]> jfr.im git - yt-dlp.git/blob - test/test_download.py
[compat] Remove more functions
[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
7 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
8
9
10 import hashlib
11 import json
12 import socket
13 import urllib.error
14 import http.client
15
16 from test.helper import (
17 assertGreaterEqual,
18 expect_info_dict,
19 expect_warnings,
20 get_params,
21 gettestcases,
22 is_download_test,
23 report_warning,
24 try_rm,
25 )
26 import yt_dlp.YoutubeDL # isort: split
27 from yt_dlp.compat import compat_HTTPError
28 from yt_dlp.extractor import get_info_extractor
29 from yt_dlp.utils import (
30 DownloadError,
31 ExtractorError,
32 UnavailableVideoError,
33 format_bytes,
34 )
35
36 RETRIES = 3
37
38
39 class YoutubeDL(yt_dlp.YoutubeDL):
40 def __init__(self, *args, **kwargs):
41 self.to_stderr = self.to_screen
42 self.processed_info_dicts = []
43 super().__init__(*args, **kwargs)
44
45 def report_warning(self, message, *args, **kwargs):
46 # Don't accept warnings during tests
47 raise ExtractorError(message)
48
49 def process_info(self, info_dict):
50 self.processed_info_dicts.append(info_dict.copy())
51 return super().process_info(info_dict)
52
53
54 def _file_md5(fn):
55 with open(fn, 'rb') as f:
56 return hashlib.md5(f.read()).hexdigest()
57
58
59 defs = gettestcases()
60
61
62 @is_download_test
63 class TestDownload(unittest.TestCase):
64 # Parallel testing in nosetests. See
65 # http://nose.readthedocs.org/en/latest/doc_tests/test_multiprocess/multiprocess.html
66 _multiprocess_shared_ = True
67
68 maxDiff = None
69
70 COMPLETED_TESTS = {}
71
72 def __str__(self):
73 """Identify each test with the `add_ie` attribute, if available."""
74
75 def strclass(cls):
76 """From 2.7's unittest; 2.6 had _strclass so we can't import it."""
77 return f'{cls.__module__}.{cls.__name__}'
78
79 add_ie = getattr(self, self._testMethodName).add_ie
80 return '%s (%s)%s:' % (self._testMethodName,
81 strclass(self.__class__),
82 ' [%s]' % add_ie if add_ie else '')
83
84 def setUp(self):
85 self.defs = defs
86
87 # Dynamically generate tests
88
89
90 def generator(test_case, tname):
91
92 def test_template(self):
93 if self.COMPLETED_TESTS.get(tname):
94 return
95 self.COMPLETED_TESTS[tname] = True
96 ie = yt_dlp.extractor.get_info_extractor(test_case['name'])()
97 other_ies = [get_info_extractor(ie_key)() for ie_key in test_case.get('add_ie', [])]
98 is_playlist = any(k.startswith('playlist') for k in test_case)
99 test_cases = test_case.get(
100 'playlist', [] if is_playlist else [test_case])
101
102 def print_skipping(reason):
103 print('Skipping %s: %s' % (test_case['name'], reason))
104 self.skipTest(reason)
105
106 if not ie.working():
107 print_skipping('IE marked as not _WORKING')
108
109 for tc in test_cases:
110 info_dict = tc.get('info_dict', {})
111 params = tc.get('params', {})
112 if not info_dict.get('id'):
113 raise Exception('Test definition incorrect. \'id\' key is not present')
114 elif not info_dict.get('ext'):
115 if params.get('skip_download') and params.get('ignore_no_formats_error'):
116 continue
117 raise Exception('Test definition incorrect. The output file cannot be known. \'ext\' key is not present')
118
119 if 'skip' in test_case:
120 print_skipping(test_case['skip'])
121
122 for other_ie in other_ies:
123 if not other_ie.working():
124 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
125
126 params = get_params(test_case.get('params', {}))
127 params['outtmpl'] = tname + '_' + params['outtmpl']
128 if is_playlist and 'playlist' not in test_case:
129 params.setdefault('extract_flat', 'in_playlist')
130 params.setdefault('playlistend', test_case.get('playlist_mincount'))
131 params.setdefault('skip_download', True)
132
133 ydl = YoutubeDL(params, auto_init=False)
134 ydl.add_default_info_extractors()
135 finished_hook_called = set()
136
137 def _hook(status):
138 if status['status'] == 'finished':
139 finished_hook_called.add(status['filename'])
140 ydl.add_progress_hook(_hook)
141 expect_warnings(ydl, test_case.get('expected_warnings', []))
142
143 def get_tc_filename(tc):
144 return ydl.prepare_filename(dict(tc.get('info_dict', {})))
145
146 res_dict = None
147
148 def try_rm_tcs_files(tcs=None):
149 if tcs is None:
150 tcs = test_cases
151 for tc in tcs:
152 tc_filename = get_tc_filename(tc)
153 try_rm(tc_filename)
154 try_rm(tc_filename + '.part')
155 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
156 try_rm_tcs_files()
157 try:
158 try_num = 1
159 while True:
160 try:
161 # We're not using .download here since that is just a shim
162 # for outside error handling, and returns the exit code
163 # instead of the result dict.
