]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/FileDownloader.py
debian/NEWS: Add link to my webpage to give more details.
[youtubedl] / youtube_dl / FileDownloader.py
index d6673fd3ab88d543086417f903cf2ff7d4019944..5c8e676a20bdb112452e2a8b0815422ca3bf7c73 100644 (file)
-import math
-import os
-import re
-import subprocess
-import sys
-import time
-import traceback
-
-if os.name == 'nt':
-    import ctypes
-
-from .utils import *
-
-
-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_bytes(bytes):
-        if bytes is None:
-            return 'N/A'
-        if type(bytes) is str:
-            bytes = float(bytes)
-        if bytes == 0.0:
-            exponent = 0
-        else:
-            exponent = int(math.log(bytes, 1024.0))
-        suffix = ['B','KiB','MiB','GiB','TiB','PiB','EiB','ZiB','YiB'][exponent]
-        converted = float(bytes) / float(1024 ** exponent)
-        return '%.2f%s' % (converted, suffix)
-
-    @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' % FileDownloader.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_cons_title(self, message):
-        """Set console/terminal window title to message."""
-        if not self.params.get('consoletitle', False):
-            return
-        if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
-            # c_wchar_p() might not be necessary if `message` is
-            # already of type unicode()
-            ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
-        elif 'TERM' in os.environ:
-            self.to_screen('\033]0;%s\007' % message, skip_eol=True)
-
-    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) as err:
-            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(self, percent, data_len_str, speed, eta):
-        """Report download progress."""
-        if self.params.get('noprogress', False):
-            return
-        clear_line = (u'\x1b[K' if sys.stderr.isatty() and os.name != 'nt' else u'')
-        eta_str = self.format_eta(eta)
-        percent_str = self.format_percent(percent)
-        speed_str = self.format_speed(speed)
-        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%s[download] %s of %s at %s ETA %s' %
-                (clear_line, 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()))
-
-    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) as err:
-            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 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:
-            clear_line = (u'\x1b[K' if sys.stderr.isatty() and os.name != 'nt' else u'')
-            self.to_screen(u'\r%s[download] 100%% of %s in %s' %
-                (clear_line, data_len_str, self.format_seconds(tot_time)))
-
-    def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path, tc_url):
-        self.report_destination(filename)
-        tmpfilename = self.temp_name(filename)
-
-        # 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
-        verbosity_option = '--verbose' if self.params.get('verbose', False) else '--quiet'
-
-        # 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', verbosity_option, '-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]
-        args = basic_args + [[], ['--resume', '--skip', '1']][self.params.get('continuedl', False)]
-        if self.params.get('verbose', False):
-            try:
-                import pipes
-                shell_quote = lambda args: ' '.join(map(pipes.quote, args))
-            except ImportError:
-                shell_quote = repr
-            self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
-        retval = subprocess.call(args)
-        while retval == 2 or retval == 1:
-            prevsize = os.path.getsize(encodeFilename(tmpfilename))
-            self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
-            time.sleep(5.0) # This seems to be needed
-            retval = subprocess.call(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'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
-                retval = 0
-                break
-        if retval == 0:
-            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.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 = ['ffmpeg', '-y', '-i', url, '-f', 'mp4', tmpfilename]
-        # Check for ffmpeg first
-        try:
-            subprocess.call(['ffmpeg', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
-        except (OSError, IOError):
-            self.report_error(u'm3u8 download detected but "%s" could not be run' % args[0] )
-            return False
-
-        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'ffmpeg exited with code %d' % retval)
-            return False
+# Legacy file for backwards compatibility, use youtube_dl.downloader instead!
