]> Raphaƫl G. Git Repositories - youtubedl/blob - test/test_download.py
Imported Upstream version 2014.11.21
[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 class YoutubeDL(youtube_dl.YoutubeDL):
44 def __init__(self, *args, **kwargs):
45 self.to_stderr = self.to_screen
46 self.processed_info_dicts = []
47 super(YoutubeDL, self).__init__(*args, **kwargs)
48 def report_warning(self, message):
49 # Don't accept warnings during tests
50 raise ExtractorError(message)
51 def process_info(self, info_dict):
52 self.processed_info_dicts.append(info_dict)
53 return super(YoutubeDL, self).process_info(info_dict)
54
55 def _file_md5(fn):
56 with open(fn, 'rb') as f:
57 return hashlib.md5(f.read()).hexdigest()
58
59 defs = gettestcases()
60
61
62 class TestDownload(unittest.TestCase):
63 maxDiff = None
64 def setUp(self):
65 self.defs = defs
66
67 ### Dynamically generate tests
68 def generator(test_case):
69
70 def test_template(self):
71 ie = youtube_dl.extractor.get_info_extractor(test_case['name'])
72 other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])]
73 is_playlist = any(k.startswith('playlist') for k in test_case)
74 test_cases = test_case.get(
75 'playlist', [] if is_playlist else [test_case])
76
77 def print_skipping(reason):
78 print('Skipping %s: %s' % (test_case['name'], reason))
79 if not ie.working():
80 print_skipping('IE marked as not _WORKING')
81 return
82
83 for tc in test_cases:
84 info_dict = tc.get('info_dict', {})
85 if not tc.get('file') and not (info_dict.get('id') and info_dict.get('ext')):
86 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
87
88 if 'skip' in test_case:
89 print_skipping(test_case['skip'])
90 return
91 for other_ie in other_ies:
92 if not other_ie.working():
93 print_skipping(u'test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
94 return
95
96 params = get_params(test_case.get('params', {}))
97 if is_playlist and 'playlist' not in test_case:
98 params.setdefault('extract_flat', True)
99 params.setdefault('skip_download', True)
100
101 ydl = YoutubeDL(params, auto_init=False)
102 ydl.add_default_info_extractors()
103 finished_hook_called = set()
104 def _hook(status):
105 if status['status'] == 'finished':
106 finished_hook_called.add(status['filename'])
107 ydl.add_progress_hook(_hook)
108 expect_warnings(ydl, test_case.get('expected_warnings', []))
109
110 def get_tc_filename(tc):
111 return tc.get('file') or ydl.prepare_filename(tc.get('info_dict', {}))
112
113 res_dict = None
114 def try_rm_tcs_files(tcs=None):
115 if tcs is None:
116 tcs = test_cases
117 for tc in tcs:
118 tc_filename = get_tc_filename(tc)
119 try_rm(tc_filename)
120 try_rm(tc_filename + '.part')
121 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
122 try_rm_tcs_files()
123 try:
124 try_num = 1
125 while True:
126 try:
127 # We're not using .download here sine that is just a shim
128 # for outside error handling, and returns the exit code
129 # instead of the result dict.
130 res_dict = ydl.extract_info(test_case['url'])
131 except (DownloadError, ExtractorError) as err:
132 # Check if the exception is not a network related one
133 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):
134 raise
135
136 if try_num == RETRIES:
137 report_warning(u'Failed due to network errors, skipping...')
138 return
139
140 print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
141
142 try_num += 1
143 else:
144 break
145
146 if is_playlist:
147 self.assertEqual(res_dict['_type'], 'playlist')
148 self.assertTrue('entries' in res_dict)
149 expect_info_dict(self, test_case.get('info_dict', {}), res_dict)
150
151 if 'playlist_mincount' in test_case:
152 assertGreaterEqual(
153 self,
154 len(res_dict['entries']),
155 test_case['playlist_mincount'],
156 'Expected at least %d in playlist %s, but got only %d' % (
157 test_case['playlist_mincount'], test_case['url'],
158 len(res_dict['entries'])))
159 if 'playlist_count' in test_case:
160 self.assertEqual(
161 len(res_dict['entries']),
162 test_case['playlist_count'],
163 'Expected %d entries in playlist %s, but got %d.' % (
164 test_case['playlist_count'],
165 test_case['url'],
166 len(res_dict['entries']),
167 ))
168 if 'playlist_duration_sum' in test_case:
169 got_duration = sum(e['duration'] for e in res_dict['entries'])
170 self.assertEqual(
171 test_case['playlist_duration_sum'], got_duration)
172
173 for tc in test_cases:
174 tc_filename = get_tc_filename(tc)
175 if not test_case.get('params', {}).get('skip_download', False):
176 self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
177 self.assertTrue(tc_filename in finished_hook_called)
178 expected_minsize = tc.get('file_minsize', 10000)
179 if expected_minsize is not None:
180 if params.get('test'):
181 expected_minsize = max(expected_minsize, 10000)
182 got_fsize = os.path.getsize(tc_filename)
183 assertGreaterEqual(
184 self, got_fsize, expected_minsize,
185 'Expected %s to be at least %s, but it\'s only %s ' %
186 (tc_filename, format_bytes(expected_minsize),
187 format_bytes(got_fsize)))
188 if 'md5' in tc:
189 md5_for_file = _file_md5(tc_filename)
190 self.assertEqual(md5_for_file, tc['md5'])
191 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
192 self.assertTrue(
193 os.path.exists(info_json_fn),
194 'Missing info file %s' % info_json_fn)
195 with io.open(info_json_fn, encoding='utf-8') as infof:
196 info_dict = json.load(infof)
197
198 expect_info_dict(self, tc.get('info_dict', {}), info_dict)
199 finally:
200 try_rm_tcs_files()
201 if is_playlist and res_dict is not None and res_dict.get('entries'):
202 # Remove all other files that may have been extracted if the
203 # extractor returns full results even with extract_flat
204 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
205 try_rm_tcs_files(res_tcs)
206
207 return test_template
208
209 ### And add them to TestDownload
210 for n, test_case in enumerate(defs):
211 test_method = generator(test_case)
212 tname = 'test_' + str(test_case['name'])
213 i = 1
214 while hasattr(TestDownload, tname):
215 tname = 'test_%s_%d' % (test_case['name'], i)
216 i += 1
217 test_method.__name__ = str(tname)
218 setattr(TestDownload, test_method.__name__, test_method)
219 del test_method
220
221
222 if __name__ == '__main__':
223 unittest.main()