implement HTMLParseError, additional typing

pytube/captions.py

@@ -26,7 +26,7 @@ class Caption:
         """Download the xml caption tracks."""
         return request.get(self.url)
 
-    def generate_srt_captions(self):
+    def generate_srt_captions(self) -> str:
         """Generate "SubRip Subtitle" captions.
 
         Takes the xml captions from :meth:`~pytube.Caption.xml_captions` and
@@ -49,7 +49,7 @@ class Caption:
         ms = "{:.3f}".format(frac).replace("0.", "")
         return time_fmt + ms
 
-    def xml_caption_to_srt(self, xml_captions):
+    def xml_caption_to_srt(self, xml_captions: str) -> str:
         """Convert xml caption tracks to "SubRip Subtitle (srt)".
 
         :param str xml_captions:

pytube/exceptions.py

@@ -42,3 +42,7 @@ class LiveStreamError(ExtractError):
 
 class VideoUnavailable(PytubeError):
     """Video is unavailable."""
+
+
+class HTMLParseError(PytubeError):
+    """HTML could not be parsed"""

pytube/extract.py

@@ -4,10 +4,10 @@ import json
 from collections import OrderedDict
 
 from html.parser import HTMLParser
-from typing import Any, Optional
+from typing import Any, Optional, Tuple, List
 from urllib.parse import quote
 from urllib.parse import urlencode
-from pytube.exceptions import RegexMatchError
+from pytube.exceptions import RegexMatchError, HTMLParseError
 from pytube.helpers import regex_search
 
 
@@ -37,6 +37,9 @@ class PytubeHTMLParser(HTMLParser):
         elif self.in_vid_descr:
             self.vid_descr += data
 
+    def error(self, message):
+        raise HTMLParseError(message)
+
 
 def is_age_restricted(watch_html: str) -> bool:
     """Check if content is age restricted.
@@ -106,8 +109,7 @@ def video_info_url(
     :param str watch_url:
         A YouTube watch url.
     :param str watch_html:
-        The html contents of the watch page.
-        TODO: unused
+        (Unused) The html contents of the watch page.
     :param str embed_html:
         The html contents of the embed page (for age restricted videos).
     :param bool age_restricted:
@@ -154,7 +156,7 @@ def js_url(html: str, age_restricted: bool = False) -> str:
     return "https://youtube.com" + base_js
 
 
-def mime_type_codec(mime_type_codec):
+def mime_type_codec(mime_type_codec: str) -> Tuple[str, List[str]]:
     """Parse the type data.
 
     Breaks up the data in the ``type`` key of the manifest, which contains the

pytube/helpers.py

@@ -135,7 +135,7 @@ def safe_filename(s: str, max_length: int = 255) -> str:
     return filename[:max_length].rsplit(" ", 0)[0]
 
 
-def create_logger(level=logging.ERROR):
+def create_logger(level:int = logging.ERROR) -> logging.Logger:
     """Create a configured instance of logger.
 
     :param int level:

pytube/itags.py

@@ -100,7 +100,7 @@ DASH_MP4_VIDEO = [133, 134, 135, 136, 137, 138, 160, 212, 264, 266, 298, 299]
 DASH_MP4_AUDIO = [139, 140, 141, 256, 258, 325, 328]
 
 
-def get_format_profile(itag):
+def get_format_profile(itag: int) -> Dict:
     """Get additional format information for a given itag.
 
     :param str itag:

pytube/mixins.py

@@ -130,7 +130,7 @@ def apply_descrambler(stream_data: Dict, key: str) -> None:
     )
 
 
-def install_proxy(proxy_handler):
+def install_proxy(proxy_handler: Dict[str, str]) -> None:
     proxy_support = request.ProxyHandler(proxy_handler)
     opener = request.build_opener(proxy_support)
     request.install_opener(opener)