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 (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', 'in_playlist')
 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 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(
 140                         test_case['url'],
 141                         force_generic_extractor=params.get('force_generic_extractor', False))
 142                 except (DownloadError, ExtractorError) as err:
 143                     # Check if the exception is not a network related one
 144                     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):
 145                         raise
 146
 147                     if try_num == RETRIES:
 148                         report_warning('Failed due to network errors, skipping...')
 149                         return
 150
 151                     print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
 152
 153                     try_num += 1
 154                 else:
 155                     break
 156
 157             if is_playlist:
 158                 self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video'])
 159                 self.assertTrue('entries' in res_dict)
 160                 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
 161
 162             if 'playlist_mincount' in test_case:
 163                 assertGreaterEqual(
 164                     self,
 165                     len(res_dict['entries']),
 166                     test_case['playlist_mincount'],
 167                     'Expected at least %d in playlist %s, but got only %d' % (
 168                         test_case['playlist_mincount'], test_case['url'],
 169                         len(res_dict['entries'])))
 170             if 'playlist_count' in test_case:
 171                 self.assertEqual(
 172                     len(res_dict['entries']),
 173                     test_case['playlist_count'],
 174                     'Expected %d entries in playlist %s, but got %d.' % (
 175                         test_case['playlist_count'],
 176                         test_case['url'],
 177                         len(res_dict['entries']),
 178                     ))
 179             if 'playlist_duration_sum' in test_case:
 180                 got_duration = sum(e['duration'] for e in res_dict['entries'])
 181                 self.assertEqual(
 182                     test_case['playlist_duration_sum'], got_duration)
 183
 184             for tc in test_cases:
 185                 tc_filename = get_tc_filename(tc)
 186                 if not test_case.get('params', {}).get('skip_download', False):
 187                     self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
 188                     self.assertTrue(tc_filename in finished_hook_called)
 189                     expected_minsize = tc.get('file_minsize', 10000)
 190                     if expected_minsize is not None:
 191                         if params.get('test'):
 192                             expected_minsize = max(expected_minsize, 10000)
 193                         got_fsize = os.path.getsize(tc_filename)
 194                         assertGreaterEqual(
 195                             self, got_fsize, expected_minsize,
 196                             'Expected %s to be at least %s, but it\'s only %s ' %
 197                             (tc_filename, format_bytes(expected_minsize),
 198                                 format_bytes(got_fsize)))
 199                     if 'md5' in tc:
 200                         md5_for_file = _file_md5(tc_filename)
 201                         self.assertEqual(md5_for_file, tc['md5'])
 202                 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
 203                 self.assertTrue(
 204                     os.path.exists(info_json_fn),
 205                     'Missing info file %s' % info_json_fn)
 206                 with io.open(info_json_fn, encoding='utf-8') as infof:
 207                     info_dict = json.load(infof)
 208
 209                 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
 210         finally:
 211             try_rm_tcs_files()
 212             if is_playlist and res_dict is not None and res_dict.get('entries'):
 213                 # Remove all other files that may have been extracted if the
 214                 # extractor returns full results even with extract_flat
 215                 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
 216                 try_rm_tcs_files(res_tcs)
 217
 218     return test_template
 219
 220 # And add them to TestDownload
 221 for n, test_case in enumerate(defs):
 222     test_method = generator(test_case)
 223     tname = 'test_' + str(test_case['name'])
 224     i = 1
 225     while hasattr(TestDownload, tname):
 226         tname = 'test_%s_%d' % (test_case['name'], i)
 227         i += 1
 228     test_method.__name__ = str(tname)
 229     setattr(TestDownload, test_method.__name__, test_method)
 230     del test_method
 231
 232
 233 if __name__ == '__main__':
 234     unittest.main()