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     expect_value,
  15     get_params,
  16     gettestcases,
  17     expect_info_dict,
  18     try_rm,
  19     report_warning,
  20 )
  21
  22
  23 import hashlib
  24 import io
  25 import json
  26 import socket
  27
  28 import youtube_dl.YoutubeDL
  29 from youtube_dl.compat import (
  30     compat_http_client,
  31     compat_urllib_error,
  32     compat_HTTPError,
  33 )
  34 from youtube_dl.utils import (
  35     DownloadError,
  36     ExtractorError,
  37     format_bytes,
  38     UnavailableVideoError,
  39 )
  40 from youtube_dl.extractor import get_info_extractor
  41
  42 RETRIES = 3
  43
  44
  45 class YoutubeDL(youtube_dl.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)
  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 class TestDownload(unittest.TestCase):
  69     # Parallel testing in nosetests. See
  70     # http://nose.readthedocs.org/en/latest/doc_tests/test_multiprocess/multiprocess.html
  71     _multiprocess_shared_ = True
  72
  73     maxDiff = None
  74
  75     def __str__(self):
  76         """Identify each test with the `add_ie` attribute, if available."""
  77
  78         def strclass(cls):
  79             """From 2.7's unittest; 2.6 had _strclass so we can't import it."""
  80             return '%s.%s' % (cls.__module__, cls.__name__)
  81
  82         add_ie = getattr(self, self._testMethodName).add_ie
  83         return '%s (%s)%s:' % (self._testMethodName,
  84                                strclass(self.__class__),
  85                                ' [%s]' % add_ie if add_ie else '')
  86
  87     def setUp(self):
  88         self.defs = defs
  89
  90 # Dynamically generate tests
  91
  92
  93 def generator(test_case, tname):
  94
  95     def test_template(self):
  96         ie = youtube_dl.extractor.get_info_extractor(test_case['name'])
  97         other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])]
  98         is_playlist = any(k.startswith('playlist') for k in test_case)
  99         test_cases = test_case.get(
 100             'playlist', [] if is_playlist else [test_case])
 101
 102         def print_skipping(reason):
 103             print('Skipping %s: %s' % (test_case['name'], reason))
 104         if not ie.working():
 105             print_skipping('IE marked as not _WORKING')
 106             return
 107
 108         for tc in test_cases:
 109             info_dict = tc.get('info_dict', {})
 110             if not (info_dict.get('id') and info_dict.get('ext')):
 111                 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
 112
 113         if 'skip' in test_case:
 114             print_skipping(test_case['skip'])
 115             return
 116         for other_ie in other_ies:
 117             if not other_ie.working():
 118                 print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
 119                 return
 120
 121         params = get_params(test_case.get('params', {}))
 122         params['outtmpl'] = tname + '_' + params['outtmpl']
 123         if is_playlist and 'playlist' not in test_case:
 124             params.setdefault('extract_flat', 'in_playlist')
 125             params.setdefault('skip_download', True)
 126
 127         ydl = YoutubeDL(params, auto_init=False)
 128         ydl.add_default_info_extractors()
 129         finished_hook_called = set()
 130
 131         def _hook(status):
 132             if status['status'] == 'finished':
 133                 finished_hook_called.add(status['filename'])
 134         ydl.add_progress_hook(_hook)
 135         expect_warnings(ydl, test_case.get('expected_warnings', []))
 136
 137         def get_tc_filename(tc):
 138             return ydl.prepare_filename(tc.get('info_dict', {}))
 139
 140         res_dict = None
 141
 142         def try_rm_tcs_files(tcs=None):
 143             if tcs is None:
 144                 tcs = test_cases
 145             for tc in tcs:
 146                 tc_filename = get_tc_filename(tc)
 147                 try_rm(tc_filename)
 148                 try_rm(tc_filename + '.part')
 149                 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
 150         try_rm_tcs_files()
 151         try:
 152             try_num = 1
 153             while True:
 154                 try:
 155                     # We're not using .download here since that is just a shim
 156                     # for outside error handling, and returns the exit code
 157                     # instead of the result dict.
