]> Raphaƫl G. Git Repositories - youtubedl/blob - test/test_download.py
debian/patches: Add patch from upstream to fix extraction from youtube.
[youtubedl] / test / test_download.py
1 #!/usr/bin/env python
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_warnings,
14 get_params,
15 gettestcases,
16 expect_info_dict,
17 try_rm,
18 report_warning,
19 )
20
21
22 import hashlib
23 import io
24 import json
25 import socket
26
27 import youtube_dl.YoutubeDL
28 from youtube_dl.compat import (
29 compat_http_client,
30 compat_urllib_error,
31 compat_HTTPError,
32 )
33 from youtube_dl.utils import (
34 DownloadError,
35 ExtractorError,
36 format_bytes,
37 UnavailableVideoError,
38 )
39 from youtube_dl.extractor import get_info_extractor
40
41 RETRIES = 3
42
43
44 class YoutubeDL(youtube_dl.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)
49
50 def report_warning(self, message):
51 # Don't accept warnings during tests
52 raise ExtractorError(message)
53
54 def process_info(self, info_dict):
55 self.processed_info_dicts.append(info_dict)
56 return super(YoutubeDL, self).process_info(info_dict)
57
58
59 def _file_md5(fn):
60 with open(fn, 'rb') as f:
61 return hashlib.md5(f.read()).hexdigest()
62
63
64 defs = gettestcases()
65
66
67 class TestDownload(unittest.TestCase):
68 maxDiff = None
69
70 def setUp(self):
71 self.defs = defs
72
73 # Dynamically generate tests
74
75
76 def generator(test_case):
77
78 def test_template(self):
79 ie = youtube_dl.extractor.get_info_extractor(test_case['name'])
80 other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])]
81 is_playlist = any(k.startswith('playlist') for k in test_case)
82 test_cases = test_case.get(
83 'playlist', [] if is_playlist else [test_case])
84
85 def print_skipping(reason):
86 print('Skipping %s: %s' % (test_case['name'], reason))
87 if not ie.working():
88 print_skipping('IE marked as not _WORKING')
89 return
90
91 for tc in test_cases:
92 info_dict = tc.get('info_dict', {})
93 if not (info_dict.get('id') and info_dict.get('ext')):
94 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
95
96 if 'skip' in test_case:
97 print_skipping(test_case['skip'])
98 return
99 for other_ie in other_ies:
100 if not other_ie.working():
101 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
102 return
103
104 params = get_params(test_case.get('params', {}))
105 if is_playlist and 'playlist' not in test_case:
106 params.setdefault('extract_flat', 'in_playlist')
107 params.setdefault('skip_download', True)
108
109 ydl = YoutubeDL(params, auto_init=False)
110 ydl.add_default_info_extractors()
111 finished_hook_called = set()
112
113 def _hook(status):
114 if status['status'] == 'finished':
115 finished_hook_called.add(status['filename'])
116 ydl.add_progress_hook(_hook)
117 expect_warnings(ydl, test_case.get('expected_warnings', []))
118
119 def get_tc_filename(tc):
120 return ydl.prepare_filename(tc.get('info_dict', {}))
121
122 res_dict = None
123
124 def try_rm_tcs_files(tcs=None):
125 if tcs is None:
126 tcs = test_cases
127 for tc in tcs:
128 tc_filename = get_tc_filename(tc)
129 try_rm(tc_filename)
130 try_rm(tc_filename + '.part')
131 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
132 try_rm_tcs_files()
133 try:
134 try_num = 1
135 while True:
136 try:
137 # We're not using .download here sine that is just a shim
138 # for outside error handling, and returns the exit code
139 # instead of the result dict.
140 res_dict = ydl.extract_info(
141 test_case['url'],
142 force_generic_extractor=params.get('force_generic_extractor', False))
143 except (DownloadError, ExtractorError) as err:
144 # Check if the exception is not a network related one
145 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):
146 raise
147
148 if try_num == RETRIES:
149 report_warning('Failed due to network errors, skipping...')
150 return
151
152 print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
153
154 try_num += 1
155 else:
156 break
157
158 if is_playlist:
159 self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
160 self.assertTrue('entries' in res_dict)
161 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
162
163 if 'playlist_mincount' in test_case:
164 assertGreaterEqual(
165 self,
166 len(res_dict['entries']),
167 test_case['playlist_mincount'],
168 'Expected at least %d in playlist %s, but got only %d' % (
169 test_case['playlist_mincount'], test_case['url'],
170 len(res_dict['entries'])))
171 if 'playlist_count' in test_case:
172 self.assertEqual(
173 len(res_dict['entries']),
174 test_case['playlist_count'],
175 'Expected %d entries in playlist %s, but got %d.' % (
176 test_case['playlist_count'],
177 test_case['url'],
178 len(res_dict['entries']),
179 ))
180 if 'playlist_duration_sum' in test_case:
181 got_duration = sum(e['duration'] for e in res_dict['entries'])
182 self.assertEqual(
183 test_case['playlist_duration_sum'], got_duration)
184
185 for tc in test_cases:
186 tc_filename = get_tc_filename(tc)
187 if not test_case.get('params', {}).get('skip_download', False):
188 self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
189 self.assertTrue(tc_filename in finished_hook_called)
190 expected_minsize = tc.get('file_minsize', 10000)
191 if expected_minsize is not None:
192 if params.get('test'):
193 expected_minsize = max(expected_minsize, 10000)
194 got_fsize = os.path.getsize(tc_filename)
195 assertGreaterEqual(
196 self, got_fsize, expected_minsize,
197 'Expected %s to be at least %s, but it\'s only %s ' %
198 (tc_filename, format_bytes(expected_minsize),
199 format_bytes(got_fsize)))
200 if 'md5' in tc:
201 md5_for_file = _file_md5(tc_filename)
202 self.assertEqual(md5_for_file, tc['md5'])
203 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
204 self.assertTrue(
205 os.path.exists(info_json_fn),
206 'Missing info file %s' % info_json_fn)
207 with io.open(info_json_fn, encoding='utf-8') as infof:
208 info_dict = json.load(infof)
209
210 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
211 finally:
212 try_rm_tcs_files()
213 if is_playlist and res_dict is not None and res_dict.get('entries'):
214 # Remove all other files that may have been extracted if the
215 # extractor returns full results even with extract_flat
216 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
217 try_rm_tcs_files(res_tcs)
218
219 return test_template
220
221
222 # And add them to TestDownload
223 for n, test_case in enumerate(defs):
224 test_method = generator(test_case)
225 tname = 'test_' + str(test_case['name'])
226 i = 1
227 while hasattr(TestDownload, tname):
228 tname = 'test_%s_%d' % (test_case['name'], i)
229 i += 1
230 test_method.__name__ = str(tname)
231 setattr(TestDownload, test_method.__name__, test_method)
232 del test_method
233
234
235 if __name__ == '__main__':
236 unittest.main()