]>
jfr.im git - yt-dlp.git/blob - test/test_download.py
3 from __future__
import unicode_literals
5 # Allow direct execution
9 sys
.path
.insert(0, os
.path
.dirname(os
.path
.dirname(os
.path
.abspath(__file__
))))
11 from test
.helper
import (
27 import yt_dlp
.YoutubeDL
28 from yt_dlp
.compat
import (
33 from yt_dlp
.utils
import (
37 UnavailableVideoError
,
39 from yt_dlp
.extractor
import get_info_extractor
44 class YoutubeDL(yt_dlp
.YoutubeDL
):
45 def __init__(self
, *args
, **kwargs
):
46 self
.to_stderr
= self
.to_screen
47 self
.processed_info_dicts
= []
48 super(YoutubeDL
, self
).__init
__(*args
, **kwargs
)
50 def report_warning(self
, message
):
51 # Don't accept warnings during tests
52 raise ExtractorError(message
)
54 def process_info(self
, info_dict
):
55 self
.processed_info_dicts
.append(info_dict
)
56 return super(YoutubeDL
, self
).process_info(info_dict
)
60 with open(fn
, 'rb') as f
:
61 return hashlib
.md5(f
.read()).hexdigest()
67 class TestDownload(unittest
.TestCase
):
68 # Parallel testing in nosetests. See
69 # http://nose.readthedocs.org/en/latest/doc_tests/test_multiprocess/multiprocess.html
70 _multiprocess_shared_
= True
75 """Identify each test with the `add_ie` attribute, if available."""
78 """From 2.7's unittest; 2.6 had _strclass so we can't import it."""
79 return '%s.%s' % (cls
.__module
__, cls
.__name
__)
81 add_ie
= getattr(self
, self
._testMethodName
).add_ie
82 return '%s (%s)%s:' % (self
._testMethodName
,
83 strclass(self
.__class
__),
84 ' [%s]' % add_ie
if add_ie
else '')
89 # Dynamically generate tests
92 def generator(test_case
, tname
):
94 def test_template(self
):
95 ie
= yt_dlp
.extractor
.get_info_extractor(test_case
['name'])()
96 other_ies
= [get_info_extractor(ie_key
)() for ie_key
in test_case
.get('add_ie', [])]
97 is_playlist
= any(k
.startswith('playlist') for k
in test_case
)
98 test_cases
= test_case
.get(
99 'playlist', [] if is_playlist
else [test_case
])
101 def print_skipping(reason
):
102 print('Skipping %s: %s' % (test_case
['name'], reason
))
104 print_skipping('IE marked as not _WORKING')
107 for tc
in test_cases
:
108 info_dict
= tc
.get('info_dict', {})
109 if not (info_dict
.get('id') and info_dict
.get('ext')):
110 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
112 if 'skip' in test_case
:
113 print_skipping(test_case
['skip'])
115 for other_ie
in other_ies
:
116 if not other_ie
.working():
117 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie
.ie_key())
120 params
= get_params(test_case
.get('params', {}))
121 params
['outtmpl'] = tname
+ '_' + params
['outtmpl']
122 if is_playlist
and 'playlist' not in test_case
:
123 params
.setdefault('extract_flat', 'in_playlist')
124 params
.setdefault('playlistend', test_case
.get('playlist_mincount'))
125 params
.setdefault('skip_download', True)
127 ydl
= YoutubeDL(params
, auto_init
=False)
128 ydl
.add_default_info_extractors()
129 finished_hook_called
= set()
132 if status
['status'] == 'finished':
133 finished_hook_called
.add(status
['filename'])
134 ydl
.add_progress_hook(_hook
)
135 expect_warnings(ydl
, test_case
.get('expected_warnings', []))
137 def get_tc_filename(tc
):
138 return ydl
.prepare_filename(tc
.get('info_dict', {}))
142 def try_rm_tcs_files(tcs
=None):
146 tc_filename
= get_tc_filename(tc
)
148 try_rm(tc_filename
+ '.part')
149 try_rm(os
.path
.splitext(tc_filename
)[0] + '.info.json')
155 # We're not using .download here since that is just a shim
156 # for outside error handling, and returns the exit code
157 # instead of the result dict.
