mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2024-08-23 07:40:57 +02:00
* [scrippsnetworks] Add new extractor(closes #19857)(closes #22981) * [teachable] Improve locked lessons detection (#23528) * [teachable] Fail with error message if no video URL found * [extractors] add missing import for ScrippsNetworksIE * [brightcove] cache brightcove player policy keys * [prosiebensat1] improve geo restriction handling(closes #23571) * [soundcloud] automatically update client id on failing requests * [spankbang] Fix extraction (closes #23307, closes #23423, closes #23444) * [spankbang] Improve removed video detection (#23423) * [brightcove] update policy key on failing requests * [pornhub] Fix extraction and add support for m3u8 formats (closes #22749, closes #23082) * [pornhub] Improve locked videos detection (closes #22449, closes #22780) * [brightcove] invalidate policy key cache on failing requests * [soundcloud] fix client id extraction for non fatal requests * [ChangeLog] Actualize [ci skip] * [devscripts/create-github-release] Switch to using PAT for authentication Basic authentication will be deprecated soon * release 2020.01.01 * [redtube] Detect private videos (#23518) * [vice] improve extraction(closes #23631) * [devscripts/create-github-release] Remove unused import * [wistia] improve format extraction and extract subtitles(closes #22590) * [nrktv:seriebase] Fix extraction (closes #23625) (#23537) * [discovery] fix anonymous token extraction(closes #23650) * [scrippsnetworks] add support for www.discovery.com videos * [scrippsnetworks] correct test case URL * [dctp] fix format extraction(closes #23656) * [pandatv] Remove extractor (#23630) * [naver] improve extraction - improve geo-restriction handling - extract automatic captions - extract uploader metadata - extract VLive HLS formats * [naver] improve metadata extraction * [cloudflarestream] improve extraction - add support for bytehighway.net domain - add support for signed URLs - extract thumbnail * [cloudflarestream] import embed URL extraction * [lego] fix extraction and extract subtitle(closes #23687) * [safari] Fix kaltura session extraction (closes #23679) (#23670) * [orf:fm4] Fix extraction (#23599) * [orf:radio] Clean description and improve extraction * [twitter] add support for promo_video_website cards(closes #23711) * [vodplatform] add support for embed.kwikmotion.com domain * [ndr:base:embed] Improve thumbnails extraction (closes #23731) * [canvas] Add support for new API endpoint and update tests (closes #17680, closes #18629) * [travis] Add flake8 job (#23720) * [yourporn] Fix extraction (closes #21645, closes #22255, closes #23459) * [ChangeLog] Actualize [ci skip] * release 2020.01.15 * [soundcloud] Restore previews extraction (closes #23739) * [orf:tvthek] Improve geo restricted videos detection (closes #23741) * [zype] improve extraction - extract subtitles(closes #21258) - support URLs with alternative keys/tokens(#21258) - extract more metadata * [americastestkitchen] fix extraction * [nbc] add support for nbc multi network URLs(closes #23049) * [ard] improve extraction(closes #23761) - simplify extraction - extract age limit and series - bypass geo-restriction * [ivi:compilation] Fix entries extraction (closes #23770) * [24video] Add support for 24video.vip (closes #23753) * [businessinsider] Fix jwplatform id extraction (closes #22929) (#22954) * [ard] add a missing condition * [azmedien] fix extraction(closes #23783) * [voicerepublic] fix extraction * [stretchinternet] fix extraction(closes #4319) * [youtube] Fix