import ctypes
from .utils import *
+from .InfoExtractors import get_info_extractor
class FileDownloader(object):
updatetime: Use the Last-modified header to set output file timestamps.
writedescription: Write the video description to a .description file
writeinfojson: Write the video description to a .info.json file
- writesubtitles: Write the video subtitles to a .srt file
+ writesubtitles: Write the video subtitles to a file
+ onlysubtitles: Downloads only the subtitles of the video
+ allsubtitles: Downloads all the subtitles of the video
+ listsubtitles: Lists all available subtitles for the video
+ subtitlesformat: Subtitle format [sbv/srt] (default=srt)
subtitleslang: Language of the subtitles to download
test: Download only first bytes to test the downloader.
+ keepvideo: Keep the video file after post-processing
+ min_filesize: Skip files smaller than this size
+ max_filesize: Skip files larger than this size
"""
params = None
"""Create a FileDownloader object with the given options."""
self._ies = []
self._pps = []
+ self._progress_hooks = []
self._download_retcode = 0
self._num_downloads = 0
self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
self.params = params
if '%(stitle)s' in self.params['outtmpl']:
- self.to_stderr(u'WARNING: %(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
+ 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.')
@staticmethod
def format_bytes(bytes):
# already of type unicode()
ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
elif 'TERM' in os.environ:
- sys.stderr.write('\033]0;%s\007' % message.encode(preferredencoding()))
+ self.to_screen('\033]0;%s\007' % message, skip_eol=True)
def fixed_template(self):
"""Checks if the output template is fixed."""
self.to_stderr(message)
if self.params.get('verbose'):
if tb is None:
- tb_data = traceback.format_list(traceback.extract_stack())
- tb = u''.join(tb_data)
+ if sys.exc_info()[0]: # if .trouble has been called from an except block
+ tb = u''
+ 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 += compat_str(traceback.format_exc())
+ else:
+ tb_data = traceback.format_list(traceback.extract_stack())
+ tb = u''.join(tb_data)
self.to_stderr(tb)
if not self.params.get('ignoreerrors', False):
- raise DownloadError(message)
+ if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
+ exc_info = sys.exc_info()[1].exc_info
+ else:
+ exc_info = sys.exc_info()
+ raise DownloadError(message, exc_info)
self._download_retcode = 1
+ def report_warning(self, message):
+ '''
+ Print the message to stderr, it will be prefixed with 'WARNING:'
+ If stderr is a tty file the 'WARNING:' will be colored
+ '''
+ if sys.stderr.isatty():
+ _msg_header=u'\033[0;33mWARNING:\033[0m'
+ else:
+ _msg_header=u'WARNING:'
+ warning_message=u'%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.
+ '''
+ if sys.stderr.isatty():
+ _msg_header = u'\033[0;31mERROR:\033[0m'
+ else:
+ _msg_header = u'ERROR:'
+ error_message = u'%s %s' % (_msg_header, message)
+ self.trouble(error_message, tb)
+
def slow_down(self, start_time, byte_counter):
"""Sleep if the download speed is over the rate limit."""
rate_limit = self.params.get('ratelimit', None)
return
os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
except (IOError, OSError) as err:
- self.trouble(u'ERROR: unable to rename file')
+ self.report_error(u'unable to rename file')
def try_utime(self, filename, last_modified_hdr):
"""Try to set the last-modified time of the given file."""
""" Report that the description file is being written """
self.to_screen(u'[info] Writing video description to: ' + descfn)
- def report_writesubtitles(self, srtfn):
+ def report_writesubtitles(self, sub_filename):
""" Report that the subtitles file is being written """
- self.to_screen(u'[info] Writing video subtitles to: ' + srtfn)
+ self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
def report_writeinfojson(self, infofn):
""" Report that the metadata file has been written """
"""Report download progress."""