+from .downloader import FileDownloader as RealFileDownloader
+from .downloader import get_suitable_downloader
 
 
+# This class reproduces the old behaviour of FileDownloader
+class FileDownloader(RealFileDownloader):
     def _do_download(self, filename, info_dict):
-        url = info_dict['url']
-
-        # 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',
-                'total_bytes': os.path.getsize(encodeFilename(filename)),
-            })
-            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('play_path', None),
-                                                info_dict.get('tc_url', None))
-
-        # Attempt to download using mplayer
-        if url.startswith('mms') or url.startswith('rtsp'):
-            return self._download_with_mplayer(filename, url)
-
-        # m3u8 manifest are downloaded with ffmpeg
-        if determine_ext(url) == u'm3u8':
-            return self._download_m3u8_with_ffmpeg(filename, url)
-
-        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)
-
-        if self.params.get('test', False):
-            request.add_header('Range','bytes=0-10240')
-
-        # Establish possible resume length
-        if os.path.isfile(encodeFilename(tmpfilename)):
-            resume_len = os.path.getsize(encodeFilename(tmpfilename))
-        else:
-            resume_len = 0
-
-        open_mode = 'wb'
-        if resume_len != 0:
-            if self.params.get('continuedl', False):
-                self.report_resuming_byte(resume_len)
-                request.add_header('Range','bytes=%d-' % resume_len)
-                open_mode = 'ab'
-            else:
-                resume_len = 0
-
-        count = 0
-        retries = self.params.get('retries', 0)
-        while count <= retries:
-            # Establish connection
-            try:
-                if count == 0 and 'urlhandle' in info_dict:
-                    data = info_dict['urlhandle']
-                data = compat_urllib_request.urlopen(request)
-                break
-            except (compat_urllib_error.HTTPError, ) as err:
-                if (err.code < 500 or err.code >= 600) and err.code != 416:
-                    # Unexpected HTTP error
-                    raise
-                elif err.code == 416:
-                    # Unable to resume (requested range not satisfiable)
-                    try:
-                        # Open the connection again without the range header
-                        data = compat_urllib_request.urlopen(basic_request)
-                        content_length = data.info()['Content-Length']
-                    except (compat_urllib_error.HTTPError, ) as err:
-                        if err.code < 500 or err.code >= 600:
-                            raise
-                    else:
-                        # Examine the reported length
-                        if (content_length is not None and
-                                (resume_len - 100 < int(content_length) < resume_len + 100)):
-                            # The file had already been fully downloaded.
-                            # Explanation to the above condition: in issue #175 it was revealed that
-                            # YouTube sometimes adds or removes a few bytes from the end of the file,
-                            # changing the file size slightly and causing problems for some users. So
-                            # I decided to implement a suggested change and consider the file
-                            # completely downloaded if the file size differs less than 100 bytes from
-                            # 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_unable_to_resume()
-                            open_mode = 'wb'
-                            break
-            # Retry
-            count += 1
-            if count <= retries:
-                self.report_retry(count, retries)
-
-        if count > 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)
-        start = time.time()
-        while True:
-            # Download and write
-            before = time.time()
-            data_block = data.read(block_size)
-            after = time.time()
-            if len(data_block) == 0:
-                break
-            byte_counter += len(data_block)
-
-            # Open file just in time
-            if stream is None:
-                try:
-                    (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
-                    assert stream is not None
-                    filename = self.undo_temp_name(tmpfilename)
-                    self.report_destination(filename)
-                except (OSError, IOError) as 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.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))
-
-            # Progress message
-            speed = self.calc_speed(start, time.time(), byte_counter - resume_len)
-            if data_len is None:
-                self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
-                eta = None
-            else:
-                percent = self.calc_percent(byte_counter, data_len)
-                eta = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
-                self.report_progress(percent, data_len_str, speed, eta)
-
-            self._hook_progress({
-                'downloaded_bytes': byte_counter,
-                'total_bytes': data_len,
-                'tmpfilename': tmpfilename,
-                'filename': filename,
-                'status': 'downloading',
-                'eta': eta,
-                'speed': speed,
-            })
-
-            # Apply rate limit
-            self.slow_down(start, byte_counter - resume_len)
-
-        if stream is None:
-            self.to_stderr(u"\n")
-            self.report_error(u'Did not get any data blocks')
-            return False
-        stream.close()
-        self.report_finish(data_len_str, (time.time() - start))
-        if data_len is not None and byte_counter != data_len:
-            raise ContentTooShortError(byte_counter, int(data_len))
-        self.try_rename(tmpfilename, filename)
-
-        # Update file modification time
-        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):
+        real_fd = get_suitable_downloader(info_dict)(self.ydl, self.params)
         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
-        * eta: The estimated time in seconds, None if unknown
-        * speed: The download speed in bytes/second, None if unknown
-
-        Hooks are guaranteed to be called at least once (with status "finished")
-        if the download is successful.
-        """
-        self._progress_hooks.append(ph)
+            real_fd.add_progress_hook(ph)
+        return real_fd.download(filename, info_dict)