]> jfr.im git - yt-dlp.git/blob - test/test_download.py
Add key `requested_downloads` in the root `info_dict`
[yt-dlp.git] / test / test_download.py
1 #!/usr/bin/env python3
2
3 from __future__ import unicode_literals
4
5 # Allow direct execution
6 import os
7 import sys
8 import unittest
9 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
10
11 from test.helper import (
12 assertGreaterEqual,
13 expect_info_dict,
14 expect_warnings,
15 get_params,
16 gettestcases,
17 is_download_test,
18 report_warning,
19 try_rm,
20 )
21
22
23 import hashlib
24 import io
25 import json
26 import socket
27
28 import yt_dlp.YoutubeDL
29 from yt_dlp.compat import (
30 compat_http_client,
31 compat_urllib_error,
32 compat_HTTPError,
33 )
34 from yt_dlp.utils import (
35 DownloadError,
36 ExtractorError,
37 format_bytes,
38 UnavailableVideoError,
39 )
40 from yt_dlp.extractor import get_info_extractor
41
42 RETRIES = 3
43
44
45 class YoutubeDL(yt_dlp.YoutubeDL):
46 def __init__(self, *args, **kwargs):
47 self.to_stderr = self.to_screen
48 self.processed_info_dicts = []
49 super(YoutubeDL, self).__init__(*args, **kwargs)
50
51 def report_warning(self, message):
52 # Don't accept warnings during tests
53 raise ExtractorError(message)
54
55 def process_info(self, info_dict):
56 self.processed_info_dicts.append(info_dict.copy())
57 return super(YoutubeDL, self).process_info(info_dict)
58
59
60 def _file_md5(fn):
61 with open(fn, 'rb') as f:
62 return hashlib.md5(f.read()).hexdigest()
63
64
65 defs = gettestcases()
66
67
68 @is_download_test
69 class TestDownload(unittest.TestCase):
70 # Parallel testing in nosetests. See
71 # http://nose.readthedocs.org/en/latest/doc_tests/test_multiprocess/multiprocess.html
72 _multiprocess_shared_ = True
73
74 maxDiff = None
75
76 COMPLETED_TESTS = {}
77
78 def __str__(self):
79 """Identify each test with the `add_ie` attribute, if available."""
80
81 def strclass(cls):
82 """From 2.7's unittest; 2.6 had _strclass so we can't import it."""
83 return '%s.%s' % (cls.__module__, cls.__name__)
84
85 add_ie = getattr(self, self._testMethodName).add_ie
86 return '%s (%s)%s:' % (self._testMethodName,
87 strclass(self.__class__),
88 ' [%s]' % add_ie if add_ie else '')
89
90 def setUp(self):
91 self.defs = defs
92
93 # Dynamically generate tests
94
95
96 def generator(test_case, tname):
97
98 def test_template(self):
99 if self.COMPLETED_TESTS.get(tname):
100 return
101 self.COMPLETED_TESTS[tname] = True
102 ie = yt_dlp.extractor.get_info_extractor(test_case['name'])()
103 other_ies = [get_info_extractor(ie_key)() for ie_key in test_case.get('add_ie', [])]
104 is_playlist = any(k.startswith('playlist') for k in test_case)
105 test_cases = test_case.get(
106 'playlist', [] if is_playlist else [test_case])
107
108 def print_skipping(reason):
109 print('Skipping %s: %s' % (test_case['name'], reason))
110 if not ie.working():
111 print_skipping('IE marked as not _WORKING')
112 return
113
114 for tc in test_cases:
115 info_dict = tc.get('info_dict', {})
116 params = tc.get('params', {})
117 if not info_dict.get('id'):
118 raise Exception('Test definition incorrect. \'id\' key is not present')
119 elif not info_dict.get('ext'):
120 if params.get('skip_download') and params.get('ignore_no_formats_error'):
121 continue
122 raise Exception('Test definition incorrect. The output file cannot be known. \'ext\' key is not present')
123
124 if 'skip' in test_case:
125 print_skipping(test_case['skip'])
126 return
127 for other_ie in other_ies:
128 if not other_ie.working():
129 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
130 return
131
132 params = get_params(test_case.get('params', {}))
133 params['outtmpl'] = tname + '_' + params['outtmpl']
134 if is_playlist and 'playlist' not in test_case:
135 params.setdefault('extract_flat', 'in_playlist')
136 params.setdefault('playlistend', test_case.get('playlist_mincount'))
137 params.setdefault('skip_download', True)
138
139 ydl = YoutubeDL(params, auto_init=False)
140 ydl.add_default_info_extractors()
141 finished_hook_called = set()
142
143 def _hook(status):
144 if status['status'] == 'finished':
145 finished_hook_called.add(status['filename'])
146 ydl.add_progress_hook(_hook)
147 expect_warnings(ydl, test_case.get('expected_warnings', []))
148
149 def get_tc_filename(tc):
150 return ydl.prepare_filename(dict(tc.get('info_dict', {})))
151
152 res_dict = None
153
154 def try_rm_tcs_files(tcs=None):
155 if tcs is None:
156 tcs = test_cases
157 for tc in tcs:
158 tc_filename = get_tc_filename(tc)
159 try_rm(tc_filename)
160 try_rm(tc_filename + '.part')
161 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
162 try_rm_tcs_files()
163 try:
164 try_num = 1
165 while True:
166 try:
167 # We're not using .download here since that is just a shim
168 # for outside error handling, and returns the exit code
169 # instead of the result dict.
