def test_proxy_with_idn(self):
ydl = YoutubeDL({
'proxy': '127.0.0.1:{0}'.format(self.port),
})
url = 'http://中文.tw/'
response = ydl.urlopen(url).read().decode('utf-8')
# b'xn--fiq228c' is '中文'.encode('idna')
self.assertEqual(response, 'normal: http://xn--fiq228c.tw/')
def test_post_hooks(self):
self.params['post_hooks'] = [self.hook_one, self.hook_two]
ydl = YoutubeDL(self.params)
ydl.download([TEST_ID])
self.assertEqual(self.stored_name_1, EXPECTED_NAME,
'Not the expected name from hook 1')
self.assertEqual(self.stored_name_2, EXPECTED_NAME,
'Not the expected name from hook 2')
def test_format_note(self):
ydl = YoutubeDL()
self.assertEqual(ydl._format_note({}), '')
assertRegexpMatches(self, ydl._format_note({
'vbr': 10,
}), r'^\s*10k$')
assertRegexpMatches(self, ydl._format_note({
'fps': 30,
}), r'^30fps$')
def test_unicode_path_redirection(self):
# XXX: Python 3 http server does not allow non-ASCII header values
if sys.version_info[0] == 3:
return
ydl = YoutubeDL({'logger': FakeLogger()})
r = ydl.extract_info('http://127.0.0.1:%d/302' % self.port)
self.assertEqual(r['entries'][0]['url'],
'http://127.0.0.1:%d/vid.mp4' % self.port)
def test_nocheckcertificate(self):
if sys.version_info >= (2, 7, 9): # No certificate checking anyways
ydl = YoutubeDL({'logger': FakeLogger()})
self.assertRaises(Exception, ydl.extract_info,
'https://127.0.0.1:%d/video.html' % self.port)
ydl = YoutubeDL({'logger': FakeLogger(), 'nocheckcertificate': True})
r = ydl.extract_info('https://127.0.0.1:%d/video.html' % self.port)
self.assertEqual(r['entries'][0]['url'],
'https://127.0.0.1:%d/vid.mp4' % self.port)
def test_proxy(self):
geo_proxy = '127.0.0.1:{0}'.format(self.geo_port)
ydl = YoutubeDL({
'proxy': '127.0.0.1:{0}'.format(self.port),
'geo_verification_proxy': geo_proxy,
})
url = 'http://foo.com/bar'
response = ydl.urlopen(url).read().decode('utf-8')
self.assertEqual(response, 'normal: {0}'.format(url))
req = compat_urllib_request.Request(url)
req.add_header('Ytdl-request-proxy', geo_proxy)
response = ydl.urlopen(req).read().decode('utf-8')
self.assertEqual(response, 'geo: {0}'.format(url))
def download(self, params, ep):
params['logger'] = FakeLogger()
ydl = YoutubeDL(params)
downloader = HttpFD(ydl, params)
filename = 'testfile.mp4'
try_rm(encodeFilename(filename))
self.assertTrue(downloader.real_download(filename, {
'url': 'http://127.0.0.1:%d/%s' % (self.port, ep),
}))
self.assertEqual(os.path.getsize(encodeFilename(filename)), TEST_SIZE)
try_rm(encodeFilename(filename))
def test_info_json(self):
expected = list(
EXPECTED_ANNOTATIONS) # Two annotations could have the same text.
ie = youtube_dlc.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 occurrence, 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 _download_restricted(url, filename, age):
""" Returns true if the file has been downloaded """
params = {
'age_limit': age,
'skip_download': True,
'writeinfojson': True,
'outtmpl': '%(id)s.%(ext)s',
}
ydl = YoutubeDL(params)
ydl.add_default_info_extractors()
json_filename = os.path.splitext(filename)[0] + '.info.json'
try_rm(json_filename)
ydl.download([url])
res = os.path.exists(json_filename)
try_rm(json_filename)
return res
'outtmpl': 'downloads/%(extractor)s-%(id)s-%(title)s.%(ext)s',
'restrictfilenames': True,
'noplaylist': True,
'nocheckcertificate': True,
'ignoreerrors': False,
'logtostderr': False,
'quiet': True,
'no_warnings': True,
'verbose': True,
'default_search': 'auto',
'source_address': '0.0.0.0' # ipv6 addresses cause issues sometimes
}
ffmpegopts = {'before_options': '-nostdin', 'options': '-vn'}
ytdl = YoutubeDL(ytdl_opts)
class YTDLSource(discord.PCMVolumeTransformer):
def __init__(self, source, *, data, requester):
super().__init__(source)
self.requester = requester
self.web_url = data.get('webpage_url')
self.title = data.get('title')
self.duration = data.get('duration')
def __getitem__(self, item: str):
