]> Raphaƫl G. Git Repositories - youtubedl/blob - test/test_download.py
a009aa475442ae588405a99f432af6feade92836
[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 defs = gettestcases()
64
65
66 class TestDownload(unittest.TestCase):
67 maxDiff = None
68
69 def setUp(self):
70 self.defs = defs
71
72 # Dynamically generate tests
73
74
75 def generator(test_case):
76
77 def test_template(self):
78 ie = youtube_dl.extractor.get_info_extractor(test_case['name'])
79 other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])]
80 is_playlist = any(k.startswith('playlist') for k in test_case)
81 test_cases = test_case.get(
82 'playlist', [] if is_playlist else [test_case])
83
84 def print_skipping(reason):
85 print('Skipping %s: %s' % (test_case['name'], reason))
86 if not ie.working():
87 print_skipping('IE marked as not _WORKING')
88 return
89
90 for tc in test_cases:
91 info_dict = tc.get('info_dict', {})
92 if not tc.get('file') and not (info_dict.get('id') and info_dict.get('ext')):
93 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
94
95 if 'skip' in test_case:
96 print_skipping(test_case['skip'])
97 return
98 for other_ie in other_ies:
99 if not other_ie.working():
100 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
101 return
102
103 params = get_params(test_case.get('params', {}))
104 if is_playlist and 'playlist' not in test_case:
105 params.setdefault('extract_flat', True)
106 params.setdefault('skip_download', True)
107
108 ydl = YoutubeDL(params, auto_init=False)
109 ydl.add_default_info_extractors()
110 finished_hook_called = set()
111
112 def _hook(status):
113 if status['status'] == 'finished':
114 finished_hook_called.add(status['filename'])
115 ydl.add_progress_hook(_hook)
116 expect_warnings(ydl, test_case.get('expected_warnings', []))
117
118 def get_tc_filename(tc):
119 return tc.get('file') or ydl.prepare_filename(tc.get('info_dict', {}))
120
121 res_dict = None
122
123 def try_rm_tcs_files(tcs=None):
124 if tcs is None:
125 tcs = test_cases
126 for tc in tcs:
127 tc_filename = get_tc_filename(tc)
128 try_rm(tc_filename)
129 try_rm(tc_filename + '.part')
130 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
131 try_rm_tcs_files()
132 try:
133 try_num = 1
134 while True:
135 try:
136 # We're not using .download here sine that is just a shim
137 # for outside error handling, and returns the exit code
138 # instead of the result dict.
139 res_dict = ydl.extract_info(test_case['url'])
140 except (DownloadError, ExtractorError) as err:
141 # Check if the exception is not a network related one
142 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):
143 raise
144
145 if try_num == RETRIES:
146 report_warning('Failed due to network errors, skipping...')
147 return
148
149 print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
150
151 try_num += 1
152 else:
153 break
154
155 if is_playlist:
156 self.assertEqual(res_dict['_type'], 'playlist')
157 self.assertTrue('entries' in res_dict)
158 expect_info_dict(self, test_case.get('info_dict', {}), res_dict)
159
160 if 'playlist_mincount' in test_case:
161 assertGreaterEqual(
162 self,
163 len(res_dict['entries']),
164 test_case['playlist_mincount'],
165 'Expected at least %d in playlist %s, but got only %d' % (
166 test_case['playlist_mincount'], test_case['url'],
167 len(res_dict['entries'])))
168 if 'playlist_count' in test_case:
169 self.assertEqual(
170 len(res_dict['entries']),
171 test_case['playlist_count'],
172 'Expected %d entries in playlist %s, but got %d.' % (
173 test_case['playlist_count'],
174 test_case['url'],
175 len(res_dict['entries']),
176 ))
177 if 'playlist_duration_sum' in test_case:
178 got_duration = sum(e['duration'] for e in res_dict['entries'])
179 self.assertEqual(
180 test_case['playlist_duration_sum'], got_duration)
181
182 for tc in test_cases:
183 tc_filename = get_tc_filename(tc)
184 if not test_case.get('params', {}).get('skip_download', False):
185 self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
186 self.assertTrue(tc_filename in finished_hook_called)
187 expected_minsize = tc.get('file_minsize', 10000)
188 if expected_minsize is not None:
189 if params.get('test'):
190 expected_minsize = max(expected_minsize, 10000)
191 got_fsize = os.path.getsize(tc_filename)
192 assertGreaterEqual(
193 self, got_fsize, expected_minsize,
194 'Expected %s to be at least %s, but it\'s only %s ' %
195 (tc_filename, format_bytes(expected_minsize),
196 format_bytes(got_fsize)))
197 if 'md5' in tc:
198 md5_for_file = _file_md5(tc_filename)
199 self.assertEqual(md5_for_file, tc['md5'])
200 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
201 self.assertTrue(
202 os.path.exists(info_json_fn),
203 'Missing info file %s' % info_json_fn)
204 with io.open(info_json_fn, encoding='utf-8') as infof:
205 info_dict = json.load(infof)
206
207 expect_info_dict(self, tc.get('info_dict', {}), info_dict)
208 finally:
209 try_rm_tcs_files()
210 if is_playlist and res_dict is not None and res_dict.get('entries'):
211 # Remove all other files that may have been extracted if the
212 # extractor returns full results even with extract_flat
213 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
214 try_rm_tcs_files(res_tcs)
215
216 return test_template
217
218 # And add them to TestDownload
219 for n, test_case in enumerate(defs):
220 test_method = generator(test_case)
221 tname = 'test_' + str(test_case['name'])
222 i = 1
223 while hasattr(TestDownload, tname):
224 tname = 'test_%s_%d' % (test_case['name'], i)
225 i += 1
226 test_method.__name__ = str(tname)
227 setattr(TestDownload, test_method.__name__, test_method)
228 del test_method
229
230
231 if __name__ == '__main__':
232 unittest.main()