170 res_dict = ydl.extract_info(
171 test_case['url'],
172 force_generic_extractor=params.get('force_generic_extractor', False))
173 except (DownloadError, ExtractorError) as err:
174 # Check if the exception is not a network related one
175 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):
176 raise
177
178 if try_num == RETRIES:
179 report_warning('%s failed due to network errors, skipping...' % tname)
180 return
181
182 print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
183
184 try_num += 1
185 else:
186 break
187
188 if is_playlist:
189 self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
190 self.assertTrue('entries' in res_dict)
191 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
192
193 if 'playlist_mincount' in test_case:
194 assertGreaterEqual(
195 self,
196 len(res_dict['entries']),
197 test_case['playlist_mincount'],
198 'Expected at least %d in playlist %s, but got only %d' % (
199 test_case['playlist_mincount'], test_case['url'],
200 len(res_dict['entries'])))
201 if 'playlist_count' in test_case:
202 self.assertEqual(
203 len(res_dict['entries']),
204 test_case['playlist_count'],
205 'Expected %d entries in playlist %s, but got %d.' % (
206 test_case['playlist_count'],
207 test_case['url'],
208 len(res_dict['entries']),
209 ))
210 if 'playlist_duration_sum' in test_case:
211 got_duration = sum(e['duration'] for e in res_dict['entries'])
212 self.assertEqual(
213 test_case['playlist_duration_sum'], got_duration)
214
215 # Generalize both playlists and single videos to unified format for
216 # simplicity
217 if 'entries' not in res_dict:
218 res_dict['entries'] = [res_dict]
219
220 for tc_num, tc in enumerate(test_cases):
221 tc_res_dict = res_dict['entries'][tc_num]
222 # First, check test cases' data against extracted data alone
223 expect_info_dict(self, tc_res_dict, tc.get('info_dict', {}))
224 # Now, check downloaded file consistency
225 tc_filename = get_tc_filename(tc)
226 if not test_case.get('params', {}).get('skip_download', False):
227 self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
228 self.assertTrue(tc_filename in finished_hook_called)
229 expected_minsize = tc.get('file_minsize', 10000)
230 if expected_minsize is not None:
231 if params.get('test'):
232 expected_minsize = max(expected_minsize, 10000)
233 got_fsize = os.path.getsize(tc_filename)
234 assertGreaterEqual(
235 self, got_fsize, expected_minsize,
236 'Expected %s to be at least %s, but it\'s only %s ' %
237 (tc_filename, format_bytes(expected_minsize),
238 format_bytes(got_fsize)))
239 if 'md5' in tc:
240 md5_for_file = _file_md5(tc_filename)
241 self.assertEqual(tc['md5'], md5_for_file)
242 # Finally, check test cases' data again but this time against
243 # extracted data from info JSON file written during processing
244 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
245 self.assertTrue(
246 os.path.exists(info_json_fn),
247 'Missing info file %s' % info_json_fn)
248 with io.open(info_json_fn, encoding='utf-8') as infof:
249 info_dict = json.load(infof)
250 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
251 finally:
252 try_rm_tcs_files()
253 if is_playlist and res_dict is not None and res_dict.get('entries'):
254 # Remove all other files that may have been extracted if the
255 # extractor returns full results even with extract_flat
256 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
257 try_rm_tcs_files(res_tcs)
258
259 return test_template
260
261
262 # And add them to TestDownload
263 tests_counter = {}
264 for test_case in defs:
265 name = test_case['name']
266 i = tests_counter.get(name, 0)
267 tests_counter[name] = i + 1
268 tname = f'test_{name}_{i}' if i else f'test_{name}'
269 test_method = generator(test_case, tname)
270 test_method.__name__ = str(tname)
271 ie_list = test_case.get('add_ie')
272 test_method.add_ie = ie_list and ','.join(ie_list)
273 setattr(TestDownload, test_method.__name__, test_method)
274 del test_method
275
276
277 def batch_generator(name, num_tests):
278
279 def test_template(self):
280 for i in range(num_tests):
281 getattr(self, f'test_{name}_{i}' if i else f'test_{name}')()
282
283 return test_template
284
285
286 for name, num_tests in tests_counter.items():
287 test_method = batch_generator(name, num_tests)
288 test_method.__name__ = f'test_{name}_all'
289 test_method.add_ie = ''
290 setattr(TestDownload, test_method.__name__, test_method)
291 del test_method
292
293
294 if __name__ == '__main__':
295 unittest.main()