sigfunc name extraction (closes #23819) * [ChangeLog] Actualize [ci skip] * release 2020.01.24 * [soundcloud] imporve private playlist/set tracks extraction https://github.com/ytdl-org/youtube-dl/issues/3707#issuecomment-577873539 * [svt] fix article extraction(closes #22897)(closes #22919) * [svt] fix series extraction(closes #22297) * [viewlift] improve extraction - fix extraction(closes #23851) - add add support for authentication - add support for more domains * [vimeo] fix album extraction(closes #23864) * [tva] Relax _VALID_URL (closes #23903) * [tv5mondeplus] Fix extraction (closes #23907, closes #23911) * [twitch:stream] Lowercase channel id for stream request (closes #23917) * [sportdeutschland] Update to new sportdeutschland API They switched to SSL, but under a different host AND path... Remove the old test cases because these videos have become unavailable. * [popcorntimes] Add extractor (closes #23949) * [thisoldhouse] fix extraction(closes #23951) * [toggle] Add support for mewatch.sg (closes #23895) (#23930) * [compat] Introduce compat_realpath (refs #23991) * [update] Fix updating via symlinks (closes #23991) * [nytimes] improve format sorting(closes #24010) * [abc:iview] Support 720p (#22907) (#22921) * [nova:embed] Fix extraction (closes #23672) * [nova:embed] Improve (closes #23690) * [nova] Improve extraction (refs #23690) * [jpopsuki] Remove extractor (closes #23858) * [YoutubeDL] Fix playlist entry indexing with --playlist-items (closes #10591, closes #10622) * [test_YoutubeDL] Fix get_ids * [test_YoutubeDL] Add tests for #10591 (closes #23873) * [24video] Add support for porn.24video.net (closes #23779, closes #23784) * [npr] Add support for streams (closes #24042) * [ChangeLog] Actualize [ci skip] * release 2020.02.16 * [tv2dk:bornholm:play] Fix extraction (#24076) * [imdb] Fix extraction (closes #23443) * [wistia] Add support for multiple generic embeds (closes #8347, closes #11385) * [teachable] Add support for multiple videos per lecture (closes #24101) * [pornhd] Fix extraction (closes #24128) * [options] Remove duplicate short option -v for --version (#24162) * [extractor/common] Convert ISM manifest to unicode before processing on python 2 (#24152) * [YoutubeDL] Force redirect URL to unicode on python 2 * Remove no longer needed compat_str around geturl * [youjizz] Fix extraction (closes #24181) * [test_subtitles] Remove obsolete test * [zdf:channel] Fix tests * [zapiks] Fix test * [xtube] Fix metadata extraction (closes #21073, closes #22455) * [xtube:user] Fix test * [telecinco] Fix extraction (refs #24195) * [telecinco] Add support for article opening videos * [franceculture] Fix extraction (closes #24204) * [xhamster] Fix extraction (closes #24205) * [ChangeLog] Actualize [ci skip] * release 2020.03.01 * [vimeo] Fix subtitles URLs (#24209) * [servus] Add support for new URL schema (closes #23475, closes #23583, closes #24142) * [youtube:playlist] Fix tests (closes #23872) (#23885) * [peertube] Improve extraction * [peertube] Fix issues and improve extraction (closes #23657) * [pornhub] Improve title extraction (closes #24184) * [vimeo] fix showcase password protected video extraction(closes #24224) * [youtube] Fix age-gated videos support without login (closes #24248) * [youtube] Fix tests * [ChangeLog] Actualize [ci skip] * release 2020.03.06 * [nhk] update API version(closes #24270) * [youtube] Improve extraction in 429 error conditions (closes #24283) * [youtube] Improve age-gated videos extraction in 429 error conditions (refs #24283) * [youtube] Remove outdated code Additional get_video_info requests don't seem to provide any extra itags any longer * [README.md] Clarify 429 error * [pornhub] Add support for pornhubpremium.com (#24288) * [utils] Add support for cookies with spaces used instead of tabs * [ChangeLog] Actualize [ci skip] * release 2020.03.08 * Revert "[utils] Add support for cookies with spaces used instead of tabs" According to [1] TABs must be used as separators between fields. Files produces by some tools with spaces as separators are considered malformed. 