-import os
-import re
-import subprocess
-import sys
-import time
-
-from .utils import (
- compat_urllib_error,
- compat_urllib_request,
- ContentTooShortError,
- determine_ext,
- encodeFilename,
- format_bytes,
- sanitize_open,
- timeconvert,
-)
-
-
-class FileDownloader(object):
- """File Downloader class.
-
- File downloader objects are the ones responsible of downloading the
- actual video file and writing it to disk.
-
- File downloaders accept a lot of parameters. In order not to saturate
- the object constructor with arguments, it receives a dictionary of
- options instead.
-
- Available options:
-
- verbose: Print additional info to stdout.
- quiet: Do not print messages to stdout.
- ratelimit: Download speed limit, in bytes/sec.
- retries: Number of times to retry for HTTP error 5xx
- buffersize: Size of download buffer in bytes.
- noresizebuffer: Do not automatically resize the download buffer.
- continuedl: Try to continue downloads if possible.
- noprogress: Do not print the progress bar.
- logtostderr: Log messages to stderr instead of stdout.
- consoletitle: Display progress in console window's titlebar.
- nopart: Do not use temporary .part files.
- updatetime: Use the Last-modified header to set output file timestamps.
- test: Download only first bytes to test the downloader.
- min_filesize: Skip files smaller than this size
- max_filesize: Skip files larger than this size
- """
-
- params = None
-
- def __init__(self, ydl, params):
- """Create a FileDownloader object with the given options."""
- self.ydl = ydl
- self._progress_hooks = []
- self.params = params
-
- @staticmethod
- def format_seconds(seconds):
- (mins, secs) = divmod(seconds, 60)
- (hours, mins) = divmod(mins, 60)
- if hours > 99:
- return '--:--:--'
- if hours == 0:
- return '%02d:%02d' % (mins, secs)
- else:
- return '%02d:%02d:%02d' % (hours, mins, secs)
-
- @staticmethod
- def calc_percent(byte_counter, data_len):
- if data_len is None:
- return None
- return float(byte_counter) / float(data_len) * 100.0
-
- @staticmethod
- def format_percent(percent):
- if percent is None:
- return '---.-%'
- return '%6s' % ('%3.1f%%' % percent)
-
- @staticmethod
- def calc_eta(start, now, total, current):
- if total is None:
- return None
- dif = now - start
- if current == 0 or dif < 0.001: # One millisecond
- return None
- rate = float(current) / dif
- return int((float(total) - float(current)) / rate)
-
- @staticmethod
- def format_eta(eta):
- if eta is None:
- return '--:--'
- return FileDownloader.format_seconds(eta)
-
- @staticmethod
- def calc_speed(start, now, bytes):
- dif = now - start
- if bytes == 0 or dif < 0.001: # One millisecond
- return None
- return float(bytes) / dif
-
- @staticmethod
- def format_speed(speed):
- if speed is None:
- return '%10s' % '---b/s'
- return '%10s' % ('%s/s' % format_bytes(speed))
-
- @staticmethod
- def best_block_size(elapsed_time, bytes):
- new_min = max(bytes / 2.0, 1.0)
- new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
- if elapsed_time < 0.001:
- return int(new_max)
- rate = bytes / elapsed_time
- if rate > new_max:
- return int(new_max)
- if rate < new_min:
- return int(new_min)
- return int(rate)
-
- @staticmethod
- def parse_bytes(bytestr):
- """Parse a string indicating a byte quantity into an integer."""
- matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr)
- if matchobj is None:
- return None
- number = float(matchobj.group(1))
- multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
- return int(round(number * multiplier))
-
- def to_screen(self, *args, **kargs):
- self.ydl.to_screen(*args, **kargs)
-
- def to_stderr(self, message):
- self.ydl.to_screen(message)
-
- def to_console_title(self, message):
- self.ydl.to_console_title(message)
-
- def trouble(self, *args, **kargs):
- self.ydl.trouble(*args, **kargs)
-
- def report_warning(self, *args, **kargs):
- self.ydl.report_warning(*args, **kargs)
-
- def report_error(self, *args, **kargs):
- self.ydl.report_error(*args, **kargs)
-
- 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)
- if rate_limit is None or byte_counter == 0:
- return
- now = time.time()
- elapsed = now - start_time
- if elapsed <= 0.0:
- return
- speed = float(byte_counter) / elapsed
- if speed > rate_limit:
- time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
-
- def temp_name(self, filename):
- """Returns a temporary filename for the given filename."""
- if self.params.get('nopart', False) or filename == u'-' or \
- (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
- return filename
- return filename + u'.part'
-
- def undo_temp_name(self, filename):
- if filename.endswith(u'.part'):
- return filename[:-len(u'.part')]
- return filename
-
- def try_rename(self, old_filename, new_filename):
- try:
- if old_filename == new_filename:
- return
- os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
- except (IOError, OSError):
- 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."""
- if last_modified_hdr is None:
- return
- if not os.path.isfile(encodeFilename(filename)):
- return
- timestr = last_modified_hdr
- if timestr is None:
- return
- filetime = timeconvert(timestr)
- if filetime is None:
- return filetime
- # Ignore obviously invalid dates
- if filetime == 0:
- return
- try:
- os.utime(filename, (time.time(), filetime))
- except:
- pass
- return filetime
-
- def report_destination(self, filename):
- """Report destination filename."""
- self.to_screen(u'[download] Destination: ' + filename)
-
- def _report_progress_status(self, msg, is_last_line=False):
- fullmsg = u'[download] ' + msg
- if self.params.get('progress_with_newline', False):
- self.to_screen(fullmsg)
- else:
- if os.name == 'nt':
- prev_len = getattr(self, '_report_progress_prev_line_length',
- 0)
- if prev_len > len(fullmsg):
- fullmsg += u' ' * (prev_len - len(fullmsg))
- self._report_progress_prev_line_length = len(fullmsg)
- clear_line = u'\r'
- else:
- clear_line = (u'\r\x1b[K' if sys.stderr.isatty() else u'\r')
- self.to_screen(clear_line + fullmsg, skip_eol=not is_last_line)
- self.to_console_title(u'youtube-dl ' + msg)
-
- def report_progress(self, percent, data_len_str, speed, eta):
- """Report download progress."""
- if self.params.get('noprogress', False):
- return
- if eta is not None:
- eta_str = self.format_eta(eta)
- else:
- eta_str = 'Unknown ETA'
- if percent is not None:
- percent_str = self.format_percent(percent)
- else:
- percent_str = 'Unknown %'
- speed_str = self.format_speed(speed)
-
- msg = (u'%s of %s at %s ETA %s' %
- (percent_str, data_len_str, speed_str, eta_str))
- self._report_progress_status(msg)
-
- def report_progress_live_stream(self, downloaded_data_len, speed, elapsed):
- if self.params.get('noprogress', False):
- return
- downloaded_str = format_bytes(downloaded_data_len)
- speed_str = self.format_speed(speed)
- elapsed_str = FileDownloader.format_seconds(elapsed)
- msg = u'%s at %s (%s)' % (downloaded_str, speed_str, elapsed_str)
- self._report_progress_status(msg)
-
- def report_finish(self, data_len_str, tot_time):
- """Report download finished."""
- if self.params.get('noprogress', False):
- self.to_screen(u'[download] Download completed')
- else:
- self._report_progress_status(
- (u'100%% of %s in %s' %
- (data_len_str, self.format_seconds(tot_time))),
- is_last_line=True)
-
- def report_resuming_byte(self, resume_len):
- """Report attempt to resume at given byte."""
- self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
-
- def report_retry(self, count, retries):
- """Report retry in case of HTTP error 5xx"""
- self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
-
- 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)
- except UnicodeEncodeError:
- self.to_screen(u'[download] The file has already been downloaded')
-
- def report_unable_to_resume(self):
- """Report it was impossible to resume download."""