158 res_dict
= ydl
.extract_info(
160 force_generic_extractor
=params
.get('force_generic_extractor', False))
161 except (DownloadError
, ExtractorError
) as err
:
162 # Check if the exception is not a network related one
163 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):
166 if try_num
== RETRIES
:
167 report_warning('%s failed due to network errors, skipping...' % tname
)
170 print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num
))
177 self
.assertTrue(res_dict
['_type'] in ['playlist', 'multi_video'])
178 self
.assertTrue('entries' in res_dict
)
179 expect_info_dict(self
, res_dict
, test_case
.get('info_dict', {}))
181 if 'playlist_mincount' in test_case
:
184 len(res_dict
['entries']),
185 test_case
['playlist_mincount'],
186 'Expected at least %d in playlist %s, but got only %d' % (
187 test_case
['playlist_mincount'], test_case
['url'],
188 len(res_dict
['entries'])))
189 if 'playlist_count' in test_case
:
191 len(res_dict
['entries']),
192 test_case
['playlist_count'],
193 'Expected %d entries in playlist %s, but got %d.' % (
194 test_case
['playlist_count'],
196 len(res_dict
['entries']),
198 if 'playlist_duration_sum' in test_case
:
199 got_duration
= sum(e
['duration'] for e
in res_dict
['entries'])
201 test_case
['playlist_duration_sum'], got_duration
)
203 # Generalize both playlists and single videos to unified format for
205 if 'entries' not in res_dict
:
206 res_dict
['entries'] = [res_dict
]
208 for tc_num
, tc
in enumerate(test_cases
):
209 tc_res_dict
= res_dict
['entries'][tc_num
]
210 # First, check test cases' data against extracted data alone
211 expect_info_dict(self
, tc_res_dict
, tc
.get('info_dict', {}))
212 # Now, check downloaded file consistency
213 tc_filename
= get_tc_filename(tc
)
214 if not test_case
.get('params', {}).get('skip_download', False):
215 self
.assertTrue(os
.path
.exists(tc_filename
), msg
='Missing file ' + tc_filename
)
216 self
.assertTrue(tc_filename
in finished_hook_called
)
217 expected_minsize
= tc
.get('file_minsize', 10000)
218 if expected_minsize
is not None:
219 if params
.get('test'):
220 expected_minsize
= max(expected_minsize
, 10000)
221 got_fsize
= os
.path
.getsize(tc_filename
)
223 self
, got_fsize
, expected_minsize
,
224 'Expected %s to be at least %s, but it\'s only %s ' %
225 (tc_filename
, format_bytes(expected_minsize
),
226 format_bytes(got_fsize
)))
228 md5_for_file
= _file_md5(tc_filename
)
229 self
.assertEqual(tc
['md5'], md5_for_file
)
230 # Finally, check test cases' data again but this time against
231 # extracted data from info JSON file written during processing
232 info_json_fn
= os
.path
.splitext(tc_filename
)[0] + '.info.json'
234 os
.path
.exists(info_json_fn
),
235 'Missing info file %s' % info_json_fn
)
236 with io
.open(info_json_fn
, encoding
='utf-8') as infof
:
237 info_dict
= json
.load(infof
)
238 expect_info_dict(self
, info_dict
, tc
.get('info_dict', {}))
241 if is_playlist
and res_dict
is not None and res_dict
.get('entries'):
242 # Remove all other files that may have been extracted if the
243 # extractor returns full results even with extract_flat
244 res_tcs
= [{'info_dict': e}
for e
in res_dict
['entries']]
245 try_rm_tcs_files(res_tcs
)
250 # And add them to TestDownload
251 for n
, test_case
in enumerate(defs
):
252 tname
= 'test_' + str(test_case
['name'])
254 while hasattr(TestDownload
, tname
):
255 tname
= 'test_%s_%d' % (test_case
['name'], i
)
257 test_method
= generator(test_case
, tname
)
258 test_method
.__name
__ = str(tname
)
259 ie_list
= test_case
.get('add_ie')
260 test_method
.add_ie
= ie_list
and ','.join(ie_list
)
261 setattr(TestDownload
, test_method
.__name
__, test_method
)
265 if __name__
== '__main__':