if self.params.get('noprogress', False):
return
- self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
+ if self.params.get('progress_with_newline', False):
+ self.to_screen(u'[download] %s of %s at %s ETA %s' %
+ (percent_str, data_len_str, speed_str, eta_str))
+ else:
+ self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
(percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
(percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
template_dict = dict(info_dict)
template_dict['epoch'] = int(time.time())
- template_dict['autonumber'] = u'%05d' % self._num_downloads
+ autonumber_size = self.params.get('autonumber_size')
+ if autonumber_size is None:
+ autonumber_size = 5
+ autonumber_templ = u'%0' + str(autonumber_size) + u'd'
+ template_dict['autonumber'] = autonumber_templ % self._num_downloads
+ if template_dict['playlist_index'] is not None:
+ template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
sanitize = lambda k,v: sanitize_filename(
u'NA' if v is None else compat_str(v),
filename = self.params['outtmpl'] % template_dict
return filename
- except (ValueError, KeyError) as err:
- self.trouble(u'ERROR: invalid system charset or erroneous output template')
+ except KeyError as err:
+ self.trouble(u'ERROR: Erroneous output template')
+ return None
+ except ValueError as err:
+ self.trouble(u'ERROR: Insufficient system charset ' + repr(preferredencoding()))
return None
def _match_entry(self, info_dict):
title = info_dict['title']
matchtitle = self.params.get('matchtitle', False)
if matchtitle:
- matchtitle = matchtitle.decode('utf8')
if not re.search(matchtitle, title, re.IGNORECASE):
return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
rejecttitle = self.params.get('rejecttitle', False)
if rejecttitle:
- rejecttitle = rejecttitle.decode('utf8')
if re.search(rejecttitle, title, re.IGNORECASE):
return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
return None
+
+ def extract_info(self, url, download = True, ie_name = None):
+ '''
+ Returns a list with a dictionary for each video we find.
+ If 'download', also downloads the videos.
+ '''
+ suitable_found = False
+
+ #We copy the original list
+ ies = list(self._ies)
+
+ if ie_name is not None:
+ #We put in the first place the given info extractor
+ first_ie = get_info_extractor(ie_name)()
+ first_ie.set_downloader(self)
+ ies.insert(0, first_ie)
+
+ for ie in ies:
+ # Go to next InfoExtractor if not suitable
+ if not ie.suitable(url):
+ continue
+
+ # Warn if the _WORKING attribute is False
+ if not ie.working():
+ self.to_stderr(u'WARNING: the program functionality for this site has been marked as broken, '
+ u'and will probably not work. If you want to go on, use the -i option.')
+
+ # Suitable InfoExtractor found
+ suitable_found = True
+
+ # Extract information from URL and process it
+ try:
+ ie_results = ie.extract(url)
+ if ie_results is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
+ break
+ results = []
+ for ie_result in ie_results:
+ if not 'extractor' in ie_result:
+ #The extractor has already been set somewhere else
+ ie_result['extractor'] = ie.IE_NAME
+ results.append(self.process_ie_result(ie_result, download))
+ return results
+ except ExtractorError as de: # An error we somewhat expected
+ self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback())
+ break
+ except Exception as e:
+ if self.params.get('ignoreerrors', False):
+ self.trouble(u'ERROR: ' + compat_str(e), tb=compat_str(traceback.format_exc()))
+ break
+ else:
+ raise
+ if not suitable_found:
+ self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url)
+
+ def process_ie_result(self, ie_result, download = True):
+ """
+ Take the result of the ie and return a list of videos.
+ For url elements it will search the suitable ie and get the videos
+ For playlist elements it will process each of the elements of the 'entries' key
+
+ It will also download the videos if 'download'.
