixup_policy = self.params.get('fixup')
+                if fixup_policy is None:
+                    fixup_policy = 'detect_or_warn'
+
+                stretched_ratio = info_dict.get('stretched_ratio')
+                if stretched_ratio is not None and stretched_ratio != 1:
+                    if fixup_policy == 'warn':
+                        self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
+                            info_dict['id'], stretched_ratio))
+                    elif fixup_policy == 'detect_or_warn':
+                        stretched_pp = FFmpegFixupStretchedPP(self)
+                        if stretched_pp.available:
+                            info_dict.setdefault('__postprocessors', [])
+                            info_dict['__postprocessors'].append(stretched_pp)
+                        else:
+                            self.report_warning(
+                                '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
+                                    info_dict['id'], stretched_ratio))
+                    else:
+                        assert fixup_policy in ('ignore', 'never')
+
+                if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
+                    if fixup_policy == 'warn':
+                        self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
+                            info_dict['id']))
+                    elif fixup_policy == 'detect_or_warn':
+                        fixup_pp = FFmpegFixupM4aPP(self)
+                        if fixup_pp.available:
+                            info_dict.setdefault('__postprocessors', [])
+                            info_dict['__postprocessors'].append(fixup_pp)
+                        else:
+                            self.report_warning(
+                                '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
+                                    info_dict['id']))
+                    else:
+                        assert fixup_policy in ('ignore', 'never')
+

                 try:
                     self.post_process(filename, info_dict)
                 except (PostProcessingError) as err:
                 try:
                     self.post_process(filename, info_dict)
                 except (PostProcessingError) as err:

-                    self.report_error(u'postprocessing: %s' % str(err))
+                    self.report_error('postprocessing: %s' % str(err))

                     return
                     return

-
-        self.record_download_archive(info_dict)
+                self.record_download_archive(info_dict)

 
     def download(self, url_list):
         """Download a given list of URLs."""
 
     def download(self, url_list):
         """Download a given list of URLs."""

+        outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)

         if (len(url_list) > 1 and
         if (len(url_list) > 1 and

-                '%' not in self.params['outtmpl']
-                and self.params.get('max_downloads') != 1):
-            raise SameFileError(self.params['outtmpl'])
+                '%' not in outtmpl and
+                self.params.get('max_downloads') != 1):
+            raise SameFileError(outtmpl)

 
         for url in url_list:
             try:
 
         for url in url_list:
             try:

-                #It also downloads the videos
-                self.extract_info(url)
+                # It also downloads the videos
+                res = self.extract_info(url)

             except UnavailableVideoError:
             except UnavailableVideoError:

-                self.report_error(u'unable to download video')
+                self.report_error('unable to download video')

             except MaxDownloadsReached:
             except MaxDownloadsReached:

-                self.to_screen(u'[info] Maximum number of downloaded files reached.')
+                self.to_screen('[info] Maximum number of downloaded files reached.')

                 raise
                 raise

+            else:
+                if self.params.get('dump_single_json', False):
+                    self.to_stdout(json.dumps(res))

 
         return self._download_retcode
 
 
         return self._download_retcode
 


@@ -916,7 +1459,7 @@ class YoutubeDL(object):
         except DownloadError:
             webpage_url = info.get('webpage_url')
             if webpage_url is not None:
         except DownloadError:
             webpage_url = info.get('webpage_url')
             if webpage_url is not None:

-                self.report_warning(u'The info failed to download, trying with "%s"' % webpage_url)
+                self.report_warning('The info failed to download, trying with "%s"' % webpage_url)

                 return self.download([webpage_url])
             else:
                 raise
                 return self.download([webpage_url])
             else:
                 raise


@@ -926,10 +1469,15 @@ class YoutubeDL(object):
         """Run all the postprocessors on the given file."""
         info = dict(ie_info)
         info['filepath'] = filename
         """Run all the postprocessors on the given file."""
         info = dict(ie_info)
         info['filepath'] = filename

