streamlined and simplified dynamic tests generation; readded a couple of test features

pull/8/head
Filippo Valsorda 12 years ago
parent 2b5b2cb84c
commit fd5ff02042

@ -1,72 +1,94 @@
#!/usr/bin/env python2 #!/usr/bin/env python
import hashlib import hashlib
import io
import os import os
import json import json
import unittest import unittest
import sys import sys
import socket
# Allow direct execution
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from youtube_dl.FileDownloader import FileDownloader import youtube_dl.FileDownloader
#import all the info extractor import youtube_dl.InfoExtractors
import youtube_dl from youtube_dl.utils import *
DEF_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'tests.json') 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' PARAMETERS_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)), "parameters.json")
)
# 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(300) # 5 minutes should be enough (famous last words)
class FileDownloader(youtube_dl.FileDownloader):
def __init__(self, *args, **kwargs):
youtube_dl.FileDownloader.__init__(self, *args, **kwargs)
self.to_stderr = self.to_screen
def _file_md5(fn):
with open(fn, 'rb') as f:
return hashlib.md5(f.read()).hexdigest()
with io.open(DEF_FILE, encoding='utf-8') as deff:
defs = json.load(deff)
with io.open(PARAMETERS_FILE, encoding='utf-8') as pf:
parameters = json.load(pf)
class TestDownload(unittest.TestCase): class TestDownload(unittest.TestCase):
pass def setUp(self):
self.parameters = parameters
self.defs = defs
# Clear old files
self.tearDown()
def tearDown(self):
for fn in [ test.get('file', False) for test in self.defs ]:
if fn and os.path.exists(fn):
os.remove(fn)
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()
### Dinamically generate tests
def generator(test_case):
def generator(name, url, md5, file, ie_param, optional_ie):
def test_template(self): def test_template(self):
fd = FileDownloader(ie_param) ie = getattr(youtube_dl.InfoExtractors, test_case['name'] + 'IE')
fd.add_info_extractor(getattr(youtube_dl, name + "IE")()) if not ie._WORKING:
fd.download([url]) print('Skipping: IE marked as not _WORKING')
self.assertTrue(os.path.exists(file)) return
self.assertEqual(md5_for_file(file), md5) if 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]
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']])
self.assertTrue(os.path.exists(test_case['file']))
if 'md5' in test_case:
md5_for_file = _file_md5(test_case['file'])
self.assertEqual(md5_for_file, test_case['md5'])
return test_template return test_template
#only python 2.7
### And add them to TestDownload
def clean_generator(files): for test_case in defs:
def clean_template(self): test_method = generator(test_case)
for file_name in files: test_method.__name__ = "test_{0}".format(test_case["name"])
if os.path.exists(file_name): setattr(TestDownload, test_method.__name__, test_method)
os.remove(file_name) del test_method
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
if __name__ == '__main__': if __name__ == '__main__':

@ -57,7 +57,7 @@ class TestYoutubeLists(unittest.TestCase):
self.assertEqual(DL.result[-1], ['http://www.youtube.com/watch?v=rYefUsYuEp0']) self.assertEqual(DL.result[-1], ['http://www.youtube.com/watch?v=rYefUsYuEp0'])
def test_youtube_channel(self): def test_youtube_channel(self):
"""I give up, please find a channel that does paginate and test this like test_youtube_playlist_long""" # I give up, please find a channel that does paginate and test this like test_youtube_playlist_long
pass # TODO pass # TODO
def test_youtube_user(self): def test_youtube_user(self):

@ -57,6 +57,6 @@
"url": "http://v.youku.com/v_show/id_XNDgyMDQ2NTQw.html", "url": "http://v.youku.com/v_show/id_XNDgyMDQ2NTQw.html",
"file": "XNDgyMDQ2NTQw_part00.flv", "file": "XNDgyMDQ2NTQw_part00.flv",
"md5": "ffe3f2e435663dc2d1eea34faeff5b5b", "md5": "ffe3f2e435663dc2d1eea34faeff5b5b",
"broken": true "params": { "test": false }
} }
] ]

Loading…
Cancel
Save