0
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-30 03:33:02 +00:00
yt-dlp/yt_dlp/downloader/fragment.py
Simon Sawicki 1c51c520f7
[fd/fragment] Improve progress calculation (#8241)
This uses the download speed from all threads and also adds smoothing to speed and eta

Authored by: Grub4K
2023-10-08 02:01:01 +02:00

527 lines
21 KiB
Python

import concurrent.futures
import contextlib
import json
import math
import os
import struct
import time
from .common import FileDownloader
from .http import HttpFD
from ..aes import aes_cbc_decrypt_bytes, unpad_pkcs7
from ..compat import compat_os_name
from ..networking import Request
from ..networking.exceptions import HTTPError, IncompleteRead
from ..utils import DownloadError, RetryManager, encodeFilename, traverse_obj
from ..utils.networking import HTTPHeaderDict
from ..utils.progress import ProgressCalculator
class HttpQuietDownloader(HttpFD):
def to_screen(self, *args, **kargs):
pass
to_console_title = to_screen
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). Default is 0 for API, but 10 for CLI
skip_unavailable_fragments:
Skip unavailable fragments (DASH and hlsnative only)
keep_fragments: Keep downloaded fragments on disk after downloading is
finished
concurrent_fragment_downloads: The number of threads to use for native hls and dash downloads
_no_ytdl_file: Don't use .ytdl file
For each incomplete fragment download yt-dlp 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 yt-dlp). 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.deprecation_warning('yt_dlp.downloader.FragmentFD.report_retry_fragment is deprecated. '
'Use yt_dlp.downloader.FileDownloader.report_retry instead')
return self.report_retry(err, count, retries, frag_index)
def report_skip_fragment(self, frag_index, err=None):
err = f' {err};' if err else ''
self.to_screen(f'[download]{err} Skipping fragment {frag_index:d} ...')
def _prepare_url(self, info_dict, url):
headers = info_dict.get('http_headers')
return Request(url, None, headers) if headers else url
def _prepare_and_start_frag_download(self, ctx, info_dict):
self._prepare_frag_download(ctx)
self._start_frag_download(ctx, info_dict)
def __do_ytdl_file(self, ctx):
return ctx['live'] is not True and ctx['tmpfilename'] != '-' and not self.params.get('_no_ytdl_file')
def _read_ytdl_file(self, ctx):
assert 'ytdl_corrupt' not in ctx
stream, _ = self.sanitize_open(self.ytdl_filename(ctx['filename']), 'r')
try:
ytdl_data = json.loads(stream.read())
ctx['fragment_index'] = ytdl_data['downloader']['current_fragment']['index']
if 'extra_state' in ytdl_data['downloader']:
ctx['extra_state'] = ytdl_data['downloader']['extra_state']
except Exception:
ctx['ytdl_corrupt'] = True
finally:
stream.close()
def _write_ytdl_file(self, ctx):
frag_index_stream, _ = self.sanitize_open(self.ytdl_filename(ctx['filename']), 'w')
try:
downloader = {
'current_fragment': {
'index': ctx['fragment_index'],
},
}
if 'extra_state' in ctx:
downloader['extra_state'] = ctx['extra_state']
if ctx.get('fragment_count') is not None:
downloader['fragment_count'] = ctx['fragment_count']
frag_index_stream.write(json.dumps({'downloader': downloader}))
finally:
frag_index_stream.close()
def _download_fragment(self, ctx, frag_url, info_dict, headers=None, request_data=None):
fragment_filename = '%s-Frag%d' % (ctx['tmpfilename'], ctx['fragment_index'])
fragment_info_dict = {
'url': frag_url,
'http_headers': headers or info_dict.get('http_headers'),