-        keep_video = None
-        for pp in self._pps:
+        pps_chain = []
+        if ie_info.get('__postprocessors') is not None:
+            pps_chain.extend(ie_info['__postprocessors'])
+        pps_chain.extend(self._pps)
+        for pp in pps_chain:
+            keep_video = None
+            old_filename = info['filepath']

             try:
             try:

-                keep_video_wish, new_info = pp.run(info)
+                keep_video_wish, info = pp.run(info)

                 if keep_video_wish is not None:
                     if keep_video_wish:
                         keep_video = keep_video_wish
                 if keep_video_wish is not None:
                     if keep_video_wish:
                         keep_video = keep_video_wish


@@ -938,12 +1486,12 @@ class YoutubeDL(object):
                         keep_video = keep_video_wish
             except PostProcessingError as e:
                 self.report_error(e.msg)
                         keep_video = keep_video_wish
             except PostProcessingError as e:
                 self.report_error(e.msg)

-        if keep_video is False and not self.params.get('keepvideo', False):
-            try:
-                self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
-                os.remove(encodeFilename(filename))
-            except (IOError, OSError):
-                self.report_warning(u'Unable to remove downloaded video file')
+            if keep_video is False and not self.params.get('keepvideo', False):
+                try:
+                    self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
+                    os.remove(encodeFilename(old_filename))
+                except (IOError, OSError):
+                    self.report_warning('Unable to remove downloaded video file')

 
     def _make_archive_id(self, info_dict):
         # Future-proof against any change in case
 
     def _make_archive_id(self, info_dict):
         # Future-proof against any change in case


@@ -954,7 +1502,7 @@ class YoutubeDL(object):
                 extractor = info_dict.get('ie_key')  # key in a playlist
         if extractor is None:
             return None  # Incomplete video information
                 extractor = info_dict.get('ie_key')  # key in a playlist
         if extractor is None:
             return None  # Incomplete video information

-        return extractor.lower() + u' ' + info_dict['id']
+        return extractor.lower() + ' ' + info_dict['id']

 
     def in_download_archive(self, info_dict):
         fn = self.params.get('download_archive')
 
     def in_download_archive(self, info_dict):
         fn = self.params.get('download_archive')


@@ -982,81 +1530,161 @@ class YoutubeDL(object):
         vid_id = self._make_archive_id(info_dict)
         assert vid_id
         with locked_file(fn, 'a', encoding='utf-8') as archive_file:
         vid_id = self._make_archive_id(info_dict)
         assert vid_id
         with locked_file(fn, 'a', encoding='utf-8') as archive_file:

-            archive_file.write(vid_id + u'\n')
+            archive_file.write(vid_id + '\n')

 
     @staticmethod
     def format_resolution(format, default='unknown'):
         if format.get('vcodec') == 'none':
             return 'audio only'
 
     @staticmethod
     def format_resolution(format, default='unknown'):
         if format.get('vcodec') == 'none':
             return 'audio only'

-        if format.get('_resolution') is not None:
-            return format['_resolution']
+        if format.get('resolution') is not None:
+            return format['resolution']

         if format.get('height') is not None:
             if format.get('width') is not None:
         if format.get('height') is not None:
             if format.get('width') is not None:

-                res = u'%sx%s' % (format['width'], format['height'])
+                res = '%sx%s' % (format['width'], format['height'])

             else:
             else:

-                res = u'%sp' % format['height']
+                res = '%sp' % format['height']
+        elif format.get('width') is not None:
+            res = '?x%d' % format['width']

         else:
             res = default
         return res
 
         else:
             res = default
         return res
 

-    def list_formats(self, info_dict):
-        def format_note(fdict):
-            res = u''
-            if fdict.get('format_note') is not None:
-                res += fdict['format_note'] + u' '
-            if (fdict.get('vcodec') is not None and
-                    fdict.get('vcodec') != 'none'):
-                res += u'%-5s' % fdict['vcodec']
-            elif fdict.get('vbr') is not None:
-                res += u'video'
+    def _format_note(self, fdict):
+        res = ''
+        if fdict.get('ext') in ['f4f', 'f4m']:
+            res += '(unsupported) '
+        if fdict.get('format_note') is not None:
+            res += fdict['format_note'] + ' '
+        if fdict.get('tbr') is not None:
+            res += '%4dk ' % fdict['tbr']
+        if fdict.get('container') is not None:
+            if res:
+                res += ', '
+            res += '%s container' % fdict['container']
+        if (fdict.get('vcodec') is not None and
+                fdict.get('vcodec') != 'none'):
+            if res:
+                res += ', '
+            res += fdict['vcodec']

             if fdict.get('vbr') is not None:
             if fdict.get('vbr') is not None:

