X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=test%2Ftest_download.py;h=8b9b833735990ce2655fa7467488d95951f079ae;hb=8936f68a0ba3284c88ec619fb4cc22eb0499e7f3;hp=d1d6b119b9290fb60f1ce56fa68ff1e67ce034d9;hpb=1a9c655e3b1569f315d4193e877cba0b4a863c63;p=youtube-dl diff --git a/test/test_download.py b/test/test_download.py index d1d6b119b..30034f978 100644 --- a/test/test_download.py +++ b/test/test_download.py @@ -1,198 +1,241 @@ -#!/usr/bin/env python2 +#!/usr/bin/env python + +from __future__ import unicode_literals + +# Allow direct execution +import os +import sys import unittest +sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) + +from test.helper import ( + assertGreaterEqual, + expect_warnings, + get_params, + gettestcases, + expect_info_dict, + try_rm, + report_warning, +) + + import hashlib -import os +import io import json +import socket -from youtube_dl.FileDownloader import FileDownloader -from youtube_dl.InfoExtractors import YoutubeIE, DailymotionIE -from youtube_dl.InfoExtractors import MetacafeIE, BlipTVIE -from youtube_dl.InfoExtractors import XVideosIE, VimeoIE -from youtube_dl.InfoExtractors import SoundcloudIE, StanfordOpenClassroomIE -from youtube_dl.InfoExtractors import CollegeHumorIE, XNXXIE - - -class DownloadTest(unittest.TestCase): - PARAMETERS_FILE = "test/parameters.json" - #calculated with md5sum: - #md5sum (GNU coreutils) 8.19 - - YOUTUBE_SIZE = 1993883 - YOUTUBE_URL = "http://www.youtube.com/watch?v=BaW_jenozKc" - YOUTUBE_FILE = "BaW_jenozKc.mp4" - - DAILYMOTION_MD5 = "d363a50e9eb4f22ce90d08d15695bb47" - DAILYMOTION_URL = "http://www.dailymotion.com/video/x33vw9_tutoriel-de-youtubeur-dl-des-video_tech" - DAILYMOTION_FILE = "x33vw9.mp4" - - METACAFE_SIZE = 5754305 - METACAFE_URL = "http://www.metacafe.com/watch/yt-_aUehQsCQtM/the_electric_company_short_i_pbs_kids_go/" - METACAFE_FILE = "_aUehQsCQtM.flv" - - BLIP_MD5 = "93c24d2f4e0782af13b8a7606ea97ba7" - BLIP_URL = "http://blip.tv/cbr/cbr-exclusive-gotham-city-imposters-bats-vs-jokerz-short-3-5796352" - BLIP_FILE = "5779306.m4v" - - XVIDEO_MD5 = "1ab4dedc01f771cb2a65e91caa801aaf" - XVIDEO_URL = "http://www.xvideos.com/video939581/funny_porns_by_s_-1" - XVIDEO_FILE = "939581.flv" - - VIMEO_MD5 = "1ab4dedc01f771cb2a65e91caa801aaf" - VIMEO_URL = "http://vimeo.com/14160053" - VIMEO_FILE = "" - - VIMEO2_MD5 = "" - VIMEO2_URL = "http://player.vimeo.com/video/47019590" - VIMEO2_FILE = "" - - SOUNDCLOUD_MD5 = "ce3775768ebb6432fa8495d446a078ed" - SOUNDCLOUD_URL = "http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy" - SOUNDCLOUD_FILE = "n6FLbx6ZzMiu.mp3" - - STANDFORD_MD5 = "22c8206291368c4e2c9c1a307f0ea0f4" - STANDFORD_URL = "http://openclassroom.stanford.edu/MainFolder/VideoPage.php?course=PracticalUnix&video=intro-environment&speed=100" - STANDFORD_FILE = "PracticalUnix_intro-environment.mp4" - - COLLEGEHUMOR_MD5 = "" - COLLEGEHUMOR_URL = "http://www.collegehumor.com/video/6830834/mitt-romney-style-gangnam-style-parody" - COLLEGEHUMOR_FILE = "" - - XNXX_MD5 = "5f0469c8d1dfd1bc38c8e6deb5e0a21d" - XNXX_URL = "http://video.xnxx.com/video1135332/lida_naked_funny_actress_5_" - XNXX_FILE = "1135332.flv" - - def test_youtube(self): - #let's download a file from youtube - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(YoutubeIE()) - fd.download([DownloadTest.YOUTUBE_URL]) - self.assertTrue(os.path.exists(DownloadTest.YOUTUBE_FILE)) - self.assertEqual(os.path.getsize(DownloadTest.YOUTUBE_FILE), DownloadTest.YOUTUBE_SIZE) - - def test_dailymotion(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(DailymotionIE()) - fd.download([DownloadTest.DAILYMOTION_URL]) - self.assertTrue(os.path.exists(DownloadTest.DAILYMOTION_FILE)) - md5_down_file = md5_for_file(DownloadTest.DAILYMOTION_FILE) - self.assertEqual(md5_down_file, DownloadTest.DAILYMOTION_MD5) - - def test_metacafe(self): - #this emulate a skip,to be 2.6 compatible - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(MetacafeIE()) - fd.add_info_extractor(YoutubeIE()) - fd.download([DownloadTest.METACAFE_URL]) - self.assertTrue(os.path.exists(DownloadTest.METACAFE_FILE)) - self.assertEqual(os.path.getsize(DownloadTest.METACAFE_FILE), DownloadTest.METACAFE_SIZE) - - def test_blip(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(BlipTVIE()) - fd.download([DownloadTest.BLIP_URL]) - self.assertTrue(os.path.exists(DownloadTest.BLIP_FILE)) - md5_down_file = md5_for_file(DownloadTest.BLIP_FILE) - self.assertEqual(md5_down_file, DownloadTest.BLIP_MD5) - - def test_xvideo(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(XVideosIE()) - fd.download([DownloadTest.XVIDEO_URL]) - self.assertTrue(os.path.exists(DownloadTest.XVIDEO_FILE)) - md5_down_file = md5_for_file(DownloadTest.XVIDEO_FILE) - self.assertEqual(md5_down_file, DownloadTest.XVIDEO_MD5) - - def test_vimeo(self): - #skipped for the moment produce an error - return - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(VimeoIE()) - fd.download([DownloadTest.VIMEO_URL]) - self.assertTrue(os.path.exists(DownloadTest.VIMEO_FILE)) - md5_down_file = md5_for_file(DownloadTest.VIMEO_FILE) - self.assertEqual(md5_down_file, DownloadTest.VIMEO_MD5) - - def test_vimeo2(self): - #skipped for the moment produce an error - return - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(VimeoIE()) - fd.download([DownloadTest.VIMEO2_URL]) - self.assertTrue(os.path.exists(DownloadTest.VIMEO2_FILE)) - md5_down_file = md5_for_file(DownloadTest.VIMEO2_FILE) - self.assertEqual(md5_down_file, DownloadTest.VIMEO2_MD5) - - def test_soundcloud(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(SoundcloudIE()) - fd.download([DownloadTest.SOUNDCLOUD_URL]) - self.assertTrue(os.path.exists(DownloadTest.SOUNDCLOUD_FILE)) - md5_down_file = md5_for_file(DownloadTest.SOUNDCLOUD_FILE) - self.assertEqual(md5_down_file, DownloadTest.SOUNDCLOUD_MD5) - - def test_standford(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(StanfordOpenClassroomIE()) - fd.download([DownloadTest.STANDFORD_URL]) - self.assertTrue(os.path.exists(DownloadTest.STANDFORD_FILE)) - md5_down_file = md5_for_file(DownloadTest.STANDFORD_FILE) - self.assertEqual(md5_down_file, DownloadTest.STANDFORD_MD5) - - def test_collegehumor(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(CollegeHumorIE()) - fd.download([DownloadTest.COLLEGEHUMOR_URL]) - self.assertTrue(os.path.exists(DownloadTest.COLLEGEHUMOR_FILE)) - md5_down_file = md5_for_file(DownloadTest.COLLEGEHUMOR_FILE) - self.assertEqual(md5_down_file, DownloadTest.COLLEGEHUMOR_MD5) - - def test_xnxx(self): - with open(DownloadTest.PARAMETERS_FILE) as f: - fd = FileDownloader(json.load(f)) - fd.add_info_extractor(XNXXIE()) - fd.download([DownloadTest.XNXX_URL]) - self.assertTrue(os.path.exists(DownloadTest.XNXX_FILE)) - md5_down_file = md5_for_file(DownloadTest.XNXX_FILE) - self.assertEqual(md5_down_file, DownloadTest.XNXX_MD5) - - def tearDown(self): - if os.path.exists(DownloadTest.YOUTUBE_FILE): - os.remove(DownloadTest.YOUTUBE_FILE) - if os.path.exists(DownloadTest.DAILYMOTION_FILE): - os.remove(DownloadTest.DAILYMOTION_FILE) - if os.path.exists(DownloadTest.METACAFE_FILE): - os.remove(DownloadTest.METACAFE_FILE) - if os.path.exists(DownloadTest.BLIP_FILE): - os.remove(DownloadTest.BLIP_FILE) - if