From d3f0687cf7b049b976420056e02c26b5d96adeed Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Mon, 24 Apr 2017 02:54:17 +0700 Subject: [PATCH] [downloader/fragment] Use temp file for current fragment --- youtube_dl/downloader/fragment.py | 59 ++++++++++++++++++++----------- youtube_dl/downloader/http.py | 8 ++--- 2 files changed, 41 insertions(+), 26 deletions(-) diff --git a/youtube_dl/downloader/fragment.py b/youtube_dl/downloader/fragment.py index 0a3b1ece0..62de0a75b 100644 --- a/youtube_dl/downloader/fragment.py +++ b/youtube_dl/downloader/fragment.py @@ -2,7 +2,6 @@ from __future__ import division, unicode_literals import os import time -import io import json from .common import FileDownloader @@ -48,28 +47,42 @@ class FragmentFD(FileDownloader): self._prepare_frag_download(ctx) self._start_frag_download(ctx) + def _read_ytdl_file(self, ctx): + stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'r') + ctx['fragment_index'] = json.loads(stream.read())['download']['current_fragment_index'] + stream.close() + + def _write_ytdl_file(self, ctx): + frag_index_stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'w') + frag_index_stream.write(json.dumps({ + 'download': { + 'current_fragment_index': ctx['fragment_index'] + }, + })) + frag_index_stream.close() + def _download_fragment(self, ctx, frag_url, info_dict, headers=None): - down = io.BytesIO() - success = ctx['dl'].download(down, { + 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 - frag_content = down.getvalue() + 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): - ctx['dest_stream'].write(frag_content) - if not (ctx.get('live') or ctx['tmpfilename'] == '-'): - frag_index_stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'w') - frag_index_stream.write(json.dumps({ - 'download': { - 'last_fragment_index': ctx['fragment_index'] - }, - })) - frag_index_stream.close() + try: + ctx['dest_stream'].write(frag_content) + finally: + if not (ctx.get('live') or ctx['tmpfilename'] == '-'): + self._write_ytdl_file(ctx) + os.remove(ctx['fragment_filename_sanitized']) + del ctx['fragment_filename_sanitized'] def _prepare_frag_download(self, ctx): if 'live' not in ctx: @@ -93,23 +106,29 @@ class FragmentFD(FileDownloader): tmpfilename = self.temp_name(ctx['filename']) open_mode = 'wb' resume_len = 0 - frag_index = 0 + # Establish possible resume length if os.path.isfile(encodeFilename(tmpfilename)): open_mode = 'ab' resume_len = os.path.getsize(encodeFilename(tmpfilename)) - ytdl_filename = encodeFilename(self.ytdl_filename(ctx['filename'])) - if os.path.isfile(ytdl_filename): - frag_index_stream, _ = sanitize_open(ytdl_filename, 'r') - frag_index = json.loads(frag_index_stream.read())['download']['last_fragment_index'] - frag_index_stream.close() + + ctx['fragment_index'] = 0 + if os.path.isfile(encodeFilename(self.ytdl_filename(ctx['filename']))): + self._read_ytdl_file(ctx) + else: + self._write_ytdl_file(ctx) + + if ctx['fragment_index'] > 0: + assert resume_len > 0 + else: + assert resume_len == 0 + dest_stream, tmpfilename = sanitize_open(tmpfilename, open_mode) ctx.update({ 'dl': dl, 'dest_stream': dest_stream, 'tmpfilename': tmpfilename, - 'fragment_index': frag_index, # Total complete fragments downloaded so far in bytes 'complete_frags_downloaded_bytes': resume_len, }) diff --git a/youtube_dl/downloader/http.py b/youtube_dl/downloader/http.py index 2896a17af..af405b950 100644 --- a/youtube_dl/downloader/http.py +++ b/youtube_dl/downloader/http.py @@ -20,14 +20,10 @@ from ..utils import ( class HttpFD(FileDownloader): - def real_download(self, filename_or_stream, info_dict): + def real_download(self, filename, info_dict): url = info_dict['url'] - filename = filename_or_stream - stream = None - if hasattr(filename_or_stream, 'write'): - stream = filename_or_stream - filename = '-' tmpfilename = self.temp_name(filename) + stream = None # Do not include the Accept-Encoding header headers = {'Youtubedl-no-compression': 'True'}