-                res += u'@%4dk' % fdict['vbr']
-            if fdict.get('acodec') is not None:
-                if res:
-                    res += u', '
-                res += u'%-5s' % fdict['acodec']
-            elif fdict.get('abr') is not None:
-                if res:
-                    res += u', '
-                res += 'audio'
-            if fdict.get('abr') is not None:
-                res += u'@%3dk' % fdict['abr']
-            if fdict.get('filesize') is not None:
-                if res:
-                    res += u', '
-                res += format_bytes(fdict['filesize'])
-            return res
-
-        def line(format, idlen=20):
-            return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % (
-                format['format_id'],
-                format['ext'],
-                self.format_resolution(format),
-                format_note(format),
-            ))
+                res += '@'
+        elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
+            res += 'video@'
+        if fdict.get('vbr') is not None:
+            res += '%4dk' % fdict['vbr']
+        if fdict.get('fps') is not None:
+            res += ', %sfps' % fdict['fps']
+        if fdict.get('acodec') is not None:
+            if res:
+                res += ', '
+            if fdict['acodec'] == 'none':
+                res += 'video only'
+            else:
+                res += '%-5s' % fdict['acodec']
+        elif fdict.get('abr') is not None:
+            if res:
+                res += ', '
+            res += 'audio'
+        if fdict.get('abr') is not None:
+            res += '@%3dk' % fdict['abr']
+        if fdict.get('asr') is not None:
+            res += ' (%5dHz)' % fdict['asr']
+        if fdict.get('filesize') is not None:
+            if res:
+                res += ', '
+            res += format_bytes(fdict['filesize'])
+        elif fdict.get('filesize_approx') is not None:
+            if res:
+                res += ', '
+            res += '~' + format_bytes(fdict['filesize_approx'])
+        return res

 
 

+    def list_formats(self, info_dict):

         formats = info_dict.get('formats', [info_dict])
         formats = info_dict.get('formats', [info_dict])

-        idlen = max(len(u'format code'),
-                    max(len(f['format_id']) for f in formats))
-        formats_s = [line(f, idlen) for f in formats]
+        table = [
+            [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
+            for f in formats
+            if f.get('preference') is None or f['preference'] >= -1000]

         if len(formats) > 1:
         if len(formats) > 1:

-            formats_s[0] += (' ' if format_note(formats[0]) else '') + '(worst)'
-            formats_s[-1] += (' ' if format_note(formats[-1]) else '') + '(best)'
-
-        header_line = line({
-            'format_id': u'format code', 'ext': u'extension',
-            '_resolution': u'resolution', 'format_note': u'note'}, idlen=idlen)
-        self.to_screen(u'[info] Available formats for %s:\n%s\n%s' %
-                       (info_dict['id'], header_line, u"\n".join(formats_s)))
+            table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
+
+        header_line = ['format code', 'extension', 'resolution', 'note']
+        self.to_screen(
+            '[info] Available formats for %s:\n%s' %
+            (info_dict['id'], render_table(header_line, table)))
+
+    def list_thumbnails(self, info_dict):
+        thumbnails = info_dict.get('thumbnails')
+        if not thumbnails:
+            tn_url = info_dict.get('thumbnail')
+            if tn_url:
+                thumbnails = [{'id': '0', 'url': tn_url}]
+            else:
+                self.to_screen(
+                    '[info] No thumbnails present for %s' % info_dict['id'])
+                return
+
+        self.to_screen(
+            '[info] Thumbnails for %s:' % info_dict['id'])
+        self.to_screen(render_table(
+            ['ID', 'width', 'height', 'URL'],
+            [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
+
+    def list_subtitles(self, video_id, subtitles, name='subtitles'):
+        if not subtitles:
+            self.to_screen('%s has no %s' % (video_id, name))
+            return
+        self.to_screen(
+            'Available %s for %s:' % (name, video_id))
+        self.to_screen(render_table(
+            ['Language', 'formats'],
+            [[lang, ', '.join(f['ext'] for f in reversed(formats))]
+                for lang, formats in subtitles.items()]))