1. https://curl.haxx.se/docs/http-cookies.html This reverts commitcff99c91d1
. * [utils] Add reference to cookie file format * Revert "[vimeo] fix showcase password protected video extraction(closes #24224)" This reverts commit12ee431676
. * [nhk] Relax _VALID_URL (#24329) * [nhk] Remove obsolete rtmp formats (closes #24329) * [nhk] Update m3u8 URL and use native hls (#24329) * [ndr] Fix extraction (closes #24326) * [xtube] Fix formats extraction (closes #24348) * [xtube] Fix typo * [hellporno] Fix extraction (closes #24399) * [cbc:watch] Add support for authentication * [cbc:watch] Fix authenticated device token caching (closes #19160) * [soundcloud] fix download url extraction(closes #24394) * [limelight] remove disabled API requests(closes #24255) * [bilibili] Add support for new URL schema with BV ids (closes #24439, closes #24442) * [bilibili] Add support for player.bilibili.com (closes #24402) * [teachable] Extract chapter metadata (closes #24421) * [generic] Look for teachable embeds before wistia * [teachable] Update upskillcourses domain New version does not use teachable platform any longer * [teachable] Update gns3 domain * [teachable] Update test * [ChangeLog] Actualize [ci skip] * [ChangeLog] Actualize [ci skip] * release 2020.03.24 * [spankwire] Fix extraction (closes #18924, closes #20648) * [spankwire] Add support for generic embeds (refs #24633) * [youporn] Add support form generic embeds * [mofosex] Add support for generic embeds (closes #24633) * [tele5] Fix extraction (closes #24553) * [extractor/common] Skip malformed ISM manifest XMLs while extracting ISM formats (#24667) * [tv4] Fix ISM formats extraction (closes #24667) * [twitch:clips] Extend _VALID_URL (closes #24290) (#24642) * [motherless] Fix extraction (closes #24699) * [nova:embed] Fix extraction (closes #24700) * [youtube] Skip broken multifeed videos (closes #24711) * [soundcloud] Extract AAC format * [soundcloud] Improve AAC format extraction (closes #19173, closes #24708) * [thisoldhouse] Fix video id extraction (closes #24548) Added support for: with of without "www." and either ".chorus.build" or ".com" It now validated correctly on older URL's ``` <iframe src="https://thisoldhouse.chorus.build/videos/zype/5e33baec27d2e50001d5f52f ``` and newer ones ``` <iframe src="https://www.thisoldhouse.com/videos/zype/5e2b70e95216cc0001615120 ``` * [thisoldhouse] Improve video id extraction (closes #24549) * [youtube] Fix DRM videos detection (refs #24736) * [options] Clarify doc on --exec command (closes #19087) (#24883) * [prosiebensat1] Improve extraction and remove 7tv.de support (#24948) * [prosiebensat1] Extract series metadata * [tenplay] Relax _VALID_URL (closes #25001) * [tvplay] fix Viafree extraction(closes #15189)(closes #24473)(closes #24789) * [yahoo] fix GYAO Player extraction and relax title URL regex(closes #24178)(closes #24778) * [youtube] Use redirected video id if any (closes #25063) * [youtube] Improve player id extraction and add tests * [extractor/common] Extract multiple JSON-LD