quich fixes

src/mfs/cli.py

@@ -39,6 +39,11 @@ def main(argv=sys.argv):
         raise ValueError("{} is not a directory".format(dest))
     url = args['URL']
     follow = not args['--no-follow']
+
+    # browsers return https while the sites are still insecure, so to ease copy/paste
+    # this is a dirty hack and proper url handling is required .. but bit lazy now
+    url = url.replace('https://', 'http://')
+
     print('Processing {}'.format(url))
 
     scraper = None

src/mfs/image_download.py

@@ -11,7 +11,8 @@
 def download_images(dl):
     # avoid to many requests(coroutines) the same time.
     # limit them by setting semaphores (simultaneous requests)
-    _sema = asyncio.Semaphore(10)
+    # setting to 5 as higher number will yield 503 from vfl.ru
+    _sema = asyncio.Semaphore(5)
 
     async def wait_with_progressbar(coros):
         for f in tqdm.tqdm(asyncio.as_completed(coros), total=len(coros)):

tests/test_image_download.py

@@ -64,6 +64,12 @@ def test_live_vfl(tmpdir, testdata):
                        shallow=False), 'Downloaded image does not match the reference one'
 
 
+def test_live_vfl_2020(tmpdir, testdata):
+    fn = tmpdir.strpath + '/aaa.jpg'
+    base.download_images([('http://vfl.ru/fotos/9b930dc417050203.html', fn)])
+    assert len(os.listdir(tmpdir.strpath)) == 1, 'One images shall be downloaded'
+
+
 def test_reference_radikal(testdata):
     url = base._resolve_radikal(testdata.textdata('radikal.ru.html'))
     assert url == 'http://s39.radikal.ru/i084/1106/a9/e1fca250702b.jpg', \