aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorJaime Marquínez Ferrándiz <jaime.marquinez.ferrandiz@gmail.com>2015-07-20 19:34:24 +0200
committerJaime Marquínez Ferrándiz <jaime.marquinez.ferrandiz@gmail.com>2015-07-20 19:34:24 +0200
commit2ee8f5d80f8eb80a28c0012b32f1abff9b633c1c (patch)
treebb4448a864d358581eb1a50a2d6943de8a18dcb6
parent3f302bca8cf29f618d1f1c2fb7bc4ea9538af805 (diff)
parent8a1a26ce4c64d7a2c142718fc56f46d9a1c2c4f2 (diff)
Merge remote-tracking branch 'yan12125/download-dash-segments' (#5886)
-rw-r--r--youtube_dl/downloader/__init__.py2
-rw-r--r--youtube_dl/downloader/dash.py66
-rw-r--r--youtube_dl/extractor/youtube.py27
3 files changed, 94 insertions, 1 deletions
diff --git a/youtube_dl/downloader/__init__.py b/youtube_dl/downloader/__init__.py
index f110830c4..dccc59212 100644
--- a/youtube_dl/downloader/__init__.py
+++ b/youtube_dl/downloader/__init__.py
@@ -8,6 +8,7 @@ from .hls import NativeHlsFD
from .http import HttpFD
from .rtsp import RtspFD
from .rtmp import RtmpFD
+from .dash import DashSegmentsFD
from ..utils import (
determine_protocol,
@@ -20,6 +21,7 @@ PROTOCOL_MAP = {
'mms': RtspFD,
'rtsp': RtspFD,
'f4m': F4mFD,
+ 'http_dash_segments': DashSegmentsFD,
}
diff --git a/youtube_dl/downloader/dash.py b/youtube_dl/downloader/dash.py
new file mode 100644
index 000000000..a4685d307
--- /dev/null
+++ b/youtube_dl/downloader/dash.py
@@ -0,0 +1,66 @@
+from __future__ import unicode_literals
+
+import re
+
+from .common import FileDownloader
+from ..compat import compat_urllib_request
+
+
+class DashSegmentsFD(FileDownloader):
+ """
+ Download segments in a DASH manifest
+ """
+ def real_download(self, filename, info_dict):
+ self.report_destination(filename)
+ tmpfilename = self.temp_name(filename)
+ base_url = info_dict['url']
+ segment_urls = info_dict['segment_urls']
+
+ is_test = self.params.get('test', False)
+ remaining_bytes = self._TEST_FILE_SIZE if is_test else None
+ byte_counter = 0
+
+ def append_url_to_file(outf, target_url, target_name, remaining_bytes=None):
+ self.to_screen('[DashSegments] %s: Downloading %s' % (info_dict['id'], target_name))
+ req = compat_urllib_request.Request(target_url)
+ if remaining_bytes is not None:
+ req.add_header('Range', 'bytes=0-%d' % (remaining_bytes - 1))
+
+ data = self.ydl.urlopen(req).read()
+
+ if remaining_bytes is not None:
+ data = data[:remaining_bytes]
+
+ outf.write(data)
+ return len(data)
+
+ def combine_url(base_url, target_url):
+ if re.match(r'^https?://', target_url):
+ return target_url
+ return '%s/%s' % (base_url, target_url)
+
+ with open(tmpfilename, 'wb') as outf:
+ append_url_to_file(
+ outf, combine_url(base_url, info_dict['initialization_url']),
+ 'initialization segment')
+ for i, segment_url in enumerate(segment_urls):
+ segment_len = append_url_to_file(
+ outf, combine_url(base_url, segment_url),
+ 'segment %d / %d' % (i + 1, len(segment_urls)),
+ remaining_bytes)
+ byte_counter += segment_len
+ if remaining_bytes is not None:
+ remaining_bytes -= segment_len
+ if remaining_bytes <= 0:
+ break
+
+ self.try_rename(tmpfilename, filename)
+
+ self._hook_progress({
+ 'downloaded_bytes': byte_counter,
+ 'total_bytes': byte_counter,
+ 'filename': filename,
+ 'status': 'finished',
+ })
+
+ return True
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
index e7f5c7861..9a08924ef 100644
--- a/youtube_dl/extractor/youtube.py
+++ b/youtube_dl/extractor/youtube.py
@@ -535,7 +535,25 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'uploader': 'dorappi2000',
'formats': 'mincount:33',
},
- }
+ },
+ # DASH manifest with segment_list
+ {
+ 'url': 'https://www.youtube.com/embed/CsmdDsKjzN8',
+ 'md5': '8ce563a1d667b599d21064e982ab9e31',
+ 'info_dict': {
+ 'id': 'CsmdDsKjzN8',
+ 'ext': 'mp4',
+ 'upload_date': '20150510',
+ 'uploader': 'Airtek',
+ 'description': 'Retransmisión en directo de la XVIII media maratón de Zaragoza.',
+ 'uploader_id': 'UCzTzUmjXxxacNnL8I3m4LnQ',
+ 'title': 'Retransmisión XVIII Media maratón Zaragoza 2015',
+ },
+ 'params': {
+ 'youtube_include_dash_manifest': True,
+ 'format': '135', # bestvideo
+ }
+ },
]
def __init__(self, *args, **kwargs):
@@ -826,6 +844,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
# TODO implement WebVTT downloading
pass
elif mime_type.startswith('audio/') or mime_type.startswith('video/'):
+ segment_list = r.find('{urn:mpeg:DASH:schema:MPD:2011}SegmentList')
format_id = r.attrib['id']
video_url = url_el.text
filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
@@ -839,6 +858,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'filesize': filesize,
'fps': int_or_none(r.attrib.get('frameRate')),
}
+ if segment_list is not None:
+ f.update({
+ 'initialization_url': segment_list.find('{urn:mpeg:DASH:schema:MPD:2011}Initialization').attrib['sourceURL'],
+ 'segment_urls': [segment.attrib.get('media') for segment in segment_list.findall('{urn:mpeg:DASH:schema:MPD:2011}SegmentURL')],
+ 'protocol': 'http_dash_segments',
+ })
try:
existing_format = next(
fo for fo in formats