Raphaël G. Git Repositories - youtubedl/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     expect_warnings,
  12     get_params,
  13     gettestcases,
  14     expect_info_dict,
  15     try_rm,
  16     report_warning,
  17 )
  18
  19
  20 import hashlib
  21 import io
  22 import json
  23 import socket
  24
  25 import youtube_dl.YoutubeDL
  26 from youtube_dl.utils import (
  27     compat_http_client,
  28     compat_urllib_error,
  29     compat_HTTPError,
  30     DownloadError,
  31     ExtractorError,
  32     format_bytes,
  33     UnavailableVideoError,
  34 )
  35 from youtube_dl.extractor import get_info_extractor
  36
  37 RETRIES = 3
  38
  39 class YoutubeDL(youtube_dl.YoutubeDL):
  40     def __init__(self, *args, **kwargs):
  41         self.to_stderr = self.to_screen
  42         self.processed_info_dicts = []
  43         super(YoutubeDL, self).__init__(*args, **kwargs)
  44     def report_warning(self, message):
  45         # Don't accept warnings during tests
  46         raise ExtractorError(message)
  47     def process_info(self, info_dict):
  48         self.processed_info_dicts.append(info_dict)
  49         return super(YoutubeDL, self).process_info(info_dict)
  50
  51 def _file_md5(fn):
  52     with open(fn, 'rb') as f:
  53         return hashlib.md5(f.read()).hexdigest()
  54
  55 defs = gettestcases()
  56
  57
  58 class TestDownload(unittest.TestCase):
  59     maxDiff = None
  60     def setUp(self):
  61         self.defs = defs
  62
  63 ### Dynamically generate tests
  64 def generator(test_case):
  65
  66     def test_template(self):
  67         ie = youtube_dl.extractor.get_info_extractor(test_case['name'])
  68         other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])]
  69         is_playlist = any(k.startswith('playlist') for k in test_case)
  70         test_cases = test_case.get(
  71             'playlist', [] if is_playlist else [test_case])
  72
  73         def print_skipping(reason):
  74             print('Skipping %s: %s' % (test_case['name'], reason))
  75         if not ie.working():
  76             print_skipping('IE marked as not _WORKING')
  77             return
  78
  79         for tc in test_cases:
  80             info_dict = tc.get('info_dict', {})
  81             if not tc.get('file') and not (info_dict.get('id') and info_dict.get('ext')):
  82                 raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?')
  83
  84         if 'skip' in test_case:
  85             print_skipping(test_case['skip'])
  86             return
  87         for other_ie in other_ies:
  88             if not other_ie.working():
  89                 print_skipping(u'test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
  90                 return
  91
  92         params = get_params(test_case.get('params', {}))
  93         if is_playlist and 'playlist' not in test_case:
  94             params.setdefault('extract_flat', True)
  95             params.setdefault('skip_download', True)
  96
  97         ydl = YoutubeDL(params)
  98         ydl.add_default_info_extractors()
  99         finished_hook_called = set()
 100         def _hook(status):
 101             if status['status'] == 'finished':
 102                 finished_hook_called.add(status['filename'])
 103         ydl.add_progress_hook(_hook)
 104         expect_warnings(ydl, test_case.get('expected_warnings', []))
 105
 106         def get_tc_filename(tc):
 107             return tc.get('file') or ydl.prepare_filename(tc.get('info_dict', {}))
 108
 109         res_dict = None
 110         def try_rm_tcs_files(tcs=None):
 111             if tcs is None:
 112                 tcs = test_cases
 113             for tc in tcs:
 114                 tc_filename = get_tc_filename(tc)
 115                 try_rm(tc_filename)
 116                 try_rm(tc_filename + '.part')
 117                 try_rm(os.path.splitext(tc_filename)[0] + '.info.json')
 118         try_rm_tcs_files()
 119         try:
 120             try_num = 1
 121             while True:
 122                 try:
 123                     # We're not using .download here sine that is just a shim
 124                     # for outside error handling, and returns the exit code
 125                     # instead of the result dict.