- self.to_screen(u'[download] Unable to resume')
-
- def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path, tc_url, live, conn):
- def run_rtmpdump(args):
- start = time.time()
- resume_percent = None
- resume_downloaded_data_len = None
- proc = subprocess.Popen(args, stderr=subprocess.PIPE)
- cursor_in_new_line = True
- proc_stderr_closed = False
- while not proc_stderr_closed:
- # read line from stderr
- line = u''
- while True:
- char = proc.stderr.read(1)
- if not char:
- proc_stderr_closed = True
- break
- if char in [b'\r', b'\n']:
- break
- line += char.decode('ascii', 'replace')
- if not line:
- # proc_stderr_closed is True
- continue
- mobj = re.search(r'([0-9]+\.[0-9]{3}) kB / [0-9]+\.[0-9]{2} sec \(([0-9]{1,2}\.[0-9])%\)', line)
- if mobj:
- downloaded_data_len = int(float(mobj.group(1))*1024)
- percent = float(mobj.group(2))
- if not resume_percent:
- resume_percent = percent
- resume_downloaded_data_len = downloaded_data_len
- eta = self.calc_eta(start, time.time(), 100-resume_percent, percent-resume_percent)
- speed = self.calc_speed(start, time.time(), downloaded_data_len-resume_downloaded_data_len)
- data_len = None
- if percent > 0:
- data_len = int(downloaded_data_len * 100 / percent)
- data_len_str = u'~' + format_bytes(data_len)
- self.report_progress(percent, data_len_str, speed, eta)
- cursor_in_new_line = False
- self._hook_progress({
- 'downloaded_bytes': downloaded_data_len,
- 'total_bytes': data_len,
- 'tmpfilename': tmpfilename,
- 'filename': filename,
- 'status': 'downloading',
- 'eta': eta,
- 'speed': speed,
- })
- else:
- # no percent for live streams
- mobj = re.search(r'([0-9]+\.[0-9]{3}) kB / [0-9]+\.[0-9]{2} sec', line)
- if mobj:
- downloaded_data_len = int(float(mobj.group(1))*1024)
- time_now = time.time()
- speed = self.calc_speed(start, time_now, downloaded_data_len)
- self.report_progress_live_stream(downloaded_data_len, speed, time_now - start)
- cursor_in_new_line = False
- self._hook_progress({
- 'downloaded_bytes': downloaded_data_len,
- 'tmpfilename': tmpfilename,
- 'filename': filename,
- 'status': 'downloading',
- 'speed': speed,
- })
- elif self.params.get('verbose', False):
- if not cursor_in_new_line:
- self.to_screen(u'')
- cursor_in_new_line = True
- self.to_screen(u'[rtmpdump] '+line)
- proc.wait()
- if not cursor_in_new_line:
- self.to_screen(u'')
- return proc.returncode
-
- self.report_destination(filename)
- tmpfilename = self.temp_name(filename)
- test = self.params.get('test', False)
-
- # Check for rtmpdump first
- try:
- subprocess.call(['rtmpdump', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
- except (OSError, IOError):
- self.report_error(u'RTMP download detected but "rtmpdump" could not be run')
- return False
-
- # Download using rtmpdump. rtmpdump returns exit code 2 when
- # the connection was interrumpted and resuming appears to be
- # possible. This is part of rtmpdump's normal usage, AFAIK.