entries * [crunchyroll] Fix and improve extraction (closes #25096, closes #25060) * [ChangeLog] Actualize [ci skip] * release 2020.05.03 * [puhutv] Remove no longer available HTTP formats (closes #25124) * [utils] Improve cookie files support + Add support for UTF-8 in cookie files * Skip malformed cookie file entries instead of crashing (invalid entry len, invalid expires at) * [dailymotion] Fix typo * [compat] Introduce compat_cookiejar_Cookie * [extractor/common] Use compat_cookiejar_Cookie for _set_cookie (closes #23256, closes #24776) To always ensure cookie name and value are bytestrings on python 2. * [orf] Add support for more radio stations (closes #24938) (#24968) * [uol] fix extraction(closes #22007) * [downloader/http] Finish downloading once received data length matches expected Always do this if possible, i.e. if Content-Length or expected length is known, not only in test. This will save unnecessary last extra loop trying to read 0 bytes. * [downloader/http] Request last data block of exact remaining size Always request last data block of exact size remaining to download if possible not the current block size. * [iprima] Improve extraction (closes #25138) * [youtube] Improve signature cipher extraction (closes #25188) * [ChangeLog] Actualize [ci skip] * release 2020.05.08 * [spike] fix Bellator mgid extraction(closes #25195) * [bbccouk] PEP8 * [mailru] Fix extraction (closes #24530) (#25239) * [README.md] flake8 HTTPS URL (#25230) * [youtube] Add support for yewtu.be (#25226) * [soundcloud] reduce API playlist page limit(closes #25274) * [vimeo] improve format extraction and sorting(closes #25285) * [redtube] Improve title extraction (#25208) * [indavideo] Switch to HTTPS for API request (#25191) * [utils] Fix file permissions in write_json_file (closes #12471) (#25122) * [redtube] Improve formats extraction and extract m3u8 formats (closes #25311, closes #25321) * [ard] Improve _VALID_URL (closes #25134) (#25198) * [giantbomb] Extend _VALID_URL (#25222) * [postprocessor/ffmpeg] Embed series metadata with --add-metadata * [youtube] Add support for more invidious instances (#25417) * [ard:beta] Extend _VALID_URL (closes #25405) * [ChangeLog] Actualize [ci skip] * release 2020.05.29 * [jwplatform] Improve embeds extraction (closes #25467) * [periscope] Fix untitled broadcasts (#25482) * [twitter:broadcast] Add untitled periscope broadcast test * [malltv] Add support for sk.mall.tv (#25445) * [brightcove] Fix subtitles extraction (closes #25540) * [brightcove] Sort imports * [twitch] Pass v5 accept header and fix thumbnails extraction (closes #25531) * [twitch:stream] Fix extraction (closes #25528) * [twitch:stream] Expect 400 and 410 HTTP errors from API * [tele5] Prefer jwplatform over nexx (closes #25533) * [jwplatform] Add support for bypass geo restriction * [tele5] Bypass geo restriction * [ChangeLog] Actualize [ci skip] * release 2020.06.06 * [kaltura] Add support for multiple embeds on a webpage (closes #25523) * [youtube] Extract chapters from JSON (closes #24819) * [facebook] Support single-video ID links I stumbled upon this at https://www.facebook.com/bwfbadminton/posts/10157127020046316 . No idea how prevalent it is yet. * [youtube] Fix playlist and feed extraction (closes #25675) * [youtube] Fix thumbnails extraction and remove uploader id extraction warning (closes #25676) * [youtube] Fix upload date extraction * [youtube] Improve view count extraction * [youtube] Fix uploader id and uploader URL extraction * [ChangeLog] Actualize [ci skip] * release 2020.06.16 * [youtube] Fix