 126                     res_dict = ydl.extract_info(test_case['url'])
 127                 except (DownloadError, ExtractorError) as err:
 128                     # Check if the exception is not a network related one
 129                     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):
 130                         raise
 131
 132                     if try_num == RETRIES:
 133                         report_warning(u'Failed due to network errors, skipping...')
 134                         return
 135
 136                     print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
 137
 138                     try_num += 1
 139                 else:
 140                     break
 141
 142             if is_playlist:
 143                 self.assertEqual(res_dict['_type'], 'playlist')
 144                 self.assertTrue('entries' in res_dict)
 145                 expect_info_dict(self, test_case.get('info_dict', {}), res_dict)
 146
 147             if 'playlist_mincount' in test_case:
 148                 assertGreaterEqual(
 149                     self,
 150                     len(res_dict['entries']),
 151                     test_case['playlist_mincount'],
 152                     'Expected at least %d in playlist %s, but got only %d' % (
 153                         test_case['playlist_mincount'], test_case['url'],
 154                         len(res_dict['entries'])))
 155             if 'playlist_count' in test_case:
 156                 self.assertEqual(
 157                     len(res_dict['entries']),
 158                     test_case['playlist_count'],
 159                     'Expected %d entries in playlist %s, but got %d.' % (
 160                         test_case['playlist_count'],
 161                         test_case['url'],
 162                         len(res_dict['entries']),
 163                     ))
 164             if 'playlist_duration_sum' in test_case:
 165                 got_duration = sum(e['duration'] for e in res_dict['entries'])
 166                 self.assertEqual(
 167                     test_case['playlist_duration_sum'], got_duration)
 168
 169             for tc in test_cases:
 170                 tc_filename = get_tc_filename(tc)
 171                 if not test_case.get('params', {}).get('skip_download', False):
 172                     self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename)
 173                     self.assertTrue(tc_filename in finished_hook_called)
 174                     expected_minsize = tc.get('file_minsize', 10000)
 175                     if expected_minsize is not None:
 176                         if params.get('test'):
 177                             expected_minsize = max(expected_minsize, 10000)
 178                         got_fsize = os.path.getsize(tc_filename)
 179                         assertGreaterEqual(
 180                             self, got_fsize, expected_minsize,
 181                             'Expected %s to be at least %s, but it\'s only %s ' %
 182                             (tc_filename, format_bytes(expected_minsize),
 183                                 format_bytes(got_fsize)))
 184                     if 'md5' in tc:
 185                         md5_for_file = _file_md5(tc_filename)
 186                         self.assertEqual(md5_for_file, tc['md5'])
 187                 info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json'
 188                 self.assertTrue(
 189                     os.path.exists(info_json_fn),
 190                     'Missing info file %s' % info_json_fn)
 191                 with io.open(info_json_fn, encoding='utf-8') as infof:
 192                     info_dict = json.load(infof)
 193
 194                 expect_info_dict(self, tc.get('info_dict', {}), info_dict)
 195         finally:
 196             try_rm_tcs_files()
 197             if is_playlist and res_dict is not None and res_dict.get('entries'):
 198                 # Remove all other files that may have been extracted if the
 199                 # extractor returns full results even with extract_flat
 200                 res_tcs = [{'info_dict': e} for e in res_dict['entries']]
 201                 try_rm_tcs_files(res_tcs)
 202
 203     return test_template
 204
 205 ### And add them to TestDownload
 206 for n, test_case in enumerate(defs):
 207     test_method = generator(test_case)
 208     tname = 'test_' + str(test_case['name'])
 209     i = 1
 210     while hasattr(TestDownload, tname):
 211         tname = 'test_'  + str(test_case['name']) + '_' + str(i)
 212         i += 1
 213     test_method.__name__ = tname
 214     setattr(TestDownload, test_method.__name__, test_method)
 215     del test_method
 216
 217
 218 if __name__ == '__main__':
 219     unittest.main()