+ """
+ result_type = ie_result.get('_type', 'video') #If not given we suppose it's a video, support the dafault old system
+ if result_type == 'video':
+ if 'playlist' not in ie_result:
+ #It isn't part of a playlist
+ ie_result['playlist'] = None
+ ie_result['playlist_index'] = None
+ if download:
+ #Do the download:
+ self.process_info(ie_result)
+ return ie_result
+ elif result_type == 'url':
+ #We get the video pointed by the url
+ result = self.extract_info(ie_result['url'], download, ie_name = ie_result['ie_key'])[0]
+ return result
+ elif result_type == 'playlist':
+ #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)
+
+ playlist_results = []
+
+ n_all_entries = len(ie_result['entries'])
+ playliststart = self.params.get('playliststart', 1) - 1
+ playlistend = self.params.get('playlistend', -1)
+
+ if playlistend == -1:
+ entries = ie_result['entries'][playliststart:]
+ else:
+ entries = ie_result['entries'][playliststart:playlistend]
+
+ n_entries = len(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))
+
+ for i,entry in enumerate(entries,1):
+ self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
+ entry_result = self.process_ie_result(entry, False)
+ entry_result['playlist'] = playlist
+ entry_result['playlist_index'] = i + playliststart
+ #We must do the download here to correctly set the 'playlist' key
+ if download:
+ self.process_info(entry_result)
+ playlist_results.append(entry_result)
+ result = ie_result.copy()
+ result['entries'] = playlist_results
+ return result
def process_info(self, info_dict):
"""Process a single dictionary returned by an InfoExtractor."""
+ #We increment the download the download count here to match the previous behaviour.
+ self.increment_downloads()
+
+ info_dict['fulltitle'] = info_dict['title']
+ if len(info_dict['title']) > 200:
+ info_dict['title'] = info_dict['title'][:197] + u'...'
+
# Keep for backwards compatibility
info_dict['stitle'] = info_dict['title']
if dn != '' and not os.path.exists(dn): # dn is already encoded
os.makedirs(dn)
except (OSError, IOError) as err:
- self.trouble(u'ERROR: unable to create directory ' + compat_str(err))
+ self.report_error(u'unable to create directory ' + compat_str(err))
return
if self.params.get('writedescription', False):
with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
descfile.write(info_dict['description'])
except (OSError, IOError):
- self.trouble(u'ERROR: Cannot write description file ' + descfn)
+ self.report_error(u'Cannot write description file ' + descfn)
return
if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
# subtitles download errors are already managed as troubles in relevant IE
# that way it will silently go on when used with unsupporting IE
- try:
- srtfn = filename.rsplit('.', 1)[0] + u'.srt'
- self.report_writesubtitles(srtfn)
- with io.open(encodeFilename(srtfn), 'w', encoding='utf-8') as srtfile:
- srtfile.write(info_dict['subtitles'])
- except (OSError, IOError):
- self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
- return
+ subtitle = info_dict['subtitles'][0]
+ (sub_error, sub_lang, sub) = subtitle
+ sub_format = self.params.get('subtitlesformat')
+ if sub_error:
+ self.report_warning("Some error while getting the subtitles")
+ else:
+ try:
+ sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
+ self.report_writesubtitles(sub_filename)
+ with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
+ subfile.write(sub)
+ except (OSError, IOError):
+ self.report_error(u'Cannot write subtitles file ' + descfn)
+ return
+ if self.params.get('onlysubtitles', False):
+ return
+
+ if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
+ subtitles = info_dict['subtitles']
+ sub_format = self.params.get('subtitlesformat')
+ for subtitle in subtitles:
+ (sub_error, sub_lang, sub) = subtitle
+ if sub_error:
+ self.report_warning("Some error while getting the subtitles")
+ else:
+ try:
+ sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
+ self.report_writesubtitles(sub_filename)
+ with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
+ subfile.write(sub)
+ except (OSError, IOError):
+ self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
+ return
+ if self.params.get('onlysubtitles', False):
+ return
if self.params.get('writeinfojson', False):
infofn = filename + u'.info.json'
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))
except (OSError, IOError):
- self.trouble(u'ERROR: Cannot write metadata to JSON file ' + infofn)
+ self.report_error(u'Cannot write metadata to JSON file ' + infofn)
return
if not self.params.get('skip_download', False):
except (OSError, IOError) as err:
raise UnavailableVideoError()
except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
- self.trouble(u'ERROR: unable to download video data: %s' % str(err))
+ self.report_error(u'unable to download video data: %s' % str(err))
return
except (ContentTooShortError, ) as err:
- self.trouble(u'ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
+ self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
return
if success:
try:
self.post_process(filename, info_dict)
except (PostProcessingError) as err:
- self.trouble(u'ERROR: postprocessing: %s' % str(err))
+ self.report_error(u'postprocessing: %s' % str(err))
return
def download(self, url_list):
raise SameFileError(self.params['outtmpl'])
for url in url_list:
- suitable_found = False
- for ie in self._ies:
- # Go to next InfoExtractor if not suitable
- if not ie.suitable(url):
- continue
-
- # Warn if the _WORKING attribute is False
- if not ie.working():
- self.to_stderr(u'WARNING: the program functionality for this site has been marked as broken, '
- u'and will probably not work. If you want to go on, use the -i option.')
