Merge remote-tracking branch 'soult/br'
[youtube-dl] / test / test_download.py
1 #!/usr/bin/env python
2
3 # Allow direct execution
4 import os
5 import sys
6 import unittest
7 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
8
9 from test.helper import (
10     get_params,
11     get_testcases,
12     try_rm,
13     md5,
14     report_warning
15 )
16
17
18 import hashlib
19 import io
20 import json
21 import re
22 import socket
23
24 import youtube_dl.YoutubeDL
25 from youtube_dl.utils import (
26     compat_http_client,
27     compat_str,
28     compat_urllib_error,
29     compat_HTTPError,
30     DownloadError,
31     ExtractorError,
32     UnavailableVideoError,
33 )
34 from youtube_dl.extractor import get_info_extractor
35
36 RETRIES = 3
37
38 class YoutubeDL(youtube_dl.YoutubeDL):
39     def __init__(self, *args, **kwargs):
40         self.to_stderr = self.to_screen
41         self.processed_info_dicts = []
42         super(YoutubeDL, self).__init__(*args, **kwargs)
43     def report_warning(self, message):
44         # Don't accept warnings during tests
45         raise ExtractorError(message)
46     def process_info(self, info_dict):
47         self.processed_info_dicts.append(info_dict)
48         return super(YoutubeDL, self).process_info(info_dict)
49
50 def _file_md5(fn):
51     with open(fn, 'rb') as f:
52         return hashlib.md5(f.read()).hexdigest()
53
54 defs = get_testcases()
55
56
57 class TestDownload(unittest.TestCase):
58     maxDiff = None
59     def setUp(self):
60         self.defs = defs
61
62 ### Dynamically generate tests
63 def generator(test_case):
64
65     def test_template(self):
66         ie = youtube_dl.extractor.get_info_extractor(test_case['name'])
67         other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])]
68         def print_skipping(reason):
69             print('Skipping %s: %s' % (test_case['name'], reason))
70         if not ie.working():
71             print_skipping('IE marked as not _WORKING')
72             return
73         if 'playlist' not in test_case:
74             info_dict = test_case.get('info_dict', {})
75             if not test_case.get('file') and not (info_dict.get('id') and info_dict.get('ext')):
76                 print_skipping('The output file cannot be know, the "file" '
77                     'key is missing or the info_dict is incomplete')
78                 return
79         if 'skip' in test_case:
80             print_skipping(test_case['skip'])
81             return
82         for other_ie in other_ies:
83             if not other_ie.working():
84                 print_skipping(u'test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
85                 return
86
87         params = get_params(test_case.get('params', {}))
88
89         ydl = YoutubeDL(params)
90         ydl.add_default_info_extractors()
91         finished_hook_called = set()
92         def _hook(status):
93             if status['status'] == 'finished':
94                 finished_hook_called.add(status['filename'])
95         ydl.add_progress_hook(_hook)
96
97         def get_tc_filename(tc):
98             return tc.get('file') or ydl.prepare_filename(tc.get('info_dict', {}))
99
100         test_cases = test_case.get('playlist', [test_case])
101         def try_rm_tcs_files():
102             for tc in test_cases:
103                 tc_filename = get_tc_filename(tc)
104                 try_rm(tc_filename)
105                 try_rm(tc_filename + '.part')
106                 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
107         try_rm_tcs_files()
108         try:
109             try_num = 1
110             while True:
111                 try:
112                     ydl.download([test_case['url']])
113                 except (DownloadError, ExtractorError) as err:
114                     # Check if the exception is not a network related one
115                     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):
116                         raise
117
118                     if try_num == RETRIES:
119                         report_warning(u'Failed due to network errors, skipping...')
120                         return
121
122                     print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
123
124                     try_num += 1
125                 else:
126                     break
127
128             for tc in test_cases:
129                 tc_filename = get_tc_filename(tc)
130                 if not test_case.get('params', {}).get('skip_download', False):
131                     self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
132                     self.assertTrue(tc_filename in finished_hook_called)
133                 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
134                 self.assertTrue(os.path.exists(info_json_fn))
135                 if 'md5' in tc:
136                     md5_for_file = _file_md5(tc_filename)
137                     self.assertEqual(md5_for_file, tc['md5'])
138                 with io.open(info_json_fn, encoding='utf-8') as infof:
139                     info_dict = json.load(infof)
140                 for (info_field, expected) in tc.get('info_dict', {}).items():
141                     if isinstance(expected, compat_str) and expected.startswith('re:'):
142                         got = info_dict.get(info_field)
143                         match_str = expected[len('re:'):]
144                         match_rex = re.compile(match_str)
145
146                         self.assertTrue(
147                             isinstance(got, compat_str) and match_rex.match(got),
148                             u'field %s (value: %r) should match %r' % (info_field, got, match_str))
149                     else:
150                         if isinstance(expected, compat_str) and expected.startswith('md5:'):
151                             got = 'md5:' + md5(info_dict.get(info_field))
152                         else:
153                             got = info_dict.get(info_field)
154                         self.assertEqual(expected, got,
155                             u'invalid value for field %s, expected %r, got %r' % (info_field, expected, got))
156
157                 # If checkable fields are missing from the test case, print the info_dict
158                 test_info_dict = dict((key, value if not isinstance(value, compat_str) or len(value) < 250 else 'md5:' + md5(value))
159                     for key, value in info_dict.items()
160                     if value and key in ('title', 'description', 'uploader', 'upload_date', 'uploader_id', 'location'))
161                 if not all(key in tc.get('info_dict', {}).keys() for key in test_info_dict.keys()):
162                     sys.stderr.write(u'\n"info_dict": ' + json.dumps(test_info_dict, ensure_ascii=False, indent=4) + u'\n')
163
164                 # Check for the presence of mandatory fields
165                 for key in ('id', 'url', 'title', 'ext'):
166                     self.assertTrue(key in info_dict.keys() and info_dict[key])
167                 # Check for mandatory fields that are automatically set by YoutubeDL
168                 for key in ['webpage_url', 'extractor', 'extractor_key']:
169                     self.assertTrue(info_dict.get(key), u'Missing field: %s' % key)
170         finally:
171             try_rm_tcs_files()
172
173     return test_template
174
175 ### And add them to TestDownload
176 for n, test_case in enumerate(defs):
177     test_method = generator(test_case)
178     tname = 'test_' + str(test_case['name'])
179     i = 1
180     while hasattr(TestDownload, tname):
181         tname = 'test_'  + str(test_case['name']) + '_' + str(i)
182         i += 1
183     test_method.__name__ = tname
184     setattr(TestDownload, test_method.__name__, test_method)
185     del test_method
186
187
188 if __name__ == '__main__':
189     unittest.main()