os.path.exists(DownloadTest.XVIDEO_FILE): - os.remove(DownloadTest.XVIDEO_FILE) - if os.path.exists(DownloadTest.VIMEO_FILE): - os.remove(DownloadTest.VIMEO_FILE) - if os.path.exists(DownloadTest.SOUNDCLOUD_FILE): - os.remove(DownloadTest.SOUNDCLOUD_FILE) - if os.path.exists(DownloadTest.STANDFORD_FILE): - os.remove(DownloadTest.STANDFORD_FILE) - if os.path.exists(DownloadTest.COLLEGEHUMOR_FILE): - os.remove(DownloadTest.COLLEGEHUMOR_FILE) - if os.path.exists(DownloadTest.XNXX_FILE): - os.remove(DownloadTest.XNXX_FILE) - -def md5_for_file(filename, block_size=2**20): - with open(filename) as f: - md5 = hashlib.md5() - while True: - data = f.read(block_size) - if not data: - break - md5.update(data) - return md5.hexdigest() +import youtube_dl.YoutubeDL +from youtube_dl.compat import ( + compat_http_client, + compat_urllib_error, + compat_HTTPError, +) +from youtube_dl.utils import ( + DownloadError, + ExtractorError, + format_bytes, + UnavailableVideoError, +) +from youtube_dl.extractor import get_info_extractor + +RETRIES = 3 + + +class YoutubeDL(youtube_dl.YoutubeDL): + def __init__(self, *args, **kwargs): + self.to_stderr = self.to_screen + self.processed_info_dicts = [] + super(YoutubeDL, self).__init__(*args, **kwargs) + + def report_warning(self, message): + # Don't accept warnings during tests + raise ExtractorError(message) + + def process_info(self, info_dict): + self.processed_info_dicts.append(info_dict) + return super(YoutubeDL, self).process_info(info_dict) + + +def _file_md5(fn): + with open(fn, 'rb') as f: + return hashlib.md5(f.read()).hexdigest() + + +defs = gettestcases() + + +class TestDownload(unittest.TestCase): + # Parallel testing in nosetests. See + # http://nose.readthedocs.org/en/latest/doc_tests/test_multiprocess/multiprocess.html + _multiprocess_shared_ = True + + maxDiff = None + + def setUp(self): + self.defs = defs + +# Dynamically generate tests + + +def generator(test_case, tname): + + def test_template(self): + ie = youtube_dl.extractor.get_info_extractor(test_case['name']) + other_ies = [get_info_extractor(ie_key) for ie_key in test_case.get('add_ie', [])] + is_playlist = any(k.startswith('playlist') for k in test_case) + test_cases = test_case.get( + 'playlist', [] if is_playlist else [test_case]) + + def print_skipping(reason): + print('Skipping %s: %s' % (test_case['name'], reason)) + if not ie.working(): + print_skipping('IE marked as not _WORKING') + return + + for tc in test_cases: + info_dict = tc.get('info_dict', {}) + if not (info_dict.get('id') and info_dict.get('ext')): + raise Exception('Test definition incorrect. The output file cannot be known. Are both \'id\' and \'ext\' keys present?') + + if 'skip' in test_case: + print_skipping(test_case['skip']) + return + for other_ie in other_ies: + if not other_ie.working(): + print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key()) + return + + params = get_params(test_case.get('params', {})) + params['outtmpl'] = tname + '_' + params['outtmpl'] + if is_playlist and 'playlist' not in test_case: + params.setdefault('extract_flat', 'in_playlist') + params.setdefault('skip_download', True) + + ydl = YoutubeDL(params, auto_init=False) + ydl.add_default_info_extractors() + finished_hook_called = set() + + def _hook(status): + if status['status'] == 'finished': + finished_hook_called.add(status['filename']) + ydl.add_progress_hook(_hook) + expect_warnings(ydl, test_case.get('expected_warnings', [])) + + def get_tc_filename(tc): + return ydl.prepare_filename(tc.get('info_dict', {})) + + res_dict = None + + def try_rm_tcs_files(tcs=None): + if tcs is None: + tcs = test_cases + for tc in tcs: + tc_filename = get_tc_filename(tc) + try_rm(tc_filename) + try_rm(tc_filename + '.part') + try_rm(os.path.splitext(tc_filename)[0] + '.info.json') + try_rm_tcs_files() + try: + try_num = 1 + while True: + try: + # We're not using .download here sine that is just a shim + # for outside error handling, and returns the exit code + # instead of the result dict. + res_dict = ydl.extract_info( + test_case['url'], + force_generic_extractor=params.get('force_generic_extractor', False)) + except (DownloadError, ExtractorError) as err: + # Check if the exception is not a network related one + 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): + raise + + if try_num == RETRIES: + report_warning('%s failed due to network errors, skipping...' % tname) + return + + print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num)) + + try_num += 1 + else: + break + + if is_playlist: + self.assertTrue(res_dict['_type'] in ['playlist', 'multi_video']) + self.assertTrue('entries' in res_dict) + expect_info_dict(self, res_dict, test_case.get('info_dict', {})) + + if 'playlist_mincount' in test_case: + assertGreaterEqual( + self, + len(res_dict['entries']), + test_case['playlist_mincount'], + 'Expected at least %d in playlist %s, but got only %d' % ( + test_case['playlist_mincount'], test_case['url'], + len(res_dict['entries']))) + if 'playlist_count' in test_case: + self.assertEqual( + len(res_dict['entries']), + test_case['playlist_count'], + 'Expected %d entries in playlist %s, but got %d.' % ( + test_case['playlist_count'], + test_case['url'], + len(res_dict['entries']), + )) + if 'playlist_duration_sum' in test_case: + got_duration = sum(e['duration'] for e in res_dict['entries']) + self.assertEqual( + test_case['playlist_duration_sum'], got_duration) + + for tc in test_cases: + tc_filename = get_tc_filename(tc) + if not test_case.get('params', {}).get('skip_download', False): + self.assertTrue(os.path.exists(tc_filename), msg='Missing file ' + tc_filename) + self.assertTrue(tc_filename in finished_hook_called) + expected_minsize = tc.get('file_minsize', 10000) + if expected_minsize is not None: + if params.get('test'): + expected_minsize = max(expected_minsize, 10000) + got_fsize = os.path.getsize(tc_filename) + assertGreaterEqual( + self, got_fsize, expected_minsize, + 'Expected %s to be at least %s, but it\'s only %s ' % + (tc_filename, format_bytes(expected_minsize), + format_bytes(got_fsize))) + if 'md5' in tc: + md5_for_file = _file_md5(tc_filename) + self.assertEqual(md5_for_file, tc['md5']) + info_json_fn = os.path.splitext(tc_filename)[0] + '.info.json' + self.assertTrue( + os.path.exists(info_json_fn), + 'Missing info file %s' % info_json_fn) + with io.open(info_json_fn, encoding='utf-8') as infof: + info_dict = json.load(infof) + + expect_info_dict(self, info_dict, tc.get('info_dict', {})) + finally: + try_rm_tcs_files() + if is_playlist and res_dict is not None and res_dict.get('entries'): + # Remove all other files that may have been extracted if the + # extractor returns full results even with extract_flat + res_tcs = [{'info_dict': e} for e in res_dict['entries']] + try_rm_tcs_files(res_tcs) + + return test_template + + +# And add them to TestDownload +for n, test_case in enumerate(defs): + tname = 'test_' + str(test_case['name']) + i = 1 + while hasattr(TestDownload, tname): + tname = 'test_%s_%d' % (test_case['name'], i) + i += 1 + test_method = generator(test_case, tname) + test_method.__name__ = str(tname) + setattr(TestDownload, test_method.__name__, test_method) + del test_method + + +if __name__ == '__main__': + unittest.main()