-
- # Suitable InfoExtractor found
- suitable_found = True
-
- # Extract information from URL and process it
- try:
- videos = ie.extract(url)
- except ExtractorError as de: # An error we somewhat expected
- self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback())
- break
- except Exception as e:
- if self.params.get('ignoreerrors', False):
- self.trouble(u'ERROR: ' + compat_str(e), tb=compat_str(traceback.format_exc()))
- break
- else:
- raise
-
- if len(videos or []) > 1 and self.fixed_template():
- raise SameFileError(self.params['outtmpl'])
-
- for video in videos or []:
- video['extractor'] = ie.IE_NAME
- try:
- self.increment_downloads()
- self.process_info(video)
- except UnavailableVideoError:
- self.trouble(u'\nERROR: unable to download video')
-
- # Suitable InfoExtractor had been found; go to next URL
- break
-
- if not suitable_found:
- self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url)
+ try:
+ #It also downloads the videos
+ videos = self.extract_info(url)
+ except UnavailableVideoError:
+ self.trouble(u'\nERROR: unable to download video')
+ except MaxDownloadsReached:
+ self.to_screen(u'[info] Maximum number of downloaded files reached.')
+ raise
return self._download_retcode
def post_process(self, filename, ie_info):
- """Run the postprocessing chain on the given file."""
+ """Run all the postprocessors on the given file."""
info = dict(ie_info)
info['filepath'] = filename
+ keep_video = None
for pp in self._pps:
- info = pp.run(info)
- if info is None:
- break
+ try:
+ keep_video_wish,new_info = pp.run(info)
+ if keep_video_wish is not None:
+ if keep_video_wish:
+ keep_video = keep_video_wish
+ elif keep_video is None:
+ # No clear decision yet, let IE decide
+ keep_video = keep_video_wish
+ except PostProcessingError as e:
+ self.to_stderr(u'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')
- def _download_with_rtmpdump(self, filename, url, player_url, page_url):
+ def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path):
self.report_destination(filename)
tmpfilename = self.temp_name(filename)
# Check for rtmpdump first
try:
- subprocess.call(['rtmpdump', '-h'], stdout=(file(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
+ subprocess.call(['rtmpdump', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
except (OSError, IOError):
- self.trouble(u'ERROR: RTMP download detected but "rtmpdump" could not be run')
+ self.report_error(u'RTMP download detected but "rtmpdump" could not be run')
return False
# Download using rtmpdump. rtmpdump returns exit code 2 when
basic_args += ['-W', player_url]
if page_url is not None:
basic_args += ['--pageUrl', page_url]
+ if play_path is not None:
+ basic_args += ['-y', play_path]
args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)]
if self.params.get('verbose', False):
try:
retval = 0
break
if retval == 0:
- self.to_screen(u'\r[rtmpdump] %s bytes' % os.path.getsize(encodeFilename(tmpfilename)))
+ fsize = os.path.getsize(encodeFilename(tmpfilename))
+ self.to_screen(u'\r[rtmpdump] %s bytes' % fsize)
self.try_rename(tmpfilename, filename)
+ self._hook_progress({
+ 'downloaded_bytes': fsize,
+ 'total_bytes': fsize,
+ 'filename': filename,
+ 'status': 'finished',
+ })
return True
else:
- self.trouble(u'\nERROR: rtmpdump exited with code %d' % retval)
+ self.to_stderr(u"\n")
+ self.report_error(u'rtmpdump exited with code %d' % retval)
return False
def _do_download(self, filename, info_dict):
# Check file already present
if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
self.report_file_already_downloaded(filename)
+ self._hook_progress({
+ 'filename': filename,
+ 'status': 'finished',
+ })
return True
# Attempt to download using rtmpdump
if url.startswith('rtmp'):
return self._download_with_rtmpdump(filename, url,
info_dict.get('player_url', None),
- info_dict.get('page_url', None))
+ info_dict.get('page_url', None),
+ info_dict.get('play_path', None))
tmpfilename = self.temp_name(filename)
stream = None
# Do not include the Accept-Encoding header
headers = {'Youtubedl-no-compression': 'True'}
+ if 'user_agent' in info_dict:
+ headers['Youtubedl-user-agent'] = info_dict['user_agent']
basic_request = compat_urllib_request.Request(url, None, headers)
request = compat_urllib_request.Request(url, None, headers)