categories and improve tags extraction * [youtube] Force old layout (closes #25682, closes #25683, closes #25680, closes #25686) * [ChangeLog] Actualize [ci skip] * release 2020.06.16.1 * [brightcove] Improve embed detection (closes #25674) * [bellmedia] add support for cp24.com clip URLs(closes #25764) * [youtube:playlists] Extend _VALID_URL (closes #25810) * [youtube] Prevent excess HTTP 301 (#25786) * [wistia] Restrict embed regex (closes #25969) * [youtube] Improve description extraction (closes #25937) (#25980) * [youtube] Fix sigfunc name extraction (closes #26134, closes #26135, closes #26136, closes #26137) * [ChangeLog] Actualize [ci skip] * release 2020.07.28 * [xhamster] Extend _VALID_URL (closes #25789) (#25804) * [xhamster] Fix extraction (closes #26157) (#26254) * [xhamster] Extend _VALID_URL (closes #25927) Co-authored-by: Remita Amine <remitamine@gmail.com> Co-authored-by: Sergey M․ <dstftw@gmail.com> Co-authored-by: nmeum <soeren+github@soeren-tempel.net> Co-authored-by: Roxedus <me@roxedus.dev> Co-authored-by: Singwai Chan <c.singwai@gmail.com> Co-authored-by: cdarlint <cdarlint@users.noreply.github.com> Co-authored-by: Johannes N <31795504+jonolt@users.noreply.github.com> Co-authored-by: jnozsc <jnozsc@gmail.com> Co-authored-by: Moritz Patelscheck <moritz.patelscheck@campus.tu-berlin.de> Co-authored-by: PB <3854688+uno20001@users.noreply.github.com> Co-authored-by: Philipp Hagemeister <phihag@phihag.de> Co-authored-by: Xaver Hellauer <software@hellauer.bayern> Co-authored-by: d2au <d2au.dev@gmail.com> Co-authored-by: Jan 'Yenda' Trmal <jtrmal@gmail.com> Co-authored-by: jxu <7989982+jxu@users.noreply.github.com> Co-authored-by: Martin Ström <name@my-domain.se> Co-authored-by: The Hatsune Daishi <nao20010128@gmail.com> Co-authored-by: tsia <github@tsia.de> Co-authored-by: 3risian <59593325+3risian@users.noreply.github.com> Co-authored-by: Tristan Waddington <tristan.waddington@gmail.com> Co-authored-by: Devon Meunier <devon.meunier@gmail.com> Co-authored-by: Felix Stupp <felix.stupp@outlook.com> Co-authored-by: tom <tomster954@gmail.com> Co-authored-by: AndrewMBL <62922222+AndrewMBL@users.noreply.github.com> Co-authored-by: willbeaufoy <will@willbeaufoy.net> Co-authored-by: Philipp Stehle <anderschwiedu@googlemail.com> Co-authored-by: hh0rva1h <61889859+hh0rva1h@users.noreply.github.com> Co-authored-by: comsomisha <shmelev1996@mail.ru> Co-authored-by: TotalCaesar659 <14265316+TotalCaesar659@users.noreply.github.com> Co-authored-by: Juan Francisco Cantero Hurtado <iam@juanfra.info> Co-authored-by: Dave Loyall <dave@the-good-guys.net> Co-authored-by: tlsssl <63866177+tlsssl@users.noreply.github.com> Co-authored-by: Rob <ankenyr@gmail.com> Co-authored-by: Michael Klein <github@a98shuttle.de> Co-authored-by: JordanWeatherby <47519158+JordanWeatherby@users.noreply.github.com> Co-authored-by: striker.sh <19488257+strikersh@users.noreply.github.com> Co-authored-by: Matej Dujava <mdujava@gmail.com> Co-authored-by: Glenn Slayden <5589855+glenn-slayden@users.noreply.github.com> Co-authored-by: MRWITEK <mrvvitek@gmail.com> Co-authored-by: JChris246 <43832407+JChris246@users.noreply.github.com> Co-authored-by: TheRealDude2 <the.real.dude@gmx.de>
355 lines
15 KiB
Python
355 lines
15 KiB
Python
from __future__ import unicode_literals
|
|
|
|
import errno
|
|
import os
|
|
import socket
|
|
import time
|
|
import random
|
|
import re
|
|
|
|
from .common import FileDownloader
|
|
from ..compat import (
|
|
compat_str,
|
|
compat_urllib_error,
|
|
)
|
|
from ..utils import (
|
|
ContentTooShortError,