- basic_args = ['rtmpdump', '--verbose', '-r', url, '-o', tmpfilename]
- if player_url is not None:
- basic_args += ['--swfVfy', player_url]
- if page_url is not None:
- basic_args += ['--pageUrl', page_url]
- if play_path is not None:
- basic_args += ['--playpath', play_path]
- if tc_url is not None:
- basic_args += ['--tcUrl', url]
- if test:
- basic_args += ['--stop', '1']
- if live:
- basic_args += ['--live']
- if conn:
- basic_args += ['--conn', conn]
- args = basic_args + [[], ['--resume', '--skip', '1']][self.params.get('continuedl', False)]
-
- if sys.platform == 'win32' and sys.version_info < (3, 0):
- # Windows subprocess module does not actually support Unicode
- # on Python 2.x
- # See http://stackoverflow.com/a/9951851/35070
- subprocess_encoding = sys.getfilesystemencoding()
- args = [a.encode(subprocess_encoding, 'ignore') for a in args]
- else:
- subprocess_encoding = None
-
- if self.params.get('verbose', False):
- if subprocess_encoding:
- str_args = [
- a.decode(subprocess_encoding) if isinstance(a, bytes) else a
- for a in args]
- else:
- str_args = args
- try:
- import pipes
- shell_quote = lambda args: ' '.join(map(pipes.quote, str_args))
- except ImportError:
- shell_quote = repr
- self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(str_args))
-
- retval = run_rtmpdump(args)
-
- while (retval == 2 or retval == 1) and not test:
- prevsize = os.path.getsize(encodeFilename(tmpfilename))
- self.to_screen(u'[rtmpdump] %s bytes' % prevsize)
- time.sleep(5.0) # This seems to be needed
- retval = run_rtmpdump(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
- cursize = os.path.getsize(encodeFilename(tmpfilename))
- if prevsize == cursize and retval == 1:
- break
- # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
- if prevsize == cursize and retval == 2 and cursize > 1024:
- self.to_screen(u'[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
- retval = 0
- break
- if retval == 0 or (test and retval == 2):
- fsize = os.path.getsize(encodeFilename(tmpfilename))
- self.to_screen(u'[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.to_stderr(u"\n")
- self.report_error(u'rtmpdump exited with code %d' % retval)
- return False
-
- def _download_with_mplayer(self, filename, url):
- self.report_destination(filename)
- tmpfilename = self.temp_name(filename)
-
- args = ['mplayer', '-really-quiet', '-vo', 'null', '-vc', 'dummy', '-dumpstream', '-dumpfile', tmpfilename, url]
- # Check for mplayer first
- try:
- subprocess.call(['mplayer', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
- except (OSError, IOError):
- self.report_error(u'MMS or RTSP download detected but "%s" could not be run' % args[0] )
- return False
-
- # Download using mplayer.
- retval = subprocess.call(args)
- if retval == 0:
- fsize = os.path.getsize(encodeFilename(tmpfilename))
- self.to_screen(u'\r[%s] %s bytes' % (args[0], fsize))
- self.try_rename(tmpfilename, filename)
- self._hook_progress({
- 'downloaded_bytes': fsize,
- 'total_bytes': fsize,
- 'filename': filename,
- 'status': 'finished',
- })
- return True
- else:
- self.to_stderr(u"\n")
- self.report_error(u'mplayer exited with code %d' % retval)
- return False
-
- def _download_m3u8_with_ffmpeg(self, filename, url):
- self.report_destination(filename)
- tmpfilename = self.temp_name(filename)
-
- args = ['-y', '-i', url, '-f', 'mp4', '-c', 'copy',
- '-bsf:a', 'aac_adtstoasc', tmpfilename]
-
- for program in ['avconv', 'ffmpeg']:
- try:
- subprocess.call([program, '-version'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
- break
- except (OSError, IOError):
- pass
- else:
- self.report_error(u'm3u8 download detected but ffmpeg or avconv could not be found')
- cmd = [program] + args
-
- retval = subprocess.call(cmd)
- if retval == 0:
- fsize = os.path.getsize(encodeFilename(tmpfilename))
- self.to_screen(u'\r[%s] %s bytes' % (args[0], fsize))
- self.try_rename(tmpfilename, filename)
- self._hook_progress({
- 'downloaded_bytes': fsize,
- 'total_bytes': fsize,
- 'filename': filename,
- 'status': 'finished',
- })
- return True
- else:
- self.to_stderr(u"\n")
- self.report_error(u'ffmpeg exited with code %d' % retval)
- return False
Rapsys Git / youtubedl / commitdiff