# the one in the hard drive.
self.report_file_already_downloaded(filename)
self.try_rename(tmpfilename, filename)
+ self._hook_progress({
+ 'filename': filename,
+ 'status': 'finished',
+ })
return True
else:
# The length does not match, we start the download over
self.report_retry(count, retries)
if count > retries:
- self.trouble(u'ERROR: giving up after %s retries' % retries)
+ self.report_error(u'giving up after %s retries' % retries)
return False
data_len = data.info().get('Content-length', None)
if data_len is not None:
data_len = int(data_len) + resume_len
+ min_data_len = self.params.get("min_filesize", None)
+ max_data_len = self.params.get("max_filesize", None)
+ if min_data_len is not None and data_len < min_data_len:
+ self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
+ return False
+ if max_data_len is not None and data_len > max_data_len:
+ self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
+ return False
+
data_len_str = self.format_bytes(data_len)
byte_counter = 0 + resume_len
block_size = self.params.get('buffersize', 1024)
filename = self.undo_temp_name(tmpfilename)
self.report_destination(filename)
except (OSError, IOError) as err:
- self.trouble(u'ERROR: unable to open for writing: %s' % str(err))
+ self.report_error(u'unable to open for writing: %s' % str(err))
return False
try:
stream.write(data_block)
except (IOError, OSError) as err:
- self.trouble(u'\nERROR: unable to write data: %s' % str(err))
+ self.to_stderr(u"\n")
+ self.report_error(u'unable to write data: %s' % str(err))
return False
if not self.params.get('noresizebuffer', False):
block_size = self.best_block_size(after - before, len(data_block))
eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
self.report_progress(percent_str, data_len_str, speed_str, eta_str)
+ self._hook_progress({
+ 'downloaded_bytes': byte_counter,
+ 'total_bytes': data_len,
+ 'tmpfilename': tmpfilename,
+ 'filename': filename,
+ 'status': 'downloading',
+ })
+
# Apply rate limit
self.slow_down(start, byte_counter - resume_len)
if stream is None:
- self.trouble(u'\nERROR: Did not get any data blocks')
+ self.to_stderr(u"\n")
+ self.report_error(u'Did not get any data blocks')
return False
stream.close()
self.report_finish()
if self.params.get('updatetime', True):
info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
+ self._hook_progress({
+ 'downloaded_bytes': byte_counter,
+ 'total_bytes': byte_counter,
+ 'filename': filename,
+ 'status': 'finished',
+ })
+
return True
+
+ def _hook_progress(self, status):
+ for ph in self._progress_hooks:
+ ph(status)
+
+ def add_progress_hook(self, ph):
+ """ ph gets called on download progress, with a dictionary with the entries
+ * filename: The final filename
+ * status: One of "downloading" and "finished"
+
+ It can also have some of the following entries:
+
+ * downloaded_bytes: Bytes on disks
+ * total_bytes: Total bytes, None if unknown
+ * tmpfilename: The filename we're currently writing to
+
+ Hooks are guaranteed to be called at least once (with status "finished")
+ if the download is successful.
+ """
+ self._progress_hooks.append(ph)
projects / youtube-dl / blobdiff