test: extend the reach of info_dict checking

[youtube-dl] / test / test_download.py

diff --git a/test/test_download.py b/test/test_download.py
index 40d596fdf1cf2e4b596917f28ecf431ef024014a..86215203361057ede5b4a3ad5dd5b2a22b004a54 100644 (file)
--- a/test/test_download.py
+++ b/test/test_download.py
@@ -7,8 +7,8 @@ import os
 import json
 import unittest
 import sys
-import hashlib
 import socket
+import binascii
 
 # Allow direct execution
 sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
@@ -20,12 +20,15 @@ from youtube_dl.utils import *
 DEF_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'tests.json')
 PARAMETERS_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "parameters.json")
 
+RETRIES = 3
+
 # General configuration (from __init__, not very elegant...)
 jar = compat_cookiejar.CookieJar()
 cookie_processor = compat_urllib_request.HTTPCookieProcessor(jar)
 proxy_handler = compat_urllib_request.ProxyHandler()
 opener = compat_urllib_request.build_opener(proxy_handler, cookie_processor, YoutubeDLHandler())
 compat_urllib_request.install_opener(opener)
+socket.setdefaulttimeout(10)
 
 def _try_rm(filename):
     """ Remove a file if it exists """
@@ -35,11 +38,17 @@ def _try_rm(filename):
         if ose.errno != errno.ENOENT:
             raise
 
+def crc32(value):
+    return '%08x' % (binascii.crc32(value.encode('utf8')) & 0xffffffff)
+
 class FileDownloader(youtube_dl.FileDownloader):
     def __init__(self, *args, **kwargs):
         self.to_stderr = self.to_screen
         self.processed_info_dicts = []
         return youtube_dl.FileDownloader.__init__(self, *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 youtube_dl.FileDownloader.process_info(self, info_dict)
@@ -55,6 +64,7 @@ with io.open(PARAMETERS_FILE, encoding='utf-8') as pf:
 
 
 class TestDownload(unittest.TestCase):
+    maxDiff = None
     def setUp(self):
         self.parameters = parameters
         self.defs = defs
@@ -63,7 +73,7 @@ class TestDownload(unittest.TestCase):
 def generator(test_case):
 
     def test_template(self):
-        ie = getattr(youtube_dl.InfoExtractors, test_case['name'] + 'IE')
+        ie = youtube_dl.InfoExtractors.get_info_extractor(test_case['name'])
         if not ie._WORKING:
             print('Skipping: IE marked as not _WORKING')
             return
@@ -78,9 +88,13 @@ def generator(test_case):
         params.update(test_case.get('params', {}))
 
         fd = FileDownloader(params)
-        fd.add_info_extractor(ie())
-        for ien in test_case.get('add_ie', []):
-            fd.add_info_extractor(getattr(youtube_dl.InfoExtractors, ien + 'IE')())
+        for ie in youtube_dl.InfoExtractors.gen_extractors():
+            fd.add_info_extractor(ie)
+        finished_hook_called = set()
+        def _hook(status):
+            if status['status'] == 'finished':
+                finished_hook_called.add(status['filename'])
+        fd.add_progress_hook(_hook)
 
         test_cases = test_case.get('playlist', [test_case])
         for tc in test_cases:
@@ -88,11 +102,24 @@ def generator(test_case):
             _try_rm(tc['file'] + '.part')
             _try_rm(tc['file'] + '.info.json')
         try:
-            fd.download([test_case['url']])
+            for retry in range(1, RETRIES + 1):
+                try:
+                    fd.download([test_case['url']])
+                except (DownloadError, ExtractorError) as err:
+                    if retry == RETRIES: raise
+
+                    # Check if the exception is not a network related one
+                    if not err.exc_info[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
+                        raise
+
+                    print('Retrying: {0} failed tries\n\n##########\n\n'.format(retry))
+                else:
+                    break
 
             for tc in test_cases:
                 if not test_case.get('params', {}).get('skip_download', False):
-                    self.assertTrue(os.path.exists(tc['file']))
+                    self.assertTrue(os.path.exists(tc['file']), msg='Missing file ' + tc['file'])
+                    self.assertTrue(tc['file'] in finished_hook_called)
                 self.assertTrue(os.path.exists(tc['file'] + '.info.json'))
                 if 'md5' in tc:
                     md5_for_file = _file_md5(tc['file'])
@@ -100,11 +127,21 @@ def generator(test_case):
                 with io.open(tc['file'] + '.info.json', encoding='utf-8') as infof:
                     info_dict = json.load(infof)
                 for (info_field, value) in tc.get('info_dict', {}).items():
-                    if value.startswith('md5:'):
-                        md5_info_value = hashlib.md5(info_dict.get(info_field, '')).hexdigest()
-                        self.assertEqual(value[3:], md5_info_value)
+                    if isinstance(value, compat_str) and value.startswith('crc32:'):
+                        self.assertEqual(value, 'crc32:' + crc32(info_dict.get(info_field)))
                     else:
                         self.assertEqual(value, info_dict.get(info_field))
+
+                # If checkable fields are missing from the test case, print the info_dict
+                test_info_dict = dict((key, value if not isinstance(value, compat_str) or len(value) < 250 else 'crc32:' + crc32(value))
+                    for key, value in info_dict.items()
+                    if value and key in ('title', 'description', 'uploader', 'upload_date', 'uploader_id', 'location'))
+                if not all(key in tc.get('info_dict', {}).keys() for key in test_info_dict.keys()):
+                    sys.stderr.write(u'\n"info_dict": ' + json.dumps(test_info_dict, ensure_ascii=False, indent=2) + u'\n')
+
+                # Check for the presence of mandatory fields
+                for key in ('id', 'url', 'title', 'ext'):
+                    self.assertTrue(key in info_dict.keys() and info_dict[key])
         finally:
             for tc in test_cases:
                 _try_rm(tc['file'])