 
     def urlopen(self, req):
         """ Start an HTTP download """
 
     def urlopen(self, req):
         """ Start an HTTP download """

-        return self._opener.open(req)
+
+        # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
+        # always respected by websites, some tend to give out URLs with non percent-encoded
+        # non-ASCII characters (see telemb.py, ard.py [#3412])
+        # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
+        # To work around aforementioned issue we will replace request's original URL with
+        # percent-encoded one
+        req_is_string = isinstance(req, compat_basestring)
+        url = req if req_is_string else req.get_full_url()
+        url_escaped = escape_url(url)
+
+        # Substitute URL if any change after escaping
+        if url != url_escaped:
+            if req_is_string:
+                req = url_escaped
+            else:
+                req = compat_urllib_request.Request(
+                    url_escaped, data=req.data, headers=req.headers,
+                    origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
+
+        return self._opener.open(req, timeout=self._socket_timeout)

 
     def print_debug_header(self):
         if not self.params.get('verbose'):
             return
 
     def print_debug_header(self):
         if not self.params.get('verbose'):
             return

-        write_string(u'[debug] youtube-dl version ' + __version__ + u'\n')
+
+        if type('') is not compat_str:
+            # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
+            self.report_warning(
+                'Your Python is broken! Update to a newer and supported version')
+
+        stdout_encoding = getattr(
+            sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
+        encoding_str = (
+            '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
+                locale.getpreferredencoding(),
+                sys.getfilesystemencoding(),
+                stdout_encoding,
+                self.get_encoding()))
+        write_string(encoding_str, encoding=None)
+
+        self._write_string('[debug] youtube-dl version ' + __version__ + '\n')

         try:
             sp = subprocess.Popen(
                 ['git', 'rev-parse', '--short', 'HEAD'],
         try:
             sp = subprocess.Popen(
                 ['git', 'rev-parse', '--short', 'HEAD'],


@@ -1065,24 +1693,46 @@ class YoutubeDL(object):
             out, err = sp.communicate()
             out = out.decode().strip()
             if re.match('[0-9a-f]+', out):
             out, err = sp.communicate()
             out = out.decode().strip()
             if re.match('[0-9a-f]+', out):

-                write_string(u'[debug] Git HEAD: ' + out + u'\n')
+                self._write_string('[debug] Git HEAD: ' + out + '\n')

         except:
             try:
                 sys.exc_clear()
             except:
                 pass
         except:
             try:
                 sys.exc_clear()
             except:
                 pass

-        write_string(u'[debug] Python version %s - %s' %
-                     (platform.python_version(), platform_name()) + u'\n')
+        self._write_string('[debug] Python version %s - %s\n' % (
+            platform.python_version(), platform_name()))
+
+        exe_versions = FFmpegPostProcessor.get_versions(self)
+        exe_versions['rtmpdump'] = rtmpdump_version()
+        exe_str = ', '.join(
+            '%s %s' % (exe, v)
+            for exe, v in sorted(exe_versions.items())
+            if v
+        )
+        if not exe_str:
+            exe_str = 'none'
+        self._write_string('[debug] exe versions: %s\n' % exe_str)

 
         proxy_map = {}
         for handler in self._opener.handlers:
             if hasattr(handler, 'proxies'):
                 proxy_map.update(handler.proxies)
 
         proxy_map = {}
         for handler in self._opener.handlers:
             if hasattr(handler, 'proxies'):
                 proxy_map.update(handler.proxies)

