X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/eed74133841b43ce901ee0dd31c7dad5a234bd48..72d91f91be4810269f74cf296cab5dd61b84129c:/youtube_dl/downloader/rtmp.py?ds=sidebyside diff --git a/youtube_dl/downloader/rtmp.py b/youtube_dl/downloader/rtmp.py index 17d9631..89e98ae 100644 --- a/youtube_dl/downloader/rtmp.py +++ b/youtube_dl/downloader/rtmp.py @@ -7,11 +7,10 @@ import sys import time from .common import FileDownloader +from ..compat import compat_str from ..utils import ( check_executable, - compat_str, encodeFilename, - format_bytes, get_exe_version, ) @@ -46,49 +45,49 @@ class RtmpFD(FileDownloader): 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) + 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) + time_now = time.time() + eta = self.calc_eta(start, time_now, 100 - resume_percent, percent - resume_percent) + speed = self.calc_speed(start, time_now, downloaded_data_len - resume_downloaded_data_len) data_len = None if percent > 0: data_len = int(downloaded_data_len * 100 / percent) - data_len_str = '~' + format_bytes(data_len) - self.report_progress(percent, data_len_str, speed, eta) - cursor_in_new_line = False self._hook_progress({ + 'status': 'downloading', 'downloaded_bytes': downloaded_data_len, - 'total_bytes': data_len, + 'total_bytes_estimate': data_len, 'tmpfilename': tmpfilename, 'filename': filename, - 'status': 'downloading', 'eta': eta, + 'elapsed': time_now - start, 'speed': speed, }) + cursor_in_new_line = False 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) + 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', + 'elapsed': time_now - start, 'speed': speed, }) + cursor_in_new_line = False elif self.params.get('verbose', False): if not cursor_in_new_line: self.to_screen('') cursor_in_new_line = True - self.to_screen('[rtmpdump] '+line) + self.to_screen('[rtmpdump] ' + line) proc.wait() if not cursor_in_new_line: self.to_screen('') @@ -104,6 +103,9 @@ class RtmpFD(FileDownloader): live = info_dict.get('rtmp_live', False) conn = info_dict.get('rtmp_conn', None) protocol = info_dict.get('rtmp_protocol', None) + real_time = info_dict.get('rtmp_real_time', False) + no_resume = info_dict.get('no_resume', False) + continue_dl = info_dict.get('continuedl', False) self.report_destination(filename) tmpfilename = self.temp_name(filename) @@ -117,7 +119,9 @@ class RtmpFD(FileDownloader): # 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] + basic_args = [ + 'rtmpdump', '--verbose', '-r', url, + '-o', encodeFilename(tmpfilename, True)] if player_url is not None: basic_args += ['--swfVfy', player_url] if page_url is not None: @@ -141,7 +145,14 @@ class RtmpFD(FileDownloader): basic_args += ['--conn', conn] if protocol is not None: basic_args += ['--protocol', protocol] - args = basic_args + [[], ['--resume', '--skip', '1']][not live and self.params.get('continuedl', False)] + if real_time: + basic_args += ['--realtime'] + + args = basic_args + if not no_resume and continue_dl and not live: + args += ['--resume'] + if not live and continue_dl: + args += ['--skip', '1'] if sys.platform == 'win32' and sys.version_info < (3, 0): # Windows subprocess module does not actually support Unicode @@ -152,19 +163,7 @@ class RtmpFD(FileDownloader): 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('[debug] rtmpdump command line: ' + shell_quote(str_args)) + self._debug_cmd(args, subprocess_encoding, exe='rtmpdump') RD_SUCCESS = 0 RD_FAILED = 1 @@ -180,12 +179,12 @@ class RtmpFD(FileDownloader): while (retval == RD_INCOMPLETE or retval == RD_FAILED) and not test and not live: prevsize = os.path.getsize(encodeFilename(tmpfilename)) self.to_screen('[rtmpdump] %s bytes' % prevsize) - time.sleep(5.0) # This seems to be needed + time.sleep(5.0) # This seems to be needed retval = run_rtmpdump(basic_args + ['-e'] + [[], ['-k', '1']][retval == RD_FAILED]) cursize = os.path.getsize(encodeFilename(tmpfilename)) if prevsize == cursize and retval == RD_FAILED: break - # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those + # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those if prevsize == cursize and retval == RD_INCOMPLETE and cursize > 1024: self.to_screen('[rtmpdump] Could not download the whole video. This can happen for some advertisements.') retval = RD_SUCCESS