Merge pull request #792 from fp7/master
[youtube-dl] / test / test_download.py
index 23819505095b11fba46d48eaf028028804272410..cf8028718fc86f31da16a0910817e4f06e56f8e3 100644 (file)
@@ -1,5 +1,6 @@
 #!/usr/bin/env python
 
+import errno
 import hashlib
 import io
 import os
@@ -19,12 +20,23 @@ 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 """
+    try:
+        os.remove(filename)
+    except OSError as ose:
+        if ose.errno != errno.ENOENT:
+            raise
 
 class FileDownloader(youtube_dl.FileDownloader):
     def __init__(self, *args, **kwargs):
@@ -46,54 +58,75 @@ 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
 
-        # Clear old files
-        self.tearDown()
-
-    def tearDown(self):
-        for files in [ test['files'] for test in self.defs ]:
-            for fn, md5 in files:
-                if os.path.exists(fn):
-                    os.remove(fn)
-
-
 ### Dynamically generate tests
 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'])#getattr(youtube_dl.InfoExtractors, test_case['name'] + 'IE')
         if not ie._WORKING:
             print('Skipping: IE marked as not _WORKING')
             return
+        if 'playlist' not in test_case and not test_case['file']:
+            print('Skipping: No output file specified')
+            return
         if 'skip' in test_case:
             print('Skipping: {0}'.format(test_case['skip']))
             return
 
-        params = dict(self.parameters) # Duplicate it locally
-        for p in test_case.get('params', {}):
-            params[p] = test_case['params'][p]
+        params = self.parameters.copy()
+        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')())
-        fd.download([test_case['url']])
-
-        for filename, md5 in test_case['files']:
-            self.assertTrue(os.path.exists(filename))
-            if md5:
-                md5_for_file = _file_md5(filename)
-                self.assertEqual(md5_for_file, md5)
-        info_dict = fd.processed_info_dicts[0]
-        for (info_field, value) in test_case.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)
-            else:
-                self.assertEqual(value, info_dict.get(info_field))
+        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:
+            _try_rm(tc['file'])
+            _try_rm(tc['file'] + '.part')
+            _try_rm(tc['file'] + '.info.json')
+        try:
+            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']), 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'])
+                    self.assertEqual(md5_for_file, tc['md5'])
+                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():
+                    self.assertEqual(value, info_dict.get(info_field))
+        finally:
+            for tc in test_cases:
+                _try_rm(tc['file'])
+                _try_rm(tc['file'] + '.part')
+                _try_rm(tc['file'] + '.info.json')
 
     return test_template