git.bitcoin.ninja Git - youtube-dl/blob - 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     assertGreaterEqual,
  11     get_params,
  12     gettestcases,
  13     expect_info_dict,
  14     try_rm,
  15     report_warning,
  16 )
  17
  18
  19 import hashlib
  20 import io
  21 import json
  22 import socket
  23
  24 import youtube_dl.YoutubeDL
  25 from youtube_dl.utils import (
  26     compat_http_client,
  27     compat_urllib_error,
  28     compat_HTTPError,
  29     DownloadError,
  30     ExtractorError,
  31     format_bytes,
  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 = gettestcases()
  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         is_playlist = any(k.startswith('playlist') for k in test_case)
  69         test_cases = test_case.get(
  70             'playlist', [] if is_playlist else [test_case])
  71
  72         def print_skipping(reason):
  73             print('Skipping %s: %s' % (test_case['name'], reason))
  74         if not ie.working():
  75             print_skipping('IE marked as not _WORKING')
  76             return
  77
  78         for tc in test_cases:
  79             info_dict = tc.get('info_dict', {})
  80             if not tc.get('file') and not (info_dict.get('id') and info_dict.get('ext')):
  81                 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
  82
  83         if 'skip' in test_case:
  84             print_skipping(test_case['skip'])
  85             return
  86         for other_ie in other_ies:
  87             if not other_ie.working():
  88                 print_skipping(u'test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
  89                 return
  90
  91         params = get_params(test_case.get('params', {}))
  92         if is_playlist and 'playlist' not in test_case:
  93             params.setdefault('extract_flat', True)
  94             params.setdefault('skip_download', True)
  95
  96         ydl = YoutubeDL(params)
  97         ydl.add_default_info_extractors()
  98         finished_hook_called = set()
  99         def _hook(status):
 100             if status['status'] == 'finished':
 101                 finished_hook_called.add(status['filename'])
 102         ydl.add_progress_hook(_hook)
 103
 104         def get_tc_filename(tc):
 105             return tc.get('file') or ydl.prepare_filename(tc.get('info_dict', {}))
 106
 107         res_dict = None
 108         def try_rm_tcs_files(tcs=None):
 109             if tcs is None:
 110                 tcs = test_cases
 111             for tc in tcs:
 112                 tc_filename = get_tc_filename(tc)
 113                 try_rm(tc_filename)
 114                 try_rm(tc_filename + '.part')
 115                 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
 116         try_rm_tcs_files()
 117         try:
 118             try_num = 1
 119             while True:
 120                 try:
 121                     # We're not using .download here sine that is just a shim
 122                     # for outside error handling, and returns the exit code
 123                     # instead of the result dict.
 124                     res_dict = ydl.extract_info(test_case['url'])
 125                 except (DownloadError, ExtractorError) as err:
 126                     # Check if the exception is not a network related one
 127                     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):
 128                         raise
 129
 130                     if try_num == RETRIES:
 131                         report_warning(u'Failed due to network errors, skipping...')
 132                         return
 133
 134                     print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
 135
 136                     try_num += 1
 137                 else:
 138                     break
 139
 140             if is_playlist:
 141                 self.assertEqual(res_dict['_type'], 'playlist')
 142                 expect_info_dict(self, test_case.get('info_dict', {}), res_dict)
 143             if 'playlist_mincount' in test_case:
 144                 assertGreaterEqual(
 145                     self,
 146                     len(res_dict['entries']),
 147                     test_case['playlist_mincount'],
 148                     'Expected at least %d in playlist %s, but got only %d' % (
 149                         test_case['playlist_mincount'], test_case['url'],
 150                         len(res_dict['entries'])))
 151             if 'playlist_count' in test_case:
 152                 self.assertEqual(
 153                     len(res_dict['entries']),
 154                     test_case['playlist_count'],
 155                     'Expected %d entries in playlist %s, but got %d.' % (
 156                         test_case['playlist_count'],
 157                         test_case['url'],
 158                         len(res_dict['entries']),
 159                     ))
 160             if 'playlist_duration_sum' in test_case:
 161                 got_duration = sum(e['duration'] for e in res_dict['entries'])
 162                 self.assertEqual(
 163                     test_case['playlist_duration_sum'], got_duration)
 164
 165             for tc in test_cases:
 166                 tc_filename = get_tc_filename(tc)
 167                 if not test_case.get('params', {}).get('skip_download', False):
 168                     self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
 169                     self.assertTrue(tc_filename in finished_hook_called)
 170                     expected_minsize = tc.get('file_minsize', 10000)
 171                     if expected_minsize is not None:
 172                         if params.get('test'):
 173                             expected_minsize = max(expected_minsize, 10000)
 174                         got_fsize = os.path.getsize(tc_filename)
 175                         assertGreaterEqual(
 176                             self, got_fsize, expected_minsize,
 177                             'Expected %s to be at least %s, but it\'s only %s ' %
 178                             (tc_filename, format_bytes(expected_minsize),
 179                                 format_bytes(got_fsize)))
 180                     if 'md5' in tc:
 181                         md5_for_file = _file_md5(tc_filename)
 182                         self.assertEqual(md5_for_file, tc['md5'])
 183                 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
 184                 self.assertTrue(os.path.exists(info_json_fn))
 185                 with io.open(info_json_fn, encoding='utf-8') as infof:
 186                     info_dict = json.load(infof)
 187
 188                 expect_info_dict(self, tc.get('info_dict', {}), info_dict)
 189         finally:
 190             try_rm_tcs_files()
 191             if is_playlist and res_dict is not None:
 192                 # Remove all other files that may have been extracted if the
 193                 # extractor returns full results even with extract_flat
 194                 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
 195                 try_rm_tcs_files(res_tcs)
 196
 197     return test_template
 198
 199 ### And add them to TestDownload
 200 for n, test_case in enumerate(defs):
 201     test_method = generator(test_case)
 202     tname = 'test_' + str(test_case['name'])
 203     i = 1
 204     while hasattr(TestDownload, tname):
 205         tname = 'test_'  + str(test_case['name']) + '_' + str(i)
 206         i += 1
 207     test_method.__name__ = tname
 208     setattr(TestDownload, test_method.__name__, test_method)
 209     del test_method
 210
 211
 212 if __name__ == '__main__':
 213     unittest.main()