'request_data': request_data,
'ctx_id': ctx.get('ctx_id'),
}
frag_resume_len = 0
if ctx['dl'].params.get('continuedl', True):
frag_resume_len = self.filesize_or_none(self.temp_name(fragment_filename))
fragment_info_dict['frag_resume_len'] = ctx['frag_resume_len'] = frag_resume_len
success, _ = ctx['dl'].download(fragment_filename, fragment_info_dict)
if not success:
return False
if fragment_info_dict.get('filetime'):
ctx['fragment_filetime'] = fragment_info_dict.get('filetime')
ctx['fragment_filename_sanitized'] = fragment_filename
return True
def _read_fragment(self, ctx):
if not ctx.get('fragment_filename_sanitized'):
return None
try:
down, frag_sanitized = self.sanitize_open(ctx['fragment_filename_sanitized'], 'rb')
except FileNotFoundError:
if ctx.get('live'):
return None
raise
ctx['fragment_filename_sanitized'] = frag_sanitized
frag_content = down.read()
down.close()
return 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):
self.try_remove(encodeFilename(ctx['fragment_filename_sanitized']))
del ctx['fragment_filename_sanitized']
def _prepare_frag_download(self, ctx):
if not ctx.setdefault('live', False):
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(f'[{self.FD_NAME}] Total fragments: {total_frags_str}')
self.report_destination(ctx['filename'])
dl = HttpQuietDownloader(self.ydl, {
**self.params,
'noprogress': True,
'test': False,
'sleep_interval': 0,
'max_sleep_interval': 0,
'sleep_interval_subtitles': 0,
})
tmpfilename = self.temp_name(ctx['filename'])
open_mode = 'wb'
# Establish possible resume length
resume_len = self.filesize_or_none(tmpfilename)
if resume_len > 0:
open_mode = 'ab'
# Should be initialized before ytdl file check
ctx.update({
'tmpfilename': tmpfilename,
'fragment_index': 0,
})
if self.__do_ytdl_file(ctx):
ytdl_file_exists = os.path.isfile(encodeFilename(self.ytdl_filename(ctx['filename'])))
continuedl = self.params.get('continuedl', True)
if continuedl and ytdl_file_exists:
self._read_ytdl_file(ctx)
is_corrupt = ctx.get('ytdl_corrupt') is True
is_inconsistent = ctx['fragment_index'] > 0 and resume_len == 0
if is_corrupt or is_inconsistent:
message = (
'.ytdl file is corrupt' if is_corrupt else
'Inconsistent state of incomplete fragment download')
self.report_warning(
'%s. Restarting from the beginning ...' % message)
ctx['fragment_index'] = resume_len = 0
if 'ytdl_corrupt' in ctx:
del ctx['ytdl_corrupt']
self._write_ytdl_file(ctx)
else:
if not continuedl:
if ytdl_file_exists:
self._read_ytdl_file(ctx)
ctx['fragment_index'] = resume_len = 0
self._write_ytdl_file(ctx)
assert ctx['fragment_index'] == 0
dest_stream, tmpfilename = self.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, info_dict):
resume_len = ctx['complete_frags_downloaded_bytes']
total_frags = ctx['total_frags']
ctx_id = ctx.get('ctx_id')
# Stores the download progress, updated by the progress hook
state = {
'status': 'downloading',
'downloaded_bytes': resume_len,
'fragment_index': ctx['fragment_index'],
'fragment_count': total_frags,
'filename': ctx['filename'],
'tmpfilename': ctx['tmpfilename'],
}
ctx['started'] = time.time()
progress = ProgressCalculator(resume_len)
def frag_progress_hook(s):
if s['status'] not in ('downloading', 'finished'):
return
if not total_frags and ctx.get('fragment_count'):
state['fragment_count'] = ctx['fragment_count']
if ctx_id is not None and s.get('ctx_id') != ctx_id:
return
state['max_progress'] = ctx.get('max_progress')
state['progress_idx'] = ctx.get('progress_idx')
