]> jfr.im git - yt-dlp.git/blob - test/test_download.py
Update to ytdl-commit-8a158a9
[yt-dlp.git] / test / test_download.py
1 #!/usr/bin/env python3
2 # Allow direct execution
3 import hashlib
4 import json
5 import os
6 import socket
7 import sys
8 import unittest
9
10 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
11
12 from test.helper import (
13 assertGreaterEqual,
14 expect_info_dict,
15 expect_warnings,
16 get_params,
17 gettestcases,
18 is_download_test,
19 report_warning,
20 try_rm,
21 )
22
23 import yt_dlp.YoutubeDL
24 from yt_dlp.compat import (
25 compat_http_client,
26 compat_HTTPError,
27 compat_urllib_error,
28 )
29 from yt_dlp.extractor import get_info_extractor
30 from yt_dlp.utils import (
31 DownloadError,
32 ExtractorError,
33 UnavailableVideoError,
34 format_bytes,
35 )
36
37 RETRIES = 3
38
39
40 class YoutubeDL(yt_dlp.YoutubeDL):
41 def __init__(self, *args, **kwargs):
42 self.to_stderr = self.to_screen
43 self.processed_info_dicts = []
44 super().__init__(*args, **kwargs)
45
46 def report_warning(self, message):
47 # Don't accept warnings during tests
48 raise ExtractorError(message)
49
50 def process_info(self, info_dict):
51 self.processed_info_dicts.append(info_dict.copy())
52 return super().process_info(info_dict)
53
54
55 def _file_md5(fn):
56 with open(fn, 'rb') as f:
57 return hashlib.md5(f.read()).hexdigest()
58
59
60 defs = gettestcases()
61
62
63 @is_download_test
64 class TestDownload(unittest.TestCase):
65 # Parallel testing in nosetests. See
66 # http://nose.readthedocs.org/en/latest/doc_tests/test_multiprocess/multiprocess.html
67 _multiprocess_shared_ = True
68
69 maxDiff = None
70
71 COMPLETED_TESTS = {}
72
73 def __str__(self):
74 """Identify each test with the `add_ie` attribute, if available."""
75
76 def strclass(cls):
77 """From 2.7's unittest; 2.6 had _strclass so we can't import it."""
78 return f'{cls.__module__}.{cls.__name__}'
79
80 add_ie = getattr(self, self._testMethodName).add_ie
81 return '%s (%s)%s:' % (self._testMethodName,
82 strclass(self.__class__),
83 ' [%s]' % add_ie if add_ie else '')
84
85 def setUp(self):
86 self.defs = defs
87
88 # Dynamically generate tests
89
90
91 def generator(test_case, tname):
92
93 def test_template(self):
94 if self.COMPLETED_TESTS.get(tname):
95 return
96 self.COMPLETED_TESTS[tname] = True
97 ie = yt_dlp.extractor.get_info_extractor(test_case['name'])()
98 other_ies = [get_info_extractor(ie_key)() for ie_key in test_case.get('add_ie', [])]
99 is_playlist = any(k.startswith('playlist') for k in test_case)
100 test_cases = test_case.get(
101 'playlist', [] if is_playlist else [test_case])
102
103 def print_skipping(reason):
104 print('Skipping %s: %s' % (test_case['name'], reason))
105 self.skipTest(reason)
106
107 if not ie.working():
108 print_skipping('IE marked as not _WORKING')
109
110 for tc in test_cases:
111 info_dict = tc.get('info_dict', {})
112 params = tc.get('params', {})
113 if not info_dict.get('id'):
114 raise Exception('Test definition incorrect. \'id\' key is not present')
115 elif not info_dict.get('ext'):
116 if params.get('skip_download') and params.get('ignore_no_formats_error'):
117 continue
118 raise Exception('Test definition incorrect. The output file cannot be known. \'ext\' key is not present')
119
120 if 'skip' in test_case:
121 print_skipping(test_case['skip'])
122
123 for other_ie in other_ies:
124 if not other_ie.working():
125 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
126
127 params = get_params(test_case.get('params', {}))
128 params['outtmpl'] = tname + '_' + params['outtmpl']
129 if is_playlist and 'playlist' not in test_case:
130 params.setdefault('extract_flat', 'in_playlist')
131 params.setdefault('playlistend', test_case.get('playlist_mincount'))
132 params.setdefault('skip_download', True)
133
134 ydl = YoutubeDL(params, auto_init=False)
135 ydl.add_default_info_extractors()
136 finished_hook_called = set()
137
138 def _hook(status):
139 if status['status'] == 'finished':
140 finished_hook_called.add(status['filename'])
141 ydl.add_progress_hook(_hook)
142 expect_warnings(ydl, test_case.get('expected_warnings', []))
143
144 def get_tc_filename(tc):
145 return ydl.prepare_filename(dict(tc.get('info_dict', {})))
146
147 res_dict = None
148
149 def try_rm_tcs_files(tcs=None):
150 if tcs is None:
151 tcs = test_cases
152 for tc in tcs:
153 tc_filename = get_tc_filename(tc)
154 try_rm(tc_filename)
155 try_rm(tc_filename + '.part')
156 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
157 try_rm_tcs_files()
158 try:
159 try_num = 1
160 while True:
161 try:
162 # We're not using .download here since that is just a shim
163 # for outside error handling, and returns the exit code
164 # instead of the result dict.