-        write_string(u'[debug] Proxy map: ' + compat_str(proxy_map) + u'\n')
+        self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
+
+        if self.params.get('call_home', False):
+            ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
+            self._write_string('[debug] Public IP address: %s\n' % ipaddr)
+            latest_version = self.urlopen(
+                'https://yt-dl.org/latest/version').read().decode('utf-8')
+            if version_tuple(latest_version) > version_tuple(__version__):
+                self.report_warning(
+                    'You are using an outdated version (newest version: %s)! '
+                    'See https://yt-dl.org/update if you need help updating.' %
+                    latest_version)

 
     def _setup_opener(self):
         timeout_val = self.params.get('socket_timeout')
 
     def _setup_opener(self):
         timeout_val = self.params.get('socket_timeout')

-        timeout = 600 if timeout_val is None else float(timeout_val)
+        self._socket_timeout = 600 if timeout_val is None else float(timeout_val)

 
         opts_cookiefile = self.params.get('cookiefile')
         opts_proxy = self.params.get('proxy')
 
         opts_cookiefile = self.params.get('cookiefile')
         opts_proxy = self.params.get('proxy')


@@ -1108,16 +1758,66 @@ class YoutubeDL(object):
             if 'http' in proxies and 'https' not in proxies:
                 proxies['https'] = proxies['http']
         proxy_handler = compat_urllib_request.ProxyHandler(proxies)
             if 'http' in proxies and 'https' not in proxies:
                 proxies['https'] = proxies['http']
         proxy_handler = compat_urllib_request.ProxyHandler(proxies)

-        https_handler = make_HTTPS_handler(
-            self.params.get('nocheckcertificate', False))
+
+        debuglevel = 1 if self.params.get('debug_printtraffic') else 0
+        https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
+        ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)

         opener = compat_urllib_request.build_opener(
         opener = compat_urllib_request.build_opener(

-            https_handler, proxy_handler, cookie_processor, YoutubeDLHandler())
+            https_handler, proxy_handler, cookie_processor, ydlh)

         # Delete the default user-agent header, which would otherwise apply in
         # cases where our custom HTTP handler doesn't come into play
         # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
         opener.addheaders = []
         self._opener = opener
 
         # Delete the default user-agent header, which would otherwise apply in
         # cases where our custom HTTP handler doesn't come into play
         # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
         opener.addheaders = []
         self._opener = opener
 

-        # TODO remove this global modification
-        compat_urllib_request.install_opener(opener)
-        socket.setdefaulttimeout(timeout)
+    def encode(self, s):
+        if isinstance(s, bytes):
+            return s  # Already encoded
+
+        try:
+            return s.encode(self.get_encoding())
+        except UnicodeEncodeError as err:
+            err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
+            raise
+
+    def get_encoding(self):
+        encoding = self.params.get('encoding')
+        if encoding is None:
+            encoding = preferredencoding()
+        return encoding
+
+    def _write_thumbnails(self, info_dict, filename):
+        if self.params.get('writethumbnail', False):
+            thumbnails = info_dict.get('thumbnails')
+            if thumbnails:
+                thumbnails = [thumbnails[-1]]
+        elif self.params.get('write_all_thumbnails', False):
+            thumbnails = info_dict.get('thumbnails')
+        else:
+            return
+
+        if not thumbnails:
+            # No thumbnails present, so return immediately
+            return
+
+        for t in thumbnails:
+            thumb_ext = determine_ext(t['url'], 'jpg')
+            suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
+            thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
+            thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
+
+            if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
+                self.to_screen('[%s] %s: Thumbnail %sis already present' %
+                               (info_dict['extractor'], info_dict['id'], thumb_display_id))
+            else:
+                self.to_screen('[%s] %s: Downloading thumbnail %s...' %
+                               (info_dict['extractor'], info_dict['id'], thumb_display_id))
+                try:
+                    uf = self.urlopen(t['url'])
+                    with open(thumb_filename, 'wb') as thumbf:
+                        shutil.copyfileobj(uf, thumbf)
+                    self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
+                                   (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
+                except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
+                    self.report_warning('Unable to download thumbnail "%s": %s' %
+                                        (t['url'], compat_str(err)))