X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/3477c644417600d9ec8f8d2a44f82da0a4b15eb5..a5a28b1eabe0533f26d23c4157c97144215b22b7:/youtube_dl/downloader/fragment.py diff --git a/youtube_dl/downloader/fragment.py b/youtube_dl/downloader/fragment.py index 5a64b29..ea5e3a4 100644 --- a/youtube_dl/downloader/fragment.py +++ b/youtube_dl/downloader/fragment.py @@ -2,12 +2,15 @@ from __future__ import division, unicode_literals import os import time +import json from .common import FileDownloader from .http import HttpFD from ..utils import ( + error_to_compat_str, encodeFilename, sanitize_open, + sanitized_Request, ) @@ -19,14 +22,111 @@ class HttpQuietDownloader(HttpFD): class FragmentFD(FileDownloader): """ A base file downloader class for fragmented media (e.g. f4m/m3u8 manifests). + + Available options: + + fragment_retries: Number of times to retry a fragment for HTTP error (DASH + and hlsnative only) + skip_unavailable_fragments: + Skip unavailable fragments (DASH and hlsnative only) + keep_fragments: Keep downloaded fragments on disk after downloading is + finished + + For each incomplete fragment download youtube-dl keeps on disk a special + bookkeeping file with download state and metadata (in future such files will + be used for any incomplete download handled by youtube-dl). This file is + used to properly handle resuming, check download file consistency and detect + potential errors. The file has a .ytdl extension and represents a standard + JSON file of the following format: + + extractor: + Dictionary of extractor related data. TBD. + + downloader: + Dictionary of downloader related data. May contain following data: + current_fragment: + Dictionary with current (being downloaded) fragment data: + index: 0-based index of current fragment among all fragments + fragment_count: + Total count of fragments + + This feature is experimental and file format may change in future. """ + def report_retry_fragment(self, err, frag_index, count, retries): + self.to_screen( + '[download] Got server HTTP error: %s. Retrying fragment %d (attempt %d of %s)...' + % (error_to_compat_str(err), frag_index, count, self.format_retries(retries))) + + def report_skip_fragment(self, frag_index): + self.to_screen('[download] Skipping fragment %d...' % frag_index) + + def _prepare_url(self, info_dict, url): + headers = info_dict.get('http_headers') + return sanitized_Request(url, None, headers) if headers else url + def _prepare_and_start_frag_download(self, ctx): self._prepare_frag_download(ctx) self._start_frag_download(ctx) + @staticmethod + def __do_ytdl_file(ctx): + return not ctx['live'] and not ctx['tmpfilename'] == '-' + + def _read_ytdl_file(self, ctx): + stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'r') + ctx['fragment_index'] = json.loads(stream.read())['downloader']['current_fragment']['index'] + stream.close() + + def _write_ytdl_file(self, ctx): + frag_index_stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'w') + downloader = { + 'current_fragment': { + 'index': ctx['fragment_index'], + }, + } + if ctx.get('fragment_count') is not None: + downloader['fragment_count'] = ctx['fragment_count'] + frag_index_stream.write(json.dumps({'downloader': downloader})) + frag_index_stream.close() + + def _download_fragment(self, ctx, frag_url, info_dict, headers=None): + fragment_filename = '%s-Frag%d' % (ctx['tmpfilename'], ctx['fragment_index']) + success = ctx['dl'].download(fragment_filename, { + 'url': frag_url, + 'http_headers': headers or info_dict.get('http_headers'), + }) + if not success: + return False, None + down, frag_sanitized = sanitize_open(fragment_filename, 'rb') + ctx['fragment_filename_sanitized'] = frag_sanitized + frag_content = down.read() + down.close() + return True, frag_content + + def _append_fragment(self, ctx, frag_content): + try: + ctx['dest_stream'].write(frag_content) + ctx['dest_stream'].flush() + finally: + if self.