|
|
encodeFilename,
|
|
int_or_none,
|
|
sanitize_open,
|
|
sanitized_Request,
|
|
write_xattr,
|
|
XAttrMetadataError,
|
|
XAttrUnavailableError,
|
|
)
|
|
|
|
|
|
class HttpFD(FileDownloader):
|
|
def real_download(self, filename, info_dict):
|
|
url = info_dict['url']
|
|
|
|
class DownloadContext(dict):
|
|
__getattr__ = dict.get
|
|
__setattr__ = dict.__setitem__
|
|
__delattr__ = dict.__delitem__
|
|
|
|
ctx = DownloadContext()
|
|
ctx.filename = filename
|
|
ctx.tmpfilename = self.temp_name(filename)
|
|
ctx.stream = None
|
|
|
|
# Do not include the Accept-Encoding header
|
|
headers = {'Youtubedl-no-compression': 'True'}
|
|
add_headers = info_dict.get('http_headers')
|
|
if add_headers:
|
|
headers.update(add_headers)
|
|
|
|
is_test = self.params.get('test', False)
|
|
chunk_size = self._TEST_FILE_SIZE if is_test else (
|
|
info_dict.get('downloader_options', {}).get('http_chunk_size')
|
|
or self.params.get('http_chunk_size') or 0)
|
|
|
|
ctx.open_mode = 'wb'
|
|
ctx.resume_len = 0
|
|
ctx.data_len = None
|
|
ctx.block_size = self.params.get('buffersize', 1024)
|
|
ctx.start_time = time.time()
|
|
ctx.chunk_size = None
|
|
|
|
if self.params.get('continuedl', True):
|
|
# Establish possible resume length
|
|
if os.path.isfile(encodeFilename(ctx.tmpfilename)):
|
|
ctx.resume_len = os.path.getsize(
|
|
encodeFilename(ctx.tmpfilename))
|
|
|
|
ctx.is_resume = ctx.resume_len > 0
|
|
|
|
count = 0
|
|
retries = self.params.get('retries', 0)
|
|
|
|
class SucceedDownload(Exception):
|
|
pass
|
|
|
|
class RetryDownload(Exception):
|
|
def __init__(self, source_error):
|
|
self.source_error = source_error
|
|
|
|
class NextFragment(Exception):
|
|
pass
|
|
|
|
def set_range(req, start, end):
|
|
range_header = 'bytes=%d-' % start
|
|
if end:
|
|
range_header += compat_str(end)
|
|
req.add_header('Range', range_header)
|
|
|
|
def establish_connection():
|
|
ctx.chunk_size = (random.randint(int(chunk_size * 0.95), chunk_size)
|
|
if not is_test and chunk_size else chunk_size)
|
|
if ctx.resume_len > 0:
|
|
range_start = ctx.resume_len
|
|
if ctx.is_resume:
|
|
self.report_resuming_byte(ctx.resume_len)
|
|
ctx.open_mode = 'ab'
|
|
elif ctx.chunk_size > 0:
|
|
range_start = 0
|
|
else:
|
|
range_start = None
|
|
ctx.is_resume = False
|
|
range_end = range_start + ctx.chunk_size - 1 if ctx.chunk_size else None
|
|
if range_end and ctx.data_len is not None and range_end >= ctx.data_len:
|
|
range_end = ctx.data_len - 1
|
|
has_range = range_start is not None
|
|
ctx.has_range = has_range
|
|
request = sanitized_Request(url, None, headers)
|
|
if has_range:
|
|
set_range(request, range_start, range_end)
|
|
# Establish connection
|
|
try:
|
|
ctx.data = self.ydl.urlopen(request)
|
|
# When trying to resume, Content-Range HTTP header of response has to be checked
|
|
# to match the value of requested Range HTTP header. This is due to a webservers
|
|
# that don't support resuming and serve a whole file with no Content-Range
|
|
# set in response despite of requested Range (see
|
|
# https://github.com/ytdl-org/youtube-dl/issues/6057#issuecomment-126129799)
|
|
if has_range:
|
|
content_range = ctx.data.headers.get('Content-Range')
|
|
if content_range:
|
|
content_range_m = re.search(r'bytes (\d+)-(\d+)?(?:/(\d+))?', content_range)
|
|
# Content-Range is present and matches requested Range, resume is possible
|
|
if content_range_m:
|
|
if range_start == int(content_range_m.group(1)):
|
|
content_range_end = int_or_none(content_range_m.group(2))
|
|
content_len = int_or_none(content_range_m.group(3))