state['elapsed'] = progress.elapsed
frag_total_bytes = s.get('total_bytes') or 0
s['fragment_info_dict'] = s.pop('info_dict', {})
# XXX: Fragment resume is not accounted for here
if not ctx['live']:
estimated_size = (
(ctx['complete_frags_downloaded_bytes'] + frag_total_bytes)
/ (state['fragment_index'] + 1) * total_frags)
progress.total = estimated_size
progress.update(s.get('downloaded_bytes'))
state['total_bytes_estimate'] = progress.total
else:
progress.update(s.get('downloaded_bytes'))
if s['status'] == 'finished':
state['fragment_index'] += 1
ctx['fragment_index'] = state['fragment_index']
progress.thread_reset()
state['downloaded_bytes'] = ctx['complete_frags_downloaded_bytes'] = progress.downloaded
state['speed'] = ctx['speed'] = progress.speed.smooth
state['eta'] = progress.eta.smooth
self._hook_progress(state, info_dict)
ctx['dl'].add_progress_hook(frag_progress_hook)
return ctx['started']
def _finish_frag_download(self, ctx, info_dict):
ctx['dest_stream'].close()
if self.__do_ytdl_file(ctx):
self.try_remove(self.ytdl_filename(ctx['filename']))
elapsed = time.time() - ctx['started']
to_file = ctx['tmpfilename'] != '-'
if to_file:
downloaded_bytes = self.filesize_or_none(ctx['tmpfilename'])
else:
downloaded_bytes = ctx['complete_frags_downloaded_bytes']
if not downloaded_bytes:
if to_file:
self.try_remove(ctx['tmpfilename'])
self.report_error('The downloaded file is empty')
return False
elif to_file:
self.try_rename(ctx['tmpfilename'], ctx['filename'])
filetime = ctx.get('fragment_filetime')
if self.params.get('updatetime', True) and filetime:
with contextlib.suppress(Exception):
os.utime(ctx['filename'], (time.time(), filetime))
self._hook_progress({
'downloaded_bytes': downloaded_bytes,
'total_bytes': downloaded_bytes,
'filename': ctx['filename'],
'status': 'finished',
'elapsed': elapsed,
'ctx_id': ctx.get('ctx_id'),
'max_progress': ctx.get('max_progress'),
'progress_idx': ctx.get('progress_idx'),
}, info_dict)
return True
def _prepare_external_frag_download(self, ctx):
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(f'[{self.FD_NAME}] Total fragments: {total_frags_str}')
tmpfilename = self.temp_name(ctx['filename'])
# Should be initialized before ytdl file check
ctx.update({
'tmpfilename': tmpfilename,
'fragment_index': 0,
})
def decrypter(self, info_dict):
_key_cache = {}
def _get_key(url):
if url not in _key_cache:
_key_cache[url] = self.ydl.urlopen(self._prepare_url(info_dict, url)).read()
return _key_cache[url]
def decrypt_fragment(fragment, frag_content):
if frag_content is None:
return
decrypt_info = fragment.get('decrypt_info')
if not decrypt_info or decrypt_info['METHOD'] != 'AES-128':
return frag_content
iv = decrypt_info.get('IV') or struct.pack('>8xq', fragment['media_sequence'])
decrypt_info['KEY'] = (decrypt_info.get('KEY')
or _get_key(traverse_obj(info_dict, ('hls_aes', 'uri')) or decrypt_info['URI']))
# Don't decrypt the content in tests since the data is explicitly truncated and it's not to a valid block
# size (see https://github.com/ytdl-org/youtube-dl/pull/27660). Tests only care that the correct data downloaded,
# not what it decrypts to.
if self.params.get('test', False):
return frag_content
return unpad_pkcs7(aes_cbc_decrypt_bytes(frag_content, decrypt_info['KEY'], iv))
return decrypt_fragment
def download_and_append_fragments_multiple(self, *args, **kwargs):
'''
@params (ctx1, fragments1, info_dict1), (ctx2, fragments2, info_dict2), ...
