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