165 res_dict = ydl.extract_info(
166 test_case['url'],
167 force_generic_extractor=params.get('force_generic_extractor', False))
168 except (DownloadError, ExtractorError) as err:
169 # Check if the exception is not a network related one
170 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):
171 raise
172
173 if try_num == RETRIES:
174 report_warning('%s failed due to network errors, skipping...' % tname)
175 return
176
177 print(f'Retrying: {try_num} failed tries\n\n##########\n\n')
178
179 try_num += 1
180 else:
181 break
182
183 if is_playlist:
184 self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
185 self.assertTrue('entries' in res_dict)
186 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
187
188 if 'playlist_mincount' in test_case:
189 assertGreaterEqual(
190 self,
191 len(res_dict['entries']),
192 test_case['playlist_mincount'],
193 'Expected at least %d in playlist %s, but got only %d' % (
194 test_case['playlist_mincount'], test_case['url'],
195 len(res_dict['entries'])))
196 if 'playlist_count' in test_case:
197 self.assertEqual(
198 len(res_dict['entries']),
199 test_case['playlist_count'],
200 'Expected %d entries in playlist %s, but got %d.' % (
201 test_case['playlist_count'],
202 test_case['url'],
203 len(res_dict['entries']),
204 ))
205 if 'playlist_duration_sum' in test_case:
206 got_duration = sum(e['duration'] for e in res_dict['entries'])
207 self.assertEqual(
208 test_case['playlist_duration_sum'], got_duration)
209
210 # Generalize both playlists and single videos to unified format for
211 # simplicity
212 if 'entries' not in res_dict:
213 res_dict['entries'] = [res_dict]
214
215 for tc_num, tc in enumerate(test_cases):
216 tc_res_dict = res_dict['entries'][tc_num]
217 # First, check test cases' data against extracted data alone
218 expect_info_dict(self, tc_res_dict, tc.get('info_dict', {}))
219 # Now, check downloaded file consistency
220 tc_filename = get_tc_filename(tc)
221 if not test_case.get('params', {}).get('skip_download', False):
222 self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
223 self.assertTrue(tc_filename in finished_hook_called)
224 expected_minsize = tc.get('file_minsize', 10000)
225 if expected_minsize is not None:
226 if params.get('test'):
227 expected_minsize = max(expected_minsize, 10000)
228 got_fsize = os.path.getsize(tc_filename)
229 assertGreaterEqual(
230 self, got_fsize, expected_minsize,
231 'Expected %s to be at least %s, but it\'s only %s ' %
232 (tc_filename, format_bytes(expected_minsize),
233 format_bytes(got_fsize)))
234 if 'md5' in tc:
235 md5_for_file = _file_md5(tc_filename)
236 self.assertEqual(tc['md5'], md5_for_file)
237 # Finally, check test cases' data again but this time against
238 # extracted data from info JSON file written during processing
239 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
240 self.assertTrue(
241 os.path.exists(info_json_fn),
242 'Missing info file %s' % info_json_fn)
243 with open(info_json_fn, encoding='utf-8') as infof:
244 info_dict = json.load(infof)
245 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
246 finally:
247 try_rm_tcs_files()
248 if is_playlist and res_dict is not None and res_dict.get('entries'):
249 # Remove all other files that may have been extracted if the
250 # extractor returns full results even with extract_flat
251 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
252 try_rm_tcs_files(res_tcs)
253
254 return test_template
255
256
257 # And add them to TestDownload
258 tests_counter = {}
259 for test_case in defs:
260 name = test_case['name']
261 i = tests_counter.get(name, 0)
262 tests_counter[name] = i + 1
263 tname = f'test_{name}_{i}' if i else f'test_{name}'
264 test_method = generator(test_case, tname)
265 test_method.__name__ = str(tname)
266 ie_list = test_case.get('add_ie')
267 test_method.add_ie = ie_list and ','.join(ie_list)
268 setattr(TestDownload, test_method.__name__, test_method)
269 del test_method
270
271
272 def batch_generator(name, num_tests):
273
274 def test_template(self):
275 for i in range(num_tests):
276 getattr(self, f'test_{name}_{i}' if i else f'test_{name}')()
277
278 return test_template
279
280
281 for name, num_tests in tests_counter.items():
282 test_method = batch_generator(name, num_tests)
283 test_method.__name__ = f'test_{name}_all'
284 test_method.add_ie = ''
285 setattr(TestDownload, test_method.__name__, test_method)
286 del test_method
287
288
289 if __name__ == '__main__':
290 unittest.main()