Merge branch 'master' into rtmpdump
[youtube-dl] / test / test_download.py
index be15fb5abefe43cfabba48c8e7fcb35272810b8d..16f2008094fca751f69ee01a38f16a6fa0c3fb53 100644 (file)
-#!/usr/bin/env python2
-import hashlib
+#!/usr/bin/env python
+
+# Allow direct execution
 import os
-import json
-import unittest
 import sys
+import unittest
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from test.helper import (
+    get_params,
+    get_testcases,
+    global_setup,
+    try_rm,
+    md5,
+    report_warning
+)
+global_setup()
 
-from youtube_dl.FileDownloader import FileDownloader
-#import all the info extractor
-import youtube_dl
 
-DEF_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'tests.json')
-PARAM_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'parameters.json'
+import hashlib
+import io
+import json
+import socket
+
+import youtube_dl.YoutubeDL
+from youtube_dl.utils import (
+    compat_str,
+    compat_urllib_error,
+    compat_HTTPError,
+    DownloadError,
+    ExtractorError,
+    UnavailableVideoError,
 )
+from youtube_dl.extractor import get_info_extractor
 
+RETRIES = 3
 
-class TestDownload(unittest.TestCase):
-    pass
+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 = get_testcases()
 
 
-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()
+class TestDownload(unittest.TestCase):
+    maxDiff = None
+    def setUp(self):
+        self.defs = defs
 
+### Dynamically generate tests
+def generator(test_case):
 
-def generator(name, url, md5, file, ie_param, optional_ie):
     def test_template(self):
-        fd = FileDownloader(ie_param)
-        fd.add_info_extractor(getattr(youtube_dl, name + "IE")())
-        fd.download([url])
-        self.assertTrue(os.path.exists(file))
-        self.assertEqual(md5_for_file(file), md5)
+        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', [])]
+        def print_skipping(reason):
+            print('Skipping %s: %s' % (test_case['name'], reason))
+        if not ie.working():
+            print_skipping('IE marked as not _WORKING')
+            return
+        if 'playlist' not in test_case:
+            info_dict = test_case.get('info_dict', {})
+            if not test_case.get('file') and not (info_dict.get('id') and info_dict.get('ext')):
+                print_skipping('The output file cannot be know, the "file" '
+                    'key is missing or the info_dict is incomplete')
+                return
+        if 'skip' in test_case:
+            print_skipping(test_case['skip'])
+            return
+        for other_ie in other_ies:
+            if not other_ie.working():
+                print_skipping(u'test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
+                return
+
+        params = get_params(test_case.get('params', {}))
+
+        ydl = YoutubeDL(params)
+        ydl.add_default_info_extractors()
+        finished_hook_called = set()
+        def _hook(status):
+            if status['status'] == 'finished':
+                finished_hook_called.add(status['filename'])
+        ydl.fd.add_progress_hook(_hook)
+
+        def get_tc_filename(tc):
+            return tc.get('file') or ydl.prepare_filename(tc.get('info_dict', {}))
+
+        test_cases = test_case.get('playlist', [test_case])
+        def try_rm_tcs_files():
+            for tc in test_cases:
+                tc_filename = get_tc_filename(tc)
+                try_rm(tc_filename)
+                try_rm(tc_filename + '.part')
+                try_rm(tc_filename + '.info.json')
+        try_rm_tcs_files()
+        try:
+            try_num = 1
+            while True:
+                try:
+                    ydl.download([test_case['url']])
+                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) or (err.exc_info[0] == compat_HTTPError and err.exc_info[1].code == 503):
+                        raise
+
+                    if try_num == RETRIES:
+                        report_warning(u'Failed due to network errors, skipping...')
+                        return
+
+                    print('Retrying: {0} failed tries\n\n##########\n\n'.format(try_num))
+
+                    try_num += 1
+                else:
+                    break
+
+            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)
+                self.assertTrue(os.path.exists(tc_filename + '.info.json'))
+                if 'md5' in tc:
+                    md5_for_file = _file_md5(tc_filename)
+                    self.assertEqual(md5_for_file, tc['md5'])
+                with io.open(tc_filename + '.info.json', encoding='utf-8') as infof:
+                    info_dict = json.load(infof)
+                for (info_field, expected) in tc.get('info_dict', {}).items():
+                    if isinstance(expected, compat_str) and expected.startswith('md5:'):
+                        got = 'md5:' + md5(info_dict.get(info_field))
+                    else:
+                        got = info_dict.get(info_field)
+                    self.assertEqual(expected, got,
+                        u'invalid value for field %s, expected %r, got %r' % (info_field, expected, got))
+
+                # 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 'md5:' + md5(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])
+                # Check for mandatory fields that are automatically set by YoutubeDL
+                for key in ['webpage_url', 'extractor', 'extractor_key']:
+                    self.assertTrue(info_dict.get(key), u'Missing field: %s' % key)
+        finally:
+            try_rm_tcs_files()
+
     return test_template
-    #only python 2.7
-
-def clean_generator(files):
-    def clean_template(self):
-        for file_name in files:
-            if os.path.exists(file_name):
-                os.remove(file_name)
-    return clean_template
-
-with open(DEF_FILE, "r") as f:
-    with open(PARAM_FILE) as fp:
-        p = json.load(fp)
-        test_param = json.load(f)
-        files = set()
-        for test_case in test_param:
-            if test_case.get("broken", False):
-                continue
-            try:
-                files.add(test_case["file"])
-                test_method = generator(test_case['name'], test_case['url'], test_case['md5'], test_case['file'], p, test_case.get('add_ie', []))
-                test_method.__name__ = "test_{0}".format(test_case["name"])
-                setattr(TestDownload, test_method.__name__, test_method)
-                del test_method
-            except KeyError as e:
-                sys.stderr.write("Issue with the parameters of test {0}.\n".format(test_case.get("name", "unknown test")))
-        #clean the files
-        ff = clean_generator(files)
-        ff.__name__ = "tearDown"
-        setattr(TestDownload, ff.__name__, ff)
-        del ff
+
+### And add them to TestDownload
+for n, test_case in enumerate(defs):
+    test_method = generator(test_case)
+    tname = 'test_' + str(test_case['name'])
+    i = 1
+    while hasattr(TestDownload, tname):
+        tname = 'test_'  + str(test_case['name']) + '_' + str(i)
+        i += 1
+    test_method.__name__ = tname
+    setattr(TestDownload, test_method.__name__, test_method)
+    del test_method
 
 
 if __name__ == '__main__':