all args must be either tuple or list
'''
interrupt_trigger = [True]
max_progress = len(args)
if max_progress == 1:
return self.download_and_append_fragments(*args[0], **kwargs)
max_workers = self.params.get('concurrent_fragment_downloads', 1)
if max_progress > 1:
self._prepare_multiline_status(max_progress)
is_live = any(traverse_obj(args, (..., 2, 'is_live')))
def thread_func(idx, ctx, fragments, info_dict, tpe):
ctx['max_progress'] = max_progress
ctx['progress_idx'] = idx
return self.download_and_append_fragments(
ctx, fragments, info_dict, **kwargs, tpe=tpe, interrupt_trigger=interrupt_trigger)
class FTPE(concurrent.futures.ThreadPoolExecutor):
# has to stop this or it's going to wait on the worker thread itself
def __exit__(self, exc_type, exc_val, exc_tb):
pass
if compat_os_name == 'nt':
def future_result(future):
while True:
try:
return future.result(0.1)
except KeyboardInterrupt:
raise
except concurrent.futures.TimeoutError:
continue
else:
def future_result(future):
return future.result()
def interrupt_trigger_iter(fg):
for f in fg:
if not interrupt_trigger[0]:
break
yield f
spins = []
for idx, (ctx, fragments, info_dict) in enumerate(args):
tpe = FTPE(math.ceil(max_workers / max_progress))
job = tpe.submit(thread_func, idx, ctx, interrupt_trigger_iter(fragments), info_dict, tpe)
spins.append((tpe, job))
result = True
for tpe, job in spins:
try:
result = result and future_result(job)
except KeyboardInterrupt:
interrupt_trigger[0] = False
finally:
tpe.shutdown(wait=True)
if not interrupt_trigger[0] and not is_live:
raise KeyboardInterrupt()
# we expect the user wants to stop and DO WANT the preceding postprocessors to run;
# so returning a intermediate result here instead of KeyboardInterrupt on live
return result
def download_and_append_fragments(
self, ctx, fragments, info_dict, *, is_fatal=(lambda idx: False),
pack_func=(lambda content, idx: content), finish_func=None,
tpe=None, interrupt_trigger=(True, )):
if not self.params.get('skip_unavailable_fragments', True):
is_fatal = lambda _: True
def download_fragment(fragment, ctx):
if not interrupt_trigger[0]:
return
frag_index = ctx['fragment_index'] = fragment['frag_index']
ctx['last_error'] = None
headers = HTTPHeaderDict(info_dict.get('http_headers'))
byte_range = fragment.get('byte_range')
if byte_range:
headers['Range'] = 'bytes=%d-%d' % (byte_range['start'], byte_range['end'] - 1)
# Never skip the first fragment
fatal = is_fatal(fragment.get('index') or (frag_index - 1))
def error_callback(err, count, retries):
if fatal and count > retries:
ctx['dest_stream'].close()
self.report_retry(err, count, retries, frag_index, fatal)
ctx['last_error'] = err
for retry in RetryManager(self.params.get('fragment_retries'), error_callback):
try:
ctx['fragment_count'] = fragment.get('fragment_count')
if not self._download_fragment(
ctx, fragment['url'], info_dict, headers, info_dict.get('request_data')):
return
except (HTTPError, IncompleteRead) as err:
retry.error = err
continue
except DownloadError: # has own retry settings
if fatal:
raise
def append_fragment(frag_content, frag_index, ctx):
if frag_content:
self._append_fragment(ctx, pack_func(frag_content, frag_index))
elif not is_fatal(frag_index - 1):
self.report_skip_fragment(frag_index, 'fragment not found')
else:
ctx['dest_stream'].close()
self.report_error(f'fragment {frag_index} not found, unable to continue')
return False
return True
decrypt_fragment = self.decrypter(info_dict)
max_workers = math.ceil(
self.params.get('concurrent_fragment_downloads', 1) / ctx.get('max_progress', 1))
if max_workers > 1:
def _download_fragment(fragment):
ctx_copy = ctx.copy()
download_fragment(fragment, ctx_copy)
return fragment, fragment['frag_index'], ctx_copy.get('fragment_filename_sanitized')
with tpe or concurrent.futures.ThreadPoolExecutor(max_workers) as pool:
try:
for fragment, frag_index, frag_filename in pool.map(_download_fragment, fragments):
ctx.update({
'fragment_filename_sanitized': frag_filename,
'fragment_index': frag_index,
})
if not append_fragment(decrypt_fragment(fragment, self._read_fragment(ctx)), frag_index, ctx):
return False
except KeyboardInterrupt:
self._finish_multiline_status()
self.report_error(
'Interrupted by user. Waiting for all threads to shutdown...', is_error=False, tb=False)
pool.shutdown(wait=False)
raise
else:
for fragment in fragments:
if not interrupt_trigger[0]:
break
try:
download_fragment(fragment, ctx)
result = append_fragment(
decrypt_fragment(fragment, self._read_fragment(ctx)), fragment['frag_index'], ctx)
except KeyboardInterrupt:
if info_dict.get('is_live'):
break
raise
if not result:
return False
if finish_func is not None:
ctx['dest_stream'].write(finish_func())
ctx['dest_stream'].flush()
return self._finish_frag_download(ctx, info_dict)