2015-06-04 00:10:18 +09:00
|
|
|
from __future__ import unicode_literals
|
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
import os
|
2015-06-04 00:10:18 +09:00
|
|
|
import re
|
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
from .fragment import FragmentFD
|
2016-03-19 23:42:23 +09:00
|
|
|
from ..compat import compat_urllib_error
|
2016-02-10 01:25:02 +09:00
|
|
|
from ..utils import (
|
|
|
|
sanitize_open,
|
|
|
|
encodeFilename,
|
|
|
|
)
|
2015-06-04 23:12:05 +09:00
|
|
|
|
2015-06-04 00:10:18 +09:00
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
class DashSegmentsFD(FragmentFD):
|
2015-06-04 00:10:18 +09:00
|
|
|
"""
|
|
|
|
Download segments in a DASH manifest
|
|
|
|
"""
|
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
FD_NAME = 'dashsegments'
|
2015-06-10 15:45:54 +09:00
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
def real_download(self, filename, info_dict):
|
|
|
|
base_url = info_dict['url']
|
|
|
|
segment_urls = [info_dict['segment_urls'][0]] if self.params.get('test', False) else info_dict['segment_urls']
|
|
|
|
initialization_url = info_dict.get('initialization_url')
|
2015-06-10 15:45:54 +09:00
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
ctx = {
|
|
|
|
'filename': filename,
|
|
|
|
'total_frags': len(segment_urls) + (1 if initialization_url else 0),
|
|
|
|
}
|
2015-06-10 15:45:54 +09:00
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
self._prepare_and_start_frag_download(ctx)
|
2015-06-04 00:10:18 +09:00
|
|
|
|
|
|
|
def combine_url(base_url, target_url):
|
|
|
|
if re.match(r'^https?://', target_url):
|
|
|
|
return target_url
|
2015-07-23 19:09:30 +09:00
|
|
|
return '%s%s%s' % (base_url, '' if base_url.endswith('/') else '/', target_url)
|
2015-06-04 00:10:18 +09:00
|
|
|
|
2016-02-10 01:25:02 +09:00
|
|
|
segments_filenames = []
|
2016-02-14 17:13:09 +09:00
|
|
|
|
2016-03-19 23:42:23 +09:00
|
|
|
fragment_retries = self.params.get('fragment_retries', 0)
|
|
|
|
|
|
|
|
def append_url_to_file(target_url, tmp_filename, segment_name):
|
|
|
|
target_filename = '%s-%s' % (tmp_filename, segment_name)
|
|
|
|
count = 0
|
|
|
|
while count <= fragment_retries:
|
|
|
|
try:
|
|
|
|
success = ctx['dl'].download(target_filename, {'url': combine_url(base_url, target_url)})
|
|
|
|
if not success:
|
|
|
|
return False
|
|
|
|
down, target_sanitized = sanitize_open(target_filename, 'rb')
|
|
|
|
ctx['dest_stream'].write(down.read())
|
|
|
|
down.close()
|
|
|
|
segments_filenames.append(target_sanitized)
|
|
|
|
break
|
|
|
|
except (compat_urllib_error.HTTPError, ) as err:
|
|
|
|
# YouTube may often return 404 HTTP error for a fragment causing the
|
|
|
|
# whole download to fail. However if the same fragment is immediately
|
|
|
|
# retried with the same request data this usually succeeds (1-2 attemps
|
|
|
|
# is usually enough) thus allowing to download the whole file successfully.
|
|
|
|
# So, we will retry all fragments that fail with 404 HTTP error for now.
|
|
|
|
if err.code != 404:
|
|
|
|
raise
|
|
|
|
# Retry fragment
|
|
|
|
count += 1
|
|
|
|
if count <= fragment_retries:
|
|
|
|
self.report_retry_fragment(segment_name, count, fragment_retries)
|
|
|
|
if count > fragment_retries:
|
|
|
|
self.report_error('giving up after %s fragment retries' % fragment_retries)
|
2016-02-10 01:25:02 +09:00
|
|
|
return False
|
|
|
|
|
|
|
|
if initialization_url:
|
2016-03-19 23:42:23 +09:00
|
|
|
append_url_to_file(initialization_url, ctx['tmpfilename'], 'Init')
|
2016-02-10 01:25:02 +09:00
|
|
|
for i, segment_url in enumerate(segment_urls):
|
2016-03-19 23:42:23 +09:00
|
|
|
append_url_to_file(segment_url, ctx['tmpfilename'], 'Seg%d' % i)
|
2016-02-10 01:25:02 +09:00
|
|
|
|
|
|
|
self._finish_frag_download(ctx)
|
|
|
|
|
|
|
|
for segment_file in segments_filenames:
|
|
|
|
os.remove(encodeFilename(segment_file))
|
2015-06-04 00:10:18 +09:00
|
|
|
|
|
|
|
return True
|