__do_ytdl_file(ctx): + self._write_ytdl_file(ctx) + if not self.params.get('keep_fragments', False): + os.remove(encodeFilename(ctx['fragment_filename_sanitized'])) + del ctx['fragment_filename_sanitized'] + def _prepare_frag_download(self, ctx): - self.to_screen('[%s] Total fragments: %d' % (self.FD_NAME, ctx['total_frags'])) + if 'live' not in ctx: + ctx['live'] = False + if not ctx['live']: + total_frags_str = '%d' % ctx['total_frags'] + ad_frags = ctx.get('ad_frags', 0) + if ad_frags: + total_frags_str += ' (not including %d ad)' % ad_frags + else: + total_frags_str = 'unknown (live)' + self.to_screen( + '[%s] Total fragments: %s' % (self.FD_NAME, total_frags_str)) self.report_destination(ctx['filename']) dl = HttpQuietDownloader( self.ydl, @@ -34,17 +134,48 @@ class FragmentFD(FileDownloader): 'continuedl': True, 'quiet': True, 'noprogress': True, - 'ratelimit': self.params.get('ratelimit', None), + 'ratelimit': self.params.get('ratelimit'), 'retries': self.params.get('retries', 0), + 'nopart': self.params.get('nopart', False), 'test': self.params.get('test', False), } ) tmpfilename = self.temp_name(ctx['filename']) - dest_stream, tmpfilename = sanitize_open(tmpfilename, 'wb') + open_mode = 'wb' + resume_len = 0 + + # Establish possible resume length + if os.path.isfile(encodeFilename(tmpfilename)): + open_mode = 'ab' + resume_len = os.path.getsize(encodeFilename(tmpfilename)) + + # Should be initialized before ytdl file check + ctx.update({ + 'tmpfilename': tmpfilename, + 'fragment_index': 0, + }) + + if self.__do_ytdl_file(ctx): + if os.path.isfile(encodeFilename(self.ytdl_filename(ctx['filename']))): + self._read_ytdl_file(ctx) + if ctx['fragment_index'] > 0 and resume_len == 0: + self.report_warning( + 'Inconsistent state of incomplete fragment download. ' + 'Restarting from the beginning...') + ctx['fragment_index'] = resume_len = 0 + self._write_ytdl_file(ctx) + else: + self._write_ytdl_file(ctx) + assert ctx['fragment_index'] == 0 + + dest_stream, tmpfilename = sanitize_open(tmpfilename, open_mode) + ctx.update({ 'dl': dl, 'dest_stream': dest_stream, 'tmpfilename': tmpfilename, + # Total complete fragments downloaded so far in bytes + 'complete_frags_downloaded_bytes': resume_len, }) def _start_frag_download(self, ctx): @@ -53,43 +184,50 @@ class FragmentFD(FileDownloader): # hook state = { 'status': 'downloading', - 'downloaded_bytes': 0, - 'frag_index': 0, - 'frag_count': total_frags, + 'downloaded_bytes': ctx['complete_frags_downloaded_bytes'], + 'fragment_index': ctx['fragment_index'], + 'fragment_count': total_frags, 'filename': ctx['filename'], 'tmpfilename': ctx['tmpfilename'], } + start = time.time() - ctx['started'] = start + ctx.update({ + 'started': start, + # Amount of fragment's bytes downloaded by the time of the previous + # frag progress hook invocation + 'prev_frag_downloaded_bytes': 0, + }) def frag_progress_hook(s): if s['status'] not in ('downloading', 'finished'): return - frag_total_bytes = s.get('total_bytes', 0) - if s['status'] == 'finished': - state['downloaded_bytes'] += frag_total_bytes - state['frag_index'] += 1 - - estimated_size = ( - (state['downloaded_bytes'] + frag_total_bytes) / - (state['frag_index'] + 1) * total_frags) time_now = time.time() - state['total_bytes_estimate'] = estimated_size state['elapsed'] = time_now - start + frag_total_bytes = s.get('total_bytes') or 0 + if not ctx['live']: + estimated_size = ( + (ctx['complete_frags_downloaded_bytes'] + frag_total_bytes) / + (state['fragment_index'] + 1) * total_frags) + state['total_bytes_estimate'] = estimated_size if s['status'] == 'finished': - progress = self.calc_percent(state['frag_index'], total_frags) + state['fragment_index'] += 1 + ctx['fragment_index'] = state['fragment_index'] + state['downloaded_bytes'] += frag_total_bytes - ctx['prev_frag_downloaded_bytes'] + ctx['complete_frags_downloaded_bytes'] = state['downloaded_bytes'] + ctx['prev_frag_downloaded_bytes'] = 0 else: frag_downloaded_bytes = s['downloaded_bytes'] - frag_progress = self.calc_percent(frag_downloaded_bytes, - frag_total_bytes) - progress = self.calc_percent(state['frag_index'], total_frags) - progress += frag_progress / float(total_frags) - - state['eta'] = self.calc_eta( - start, time_now, estimated_size, state['downloaded_bytes'] + frag_downloaded_bytes) - state['speed'] = s.get('speed') + state['downloaded_bytes'] += frag_downloaded_bytes - ctx['prev_frag_downloaded_bytes'] + if not ctx['live']: + state['eta'] = self.calc_eta( + start, time_now, estimated_size, + state['downloaded_bytes']) + state['speed'] = s.get('speed') or ctx.get('speed') + ctx['speed'] = state['speed'] + ctx['prev_frag_downloaded_bytes'] = frag_downloaded_bytes self._hook_progress(state) ctx['dl'].add_progress_hook(frag_progress_hook) @@ -98,6 +236,10 @@ class FragmentFD(FileDownloader): def _finish_frag_download(self, ctx): ctx['dest_stream'].close() + if self.__do_ytdl_file(ctx): + ytdl_filename = encodeFilename(self.ytdl_filename(ctx['filename'])) + if os.path.isfile(ytdl_filename): + os.remove(ytdl_filename) elapsed = time.time() - ctx['started'] self.try_rename(ctx['tmpfilename'], ctx['filename']) fsize = os.path.getsize(encodeFilename(ctx['filename']))