 158                     res_dict = ydl.extract_info(
 159                         test_case['url'],
 160                         force_generic_extractor=params.get('force_generic_extractor', False))
 161                 except (DownloadError, ExtractorError) as err:
 162                     # Check if the exception is not a network related one
 163                     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):
 164                         raise
 165
 166                     if try_num == RETRIES:
 167                         report_warning('%s failed due to network errors, skipping...' % tname)
 168                         return
 169
 170                     print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
 171
 172                     try_num += 1
 173                 else:
 174                     break
 175
 176             if is_playlist:
 177                 self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
 178                 self.assertTrue('entries' in res_dict)
 179                 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
 180
 181             if 'playlist_mincount' in test_case:
 182                 assertGreaterEqual(
 183                     self,
 184                     len(res_dict['entries']),
 185                     test_case['playlist_mincount'],
 186                     'Expected at least %d in playlist %s, but got only %d' % (
 187                         test_case['playlist_mincount'], test_case['url'],
 188                         len(res_dict['entries'])))
 189             if 'playlist_count' in test_case:
 190                 self.assertEqual(
 191                     len(res_dict['entries']),
 192                     test_case['playlist_count'],
 193                     'Expected %d entries in playlist %s, but got %d.' % (
 194                         test_case['playlist_count'],
 195                         test_case['url'],
 196                         len(res_dict['entries']),
 197                     ))
 198             if 'playlist_duration_sum' in test_case:
 199                 got_duration = sum(e['duration'] for e in res_dict['entries'])
 200                 self.assertEqual(
 201                     test_case['playlist_duration_sum'], got_duration)
 202
 203             for tc in test_cases:
 204                 expect_value(self, res_dict['id'], tc['info_dict']['id'], 'id')
 205                 tc_filename = get_tc_filename(tc)
 206                 if not test_case.get('params', {}).get('skip_download', False):
 207                     self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
 208                     self.assertTrue(tc_filename in finished_hook_called)
 209                     expected_minsize = tc.get('file_minsize', 10000)
 210                     if expected_minsize is not None:
 211                         if params.get('test'):
 212                             expected_minsize = max(expected_minsize, 10000)
 213                         got_fsize = os.path.getsize(tc_filename)
 214                         assertGreaterEqual(
 215                             self, got_fsize, expected_minsize,
 216                             'Expected %s to be at least %s, but it\'s only %s ' %
 217                             (tc_filename, format_bytes(expected_minsize),
 218                                 format_bytes(got_fsize)))
 219                     if 'md5' in tc:
 220                         md5_for_file = _file_md5(tc_filename)
 221                         self.assertEqual(md5_for_file, tc['md5'])
 222                 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
 223                 self.assertTrue(
 224                     os.path.exists(info_json_fn),
 225                     'Missing info file %s' % info_json_fn)
 226                 with io.open(info_json_fn, encoding='utf-8') as infof:
 227                     info_dict = json.load(infof)
 228
 229                 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
 230         finally:
 231             try_rm_tcs_files()
 232             if is_playlist and res_dict is not None and res_dict.get('entries'):
 233                 # Remove all other files that may have been extracted if the
 234                 # extractor returns full results even with extract_flat
 235                 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
 236                 try_rm_tcs_files(res_tcs)
 237
 238     return test_template
 239
 240
 241 # And add them to TestDownload
 242 for n, test_case in enumerate(defs):
 243     tname = 'test_' + str(test_case['name'])
 244     i = 1
 245     while hasattr(TestDownload, tname):
 246         tname = 'test_%s_%d' % (test_case['name'], i)
 247         i += 1
 248     test_method = generator(test_case, tname)
 249     test_method.__name__ = str(tname)
 250     ie_list = test_case.get('add_ie')
 251     test_method.add_ie = ie_list and ','.join(ie_list)
 252     setattr(TestDownload, test_method.__name__, test_method)
 253     del test_method
 254
 255
 256 if __name__ == '__main__':
 257     unittest.main()