164 res_dict = ydl.extract_info(
165 test_case['url'],
166 force_generic_extractor=params.get('force_generic_extractor', False))
167 except (DownloadError, ExtractorError) as err:
168 # Check if the exception is not a network related one
169 if not err.exc_info[0] in (urllib.error.URLError, socket.timeout, UnavailableVideoError, http.client.BadStatusLine) or (err.exc_info[0] == compat_HTTPError and err.exc_info[1].code == 503):
170 raise
171
172 if try_num == RETRIES:
173 report_warning('%s failed due to network errors, skipping...' % tname)
174 return
175
176 print(f'Retrying: {try_num} failed tries\n\n##########\n\n')
177
178 try_num += 1
179 else:
180 break
181
182 if is_playlist:
183 self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
184 self.assertTrue('entries' in res_dict)
185 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
186
187 if 'playlist_mincount' in test_case:
188 assertGreaterEqual(
189 self,
190 len(res_dict['entries']),
191 test_case['playlist_mincount'],
192 'Expected at least %d in playlist %s, but got only %d' % (
193 test_case['playlist_mincount'], test_case['url'],
194 len(res_dict['entries'])))
195 if 'playlist_count' in test_case:
196 self.assertEqual(
197 len(res_dict['entries']),
198 test_case['playlist_count'],
199 'Expected %d entries in playlist %s, but got %d.' % (
200 test_case['playlist_count'],
201 test_case['url'],
202 len(res_dict['entries']),
203 ))
204 if 'playlist_duration_sum' in test_case:
205 got_duration = sum(e['duration'] for e in res_dict['entries'])
206 self.assertEqual(
207 test_case['playlist_duration_sum'], got_duration)
208
209 # Generalize both playlists and single videos to unified format for
210 # simplicity
211 if 'entries' not in res_dict:
212 res_dict['entries'] = [res_dict]
213
214 for tc_num, tc in enumerate(test_cases):
215 tc_res_dict = res_dict['entries'][tc_num]
216 # First, check test cases' data against extracted data alone
217 expect_info_dict(self, tc_res_dict, tc.get('info_dict', {}))
218 # Now, check downloaded file consistency
219 tc_filename = get_tc_filename(tc)
220 if not test_case.get('params', {}).get('skip_download', False):
221 self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
222 self.assertTrue(tc_filename in finished_hook_called)
223 expected_minsize = tc.get('file_minsize', 10000)
224 if expected_minsize is not None:
225 if params.get('test'):
226 expected_minsize = max(expected_minsize, 10000)
227 got_fsize = os.path.getsize(tc_filename)
228 assertGreaterEqual(
229 self, got_fsize, expected_minsize,
230 'Expected %s to be at least %s, but it\'s only %s ' %
231 (tc_filename, format_bytes(expected_minsize),
232 format_bytes(got_fsize)))
233 if 'md5' in tc:
234 md5_for_file = _file_md5(tc_filename)
235 self.assertEqual(tc['md5'], md5_for_file)
236 # Finally, check test cases' data again but this time against
237 # extracted data from info JSON file written during processing
238 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
239 self.assertTrue(
240 os.path.exists(info_json_fn),
241 'Missing info file %s' % info_json_fn)
242 with open(info_json_fn, encoding='utf-8') as infof:
243 info_dict = json.load(infof)
244 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
245 finally:
246 try_rm_tcs_files()
247 if is_playlist and res_dict is not None and res_dict.get('entries'):
248 # Remove all other files that may have been extracted if the
249 # extractor returns full results even with extract_flat
250 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
251 try_rm_tcs_files(res_tcs)
252
253 return test_template
254
255
256 # And add them to TestDownload
257 tests_counter = {}
258 for test_case in defs:
259 name = test_case['name']
260 i = tests_counter.get(name, 0)
261 tests_counter[name] = i + 1
262 tname = f'test_{name}_{i}' if i else f'test_{name}'
263 test_method = generator(test_case, tname)
264 test_method.__name__ = str(tname)
265 ie_list = test_case.get('add_ie')
266 test_method.add_ie = ie_list and ','.join(ie_list)
267 setattr(TestDownload, test_method.__name__, test_method)
268 del test_method
269
270
271 def batch_generator(name, num_tests):
272
273 def test_template(self):
274 for i in range(num_tests):
275 getattr(self, f'test_{name}_{i}' if i else f'test_{name}')()
276
277 return test_template
278
279
280 for name, num_tests in tests_counter.items():
281 test_method = batch_generator(name, num_tests)
282 test_method.__name__ = f'test_{name}_all'
283 test_method.add_ie = ''
284 setattr(TestDownload, test_method.__name__, test_method)
285 del test_method
286
287
288 if __name__ == '__main__':
289 unittest.main()