|
|
accept_content_len = (
|
|
# Non-chunked download
|
|
not ctx.chunk_size
|
|
# Chunked download and requested piece or
|
|
# its part is promised to be served
|
|
or content_range_end == range_end
|
|
or content_len < range_end)
|
|
if accept_content_len:
|
|
ctx.data_len = content_len
|
|
return
|
|
# Content-Range is either not present or invalid. Assuming remote webserver is
|
|
# trying to send the whole file, resume is not possible, so wiping the local file
|
|
# and performing entire redownload
|
|
self.report_unable_to_resume()
|
|
ctx.resume_len = 0
|
|
ctx.open_mode = 'wb'
|
|
ctx.data_len = int_or_none(ctx.data.info().get('Content-length', None))
|
|
return
|
|
except (compat_urllib_error.HTTPError, ) as err:
|
|
if err.code == 416:
|
|
# Unable to resume (requested range not satisfiable)
|
|
try:
|
|
# Open the connection again without the range header
|
|
ctx.data = self.ydl.urlopen(
|
|
sanitized_Request(url, None, headers))
|
|
content_length = ctx.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 (ctx.resume_len - 100 < int(content_length) < ctx.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(ctx.filename)
|
|
self.try_rename(ctx.tmpfilename, ctx.filename)
|
|
self._hook_progress({
|
|
'filename': ctx.filename,
|
|
'status': 'finished',
|
|
'downloaded_bytes': ctx.resume_len,
|
|
'total_bytes': ctx.resume_len,
|
|
})
|
|
raise SucceedDownload()
|
|
else:
|
|
# The length does not match, we start the download over
|
|
self.report_unable_to_resume()
|
|
ctx.resume_len = 0
|
|
ctx.open_mode = 'wb'
|
|
return
|
|
elif err.code < 500 or err.code >= 600:
|
|
# Unexpected HTTP error
|
|
raise
|
|
raise RetryDownload(err)
|
|
except socket.error as err:
|
|
if err.errno != errno.ECONNRESET:
|
|
# Connection reset is no problem, just retry
|
|
raise
|
|
raise RetryDownload(err)
|
|
|
|
def download():
|
|
data_len = ctx.data.info().get('Content-length', None)
|
|
|
|
# Range HTTP header may be ignored/unsupported by a webserver
|
|
# (e.g. extractor/scivee.py, extractor/bambuser.py).
|
|
# However, for a test we still would like to download just a piece of a file.
|
|
# To achieve this we limit data_len to _TEST_FILE_SIZE and manually control
|
|
# block size when downloading a file.
|
|
if is_test and (data_len is None or int(data_len) > self._TEST_FILE_SIZE):
|
|
data_len = self._TEST_FILE_SIZE
|
|
|
|
if data_len is not None:
|
|
data_len = int(data_len) + ctx.resume_len
|
|
min_data_len = self.params.get('min_filesize')
|
|
max_data_len = self.params.get('max_filesize')
|
|
if min_data_len is not None and data_len < min_data_len:
|
|
self.to_screen('\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('\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
|
|
return False
|
|
|
|
byte_counter = 0 + ctx.resume_len
|
|
block_size = ctx.block_size
|
|
start = time.time()
|
|
|
|
# measure time over whole while-loop, so slow_down() and best_block_size() work together properly
|
|
now = None # needed for slow_down() in the first loop run
|
|
before = start # start measuring
|
|
|
|
def retry(e):
|
|
to_stdout = ctx.tmpfilename == '-'
|
|
if not to_stdout:
|
|
ctx.stream.close()
|
|
ctx.stream = None
|
|
ctx.resume_len = byte_counter if to_stdout else os.path.getsize(encodeFilename(ctx.tmpfilename))
|
|
raise RetryDownload(e)
|
|
|
|
while True:
|
|
try:
|
|
# Download and write
|
|
data_block = ctx.data.read(block_size if data_len is None else min(block_size, data_len - byte_counter))
|
|
# socket.timeout is a subclass of socket.error but may not have
|
|
# errno set
|
