[youtube] Adds #1312 Download annotations

Adds #1321 Download annotations from youtube
Annotations are downloaded and written to a .annotations.xml file using the https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=$VIDEOID API.
Added unit test for annotations.
This commit is contained in:
Jai Grimshaw 2013-10-14 16:18:58 +11:00
parent 06723d47c4
commit 1fb07d10a3
4 changed files with 113 additions and 0 deletions

View File

@ -0,0 +1,82 @@
#!/usr/bin/env python
# coding: utf-8
import xml.etree.ElementTree
import os
import sys
import unittest
# Allow direct execution
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
import youtube_dl.YoutubeDL
import youtube_dl.extractor
from youtube_dl.utils import *
from .helper import try_rm
PARAMETERS_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "parameters.json")
# General configuration (from __init__, not very elegant...)
jar = compat_cookiejar.CookieJar()
cookie_processor = compat_urllib_request.HTTPCookieProcessor(jar)
proxy_handler = compat_urllib_request.ProxyHandler()
opener = compat_urllib_request.build_opener(proxy_handler, cookie_processor, YoutubeDLHandler())
compat_urllib_request.install_opener(opener)
class YoutubeDL(youtube_dl.YoutubeDL):
def __init__(self, *args, **kwargs):
super(YoutubeDL, self).__init__(*args, **kwargs)
self.to_stderr = self.to_screen
with io.open(PARAMETERS_FILE, encoding='utf-8') as pf:
params = json.load(pf)
params['writeannotations'] = True
params['skip_download'] = True
params['writeinfojson'] = False
params['format'] = 'flv'
TEST_ID = 'gr51aVj-mLg'
ANNOTATIONS_FILE = TEST_ID + '.flv.annotations.xml'
EXPECTED_ANNOTATIONS = ['Speech bubble', 'Note', 'Title', 'Spotlight', 'Label']
class TestAnnotations(unittest.TestCase):
def setUp(self):
# Clear old files
self.tearDown()
def test_info_json(self):
expected = list(EXPECTED_ANNOTATIONS) #Two annotations could have the same text.
ie = youtube_dl.extractor.YoutubeIE()
ydl = YoutubeDL(params)
ydl.add_info_extractor(ie)
ydl.download([TEST_ID])
self.assertTrue(os.path.exists(ANNOTATIONS_FILE))
annoxml = None
with io.open(ANNOTATIONS_FILE, 'r', encoding='utf-8') as annof:
annoxml = xml.etree.ElementTree.parse(annof)
self.assertTrue(annoxml is not None, 'Failed to parse annotations XML')
root = annoxml.getroot()
self.assertEqual(root.tag, 'document')
annotationsTag = root.find('annotations')
self.assertEqual(annotationsTag.tag, 'annotations')
annotations = annotationsTag.findall('annotation')
#Not all the annotations have TEXT children and the annotations are returned unsorted.
for a in annotations:
self.assertEqual(a.tag, 'annotation')
if a.get('type') == 'text':
textTag = a.find('TEXT')
text = textTag.text
self.assertTrue(text in expected) #assertIn only added in python 2.7
#remove the first occurance, there could be more than one annotation with the same text
expected.remove(text)
#We should have seen (and removed) all the expected annotation texts.
self.assertEqual(len(expected), 0, 'Not all expected annotations were found.')
def tearDown(self):
try_rm(ANNOTATIONS_FILE)
if __name__ == '__main__':
unittest.main()

View File

@ -71,6 +71,7 @@ class YoutubeDL(object):
logtostderr: Log messages to stderr instead of stdout.
writedescription: Write the video description to a .description file
writeinfojson: Write the video description to a .info.json file
writeannotations: Write the video annotations to a .annotations.xml file
writethumbnail: Write the thumbnail image to a file
writesubtitles: Write the video subtitles to a file
writeautomaticsub: Write the automatic subtitles to a file
@ -258,6 +259,10 @@ def report_writeinfojson(self, infofn):
""" Report that the metadata file has been written """
self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
def report_writeannotations(self, annofn):
""" Report that the annotations file has been written. """
self.to_screen(u'[info] Writing video annotations to: ' + annofn)
def report_file_already_downloaded(self, file_name):
"""Report file has already been fully downloaded."""
try:
@ -522,6 +527,18 @@ def process_info(self, info_dict):
self.report_error(u'Cannot write description file ' + descfn)
return
if self.params.get('writeannotations', False):
try:
annofn = filename + u'.annotations.xml'
self.report_writeannotations(annofn)
with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
annofile.write(info_dict['annotations'])
except (KeyError, TypeError):
self.report_warning(u'There are no annotations to write.')
except (OSError, IOError):
self.report_error(u'Cannot write annotations file: ' + annofn)
return
subtitles_are_requested = any([self.params.get('writesubtitles', False),
self.params.get('writeautomaticsub')])

View File

@ -339,6 +339,9 @@ def _hide_login_info(opts):
filesystem.add_option('--write-info-json',
action='store_true', dest='writeinfojson',
help='write video metadata to a .info.json file', default=False)
filesystem.add_option('--write-annotations',
action='store_true', dest='writeannotations',
help='write video annotations to a .annotation file', default=False)
filesystem.add_option('--write-thumbnail',
action='store_true', dest='writethumbnail',
help='write thumbnail image to disk', default=False)
@ -601,6 +604,7 @@ def _real_main(argv=None):
'nopart': opts.nopart,
'updatetime': opts.updatetime,
'writedescription': opts.writedescription,
'writeannotations': opts.writeannotations,
'writeinfojson': opts.writeinfojson,
'writethumbnail': opts.writethumbnail,
'writesubtitles': opts.writesubtitles,

View File

@ -1250,6 +1250,10 @@ def _get_urls(_manifest):
url_map[itag] = format_url
return url_map
def _extract_annotations(self, video_id):
url = 'https://www.youtube.com/annotations_invideo?features=1&legacy=1&video_id=%s' % video_id
return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.')
def _real_extract(self, url):
# Extract original video URL from URL with redirection, like age verification, using next_url parameter
mobj = re.search(self._NEXT_URL_RE, url)
@ -1382,6 +1386,11 @@ def _real_extract(self, url):
else:
video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
# annotations
video_annotations = None
if self._downloader.params.get('writeannotations', False):
video_annotations = self._extract_annotations(video_id)
# Decide which formats to download
try:
@ -1495,6 +1504,7 @@ def _real_extract(self, url):
'subtitles': video_subtitles,
'duration': video_duration,
'age_limit': 18 if age_gate else 0,
'annotations': video_annotations
})
return results