# Allows us to access attributes similar to a dict.
# This is only useful when you are NOT downloading.
#
def run_pp(params, PP):
with open(filename, 'wt') as f:
f.write('EXAMPLE')
ydl = YoutubeDL(params)
ydl.add_post_processor(PP())
ydl.post_process(filename, {'filepath': filename})
def fname(templ, na_placeholder='NA'):
params = {'outtmpl': templ}
if na_placeholder != 'NA':
params['outtmpl_na_placeholder'] = na_placeholder
ydl = YoutubeDL(params)
return ydl.prepare_filename(info)
def test_post_hook_exception(self):
self.params['post_hooks'] = [self.hook_three]
ydl = YoutubeDL(self.params)
self.assertRaises(DownloadError, ydl.download, [TEST_ID])
def test_template(self):
ie = youtube_dlc.extractor.get_info_extractor(test_case['name'])()
other_ies = [get_info_extractor(ie_key)() for ie_key in test_case.get('add_ie', [])]
is_playlist = any(k.startswith('playlist') for k in test_case)
test_cases = test_case.get(
'playlist', [] if is_playlist else [test_case])
def print_skipping(reason):
print('Skipping %s: %s' % (test_case['name'], reason))
if not ie.working():
print_skipping('IE marked as not _WORKING')
return
for tc in test_cases:
info_dict = tc.get('info_dict', {})
if not (info_dict.get('id') and info_dict.get('ext')):
raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
if 'skip' in test_case:
print_skipping(test_case['skip'])
return
for other_ie in other_ies:
if not other_ie.working():
print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
return
params = get_params(test_case.get('params', {}))
params['outtmpl'] = tname + '_' + params['outtmpl']
if is_playlist and 'playlist' not in test_case:
params.setdefault('extract_flat', 'in_playlist')
params.setdefault('skip_download', True)
ydl = YoutubeDL(params, auto_init=False)
ydl.add_default_info_extractors()
finished_hook_called = set()
def _hook(status):
if status['status'] == 'finished':
finished_hook_called.add(status['filename'])
ydl.add_progress_hook(_hook)
expect_warnings(ydl, test_case.get('expected_warnings', []))
def get_tc_filename(tc):
return ydl.prepare_filename(tc.get('info_dict', {}))
res_dict = None
def try_rm_tcs_files(tcs=None):
if tcs is None:
tcs = test_cases
for tc in tcs:
tc_filename = get_tc_filename(tc)
try_rm(tc_filename)
try_rm(tc_filename + '.part')
try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
try_rm_tcs_files()
try:
try_num = 1
while True:
try:
# We're not using .download here since that is just a shim
# for outside error handling, and returns the exit code
# instead of the result dict.
res_dict = ydl.extract_info(
test_case['url'],
force_generic_extractor=params.get('force_generic_extractor', False))
except (DownloadError, ExtractorError) as err:
# Check if the exception is not a network related one
if not err.exc_info[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError, compat_http_client.BadStatusLine) or (err.exc_info[0] == compat_HTTPError and err.exc_info[1].code == 503):
raise
if try_num == RETRIES:
report_warning('%s failed due to network errors, skipping...' % tname)
return
print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
try_num += 1
else:
break
if is_playlist:
self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
self.assertTrue('entries' in res_dict)
expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
if 'playlist_mincount' in test_case:
assertGreaterEqual(
self,
len(res_dict['entries']),
test_case['playlist_mincount'],
'Expected at least %d in playlist %s, but got only %d' % (
test_case['playlist_mincount'], test_case['url'],
len(res_dict['entries'])))
if 'playlist_count' in test_case:
self.assertEqual(
len(res_dict['entries']),
test_case['playlist_count'],
'Expected %d entries in playlist %s, but got %d.' % (
test_case['playlist_count'],
test_case['url'],
len(res_dict['entries']),
))
if 'playlist_duration_sum' in test_case:
got_duration = sum(e['duration'] for e in res_dict['entries'])
self.assertEqual(
test_case['playlist_duration_sum'], got_duration)
# Generalize both playlists and single videos to unified format for
# simplicity
if 'entries' not in res_dict:
res_dict['entries'] = [res_dict]
for tc_num, tc in enumerate(test_cases):
tc_res_dict = res_dict['entries'][tc_num]
# First, check test cases' data against extracted data alone
expect_info_dict(self, tc_res_dict, tc.get('info_dict', {}))
# Now, check downloaded file consistency
tc_filename = get_tc_filename(tc)
if not test_case.get('params', {}).get('skip_download', False):
self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
self.assertTrue(tc_filename in finished_hook_called)
expected_minsize = tc.get('file_minsize', 10000)
if expected_minsize is not None:
if params.get('test'):
expected_minsize = max(expected_minsize, 10000)
got_fsize = os.path.getsize(tc_filename)
assertGreaterEqual(
self, got_fsize, expected_minsize,
'Expected %s to be at least %s, but it\'s only %s ' %
(tc_filename, format_bytes(expected_minsize),
format_bytes(got_fsize)))
if 'md5' in tc:
md5_for_file = _file_md5(tc_filename)
self.assertEqual(tc['md5'], md5_for_file)
# Finally, check test cases' data again but this time against
# extracted data from info JSON file written during processing
info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
self.assertTrue(
os.path.exists(info_json_fn),
'Missing info file %s' % info_json_fn)
with io.open(info_json_fn, encoding='utf-8') as infof:
info_dict = json.load(infof)
expect_info_dict(self, info_dict, tc.get('info_dict', {}))
finally:
try_rm_tcs_files()
if is_playlist and res_dict is not None and res_dict.get('entries'):
# Remove all other files that may have been extracted if the
# extractor returns full results even with extract_flat
res_tcs = [{'info_dict': e} for e in res_dict['entries']]
try_rm_tcs_files(res_tcs)
def fname(templ):
ydl = YoutubeDL({'outtmpl': templ})
return ydl.prepare_filename(info)