|
except socket.timeout as e:
|
|
retry(e)
|
|
except socket.error as e:
|
|
if e.errno not in (errno.ECONNRESET, errno.ETIMEDOUT):
|
|
raise
|
|
retry(e)
|
|
|
|
byte_counter += len(data_block)
|
|
|
|
# exit loop when download is finished
|
|
if len(data_block) == 0:
|
|
break
|
|
|
|
# Open destination file just in time
|
|
if ctx.stream is None:
|
|
try:
|
|
ctx.stream, ctx.tmpfilename = sanitize_open(
|
|
ctx.tmpfilename, ctx.open_mode)
|
|
assert ctx.stream is not None
|
|
ctx.filename = self.undo_temp_name(ctx.tmpfilename)
|
|
self.report_destination(ctx.filename)
|
|
except (OSError, IOError) as err:
|
|
self.report_error('unable to open for writing: %s' % str(err))
|
|
return False
|
|
|
|
if self.params.get('xattr_set_filesize', False) and data_len is not None:
|
|
try:
|
|
write_xattr(ctx.tmpfilename, 'user.ytdl.filesize', str(data_len).encode('utf-8'))
|
|
except (XAttrUnavailableError, XAttrMetadataError) as err:
|
|
self.report_error('unable to set filesize xattr: %s' % str(err))
|
|
|
|
try:
|
|
ctx.stream.write(data_block)
|
|
except (IOError, OSError) as err:
|
|
self.to_stderr('\n')
|
|
self.report_error('unable to write data: %s' % str(err))
|
|
return False
|
|
|
|
# Apply rate limit
|
|
self.slow_down(start, now, byte_counter - ctx.resume_len)
|
|
|
|
# end measuring of one loop run
|
|
now = time.time()
|
|
after = now
|
|
|
|
# Adjust block size
|
|
if not self.params.get('noresizebuffer', False):
|
|
block_size = self.best_block_size(after - before, len(data_block))
|
|
|
|
before = after
|
|
|
|
# Progress message
|
|
speed = self.calc_speed(start, now, byte_counter - ctx.resume_len)
|
|
if ctx.data_len is None:
|
|
eta = None
|
|
else:
|
|
eta = self.calc_eta(start, time.time(), ctx.data_len - ctx.resume_len, byte_counter - ctx.resume_len)
|
|
|
|
self._hook_progress({
|
|
'status': 'downloading',
|
|
'downloaded_bytes': byte_counter,
|
|
'total_bytes': ctx.data_len,
|
|
'tmpfilename': ctx.tmpfilename,
|
|
'filename': ctx.filename,
|
|
'eta': eta,
|
|
'speed': speed,
|
|
'elapsed': now - ctx.start_time,
|
|
})
|
|
|
|
if data_len is not None and byte_counter == data_len:
|
|
break
|
|
|
|
if not is_test and ctx.chunk_size and ctx.data_len is not None and byte_counter < ctx.data_len:
|
|
ctx.resume_len = byte_counter
|
|
# ctx.block_size = block_size
|
|
raise NextFragment()
|
|
|
|
if ctx.stream is None:
|
|
self.to_stderr('\n')
|
|
self.report_error('Did not get any data blocks')
|
|
return False
|
|
if ctx.tmpfilename != '-':
|
|
ctx.stream.close()
|
|
|
|
if data_len is not None and byte_counter != data_len:
|
|
err = ContentTooShortError(byte_counter, int(data_len))
|
|
if count <= retries:
|
|
retry(err)
|
|
raise err
|
|
|
|
self.try_rename(ctx.tmpfilename, ctx.filename)
|
|
|
|
# Update file modification time
|
|
if self.params.get('updatetime', True):
|
|
info_dict['filetime'] = self.try_utime(ctx.filename, ctx.data.info().get('last-modified', None))
|
|
|
|
self._hook_progress({
|
|
'downloaded_bytes': byte_counter,
|
|
'total_bytes': byte_counter,
|
|
'filename': ctx.filename,
|
|
'status': 'finished',
|
|
'elapsed': time.time() - ctx.start_time,
|
|
})
|
|
|
|
return True
|
|
|
|
while count <= retries:
|
|
try:
|
|
establish_connection()
|
|
return download()
|
|
except RetryDownload as e:
|
|
count += 1
|
|
if count <= retries:
|
|
self.report_retry(e.source_error, count, retries)
|
|
continue
|
|
except NextFragment:
|
|
continue
|
|
except SucceedDownload:
|
|
return True
|
|
|
|
self.report_error('giving up after %s retries' % retries)
|
|
return False
|