Lomanic/NewLeaf
1
0
Fork 0
mirror of https://git.sr.ht/~cadence/NewLeaf synced 2024-11-24 00:27:28 +00:00
Code Issues Releases Wiki Activity

#29 Extract named channels using dynamic endpoint with second__path param instead of /user/

Browse Source
This commit is contained in:
Lomanic 2021-11-13 20:15:16 +01:00 committed by Cadence Ember
parent 29a3894337
commit ac1aa07108
2 changed files with 12 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
extractors/channel.py
View File

@ -12,12 +12,12 @@ channel_cache_lock = Lock()
channel_latest_cache = TTLCache(maxsize=500, ttl=300) channel_latest_cache = TTLCache(maxsize=500, ttl=300)
channel_latest_cache_lock = Lock() channel_latest_cache_lock = Lock()
def extract_channel(ucid): def extract_channel(ucid, second__path="user"):
with channel_cache_lock: with channel_cache_lock:
if ucid in channel_cache: if ucid in channel_cache:
return channel_cache[ucid] return channel_cache[ucid]
channel_type = "channel" if len(ucid) == 24 and ucid[:2] == "UC" else "user" channel_type = "channel" if len(ucid) == 24 and ucid[:2] == "UC" else second__path
r = requests.get("https://www.youtube.com/{}/{}/videos?hl=en".format(channel_type, ucid), cookies=eu_consent_cookie()) r = requests.get("https://www.youtube.com/{}/{}/videos?hl=en".format(channel_type, ucid), cookies=eu_consent_cookie())
r.raise_for_status() r.raise_for_status()
yt_initial_data = extract_yt_initial_data(r.content.decode("utf8")) yt_initial_data = extract_yt_initial_data(r.content.decode("utf8"))

14
index.py
View File

@ -58,7 +58,7 @@ class NewLeaf(object):
@cherrypy.expose @cherrypy.expose
@cherrypy.tools.json_out() @cherrypy.tools.json_out()
def channels(self, *suffix, **kwargs): def channels(self, *suffix, second__path="user", **kwargs):
ucid = "" ucid = ""
part = "" part = ""
possible_parts = ("videos", "latest", "playlists") possible_parts = ("videos", "latest", "playlists")
@ -74,15 +74,21 @@ class NewLeaf(object):
"error": "Two components specified in URL, but neither component was recognised as a part keyword.", "error": "Two components specified in URL, but neither component was recognised as a part keyword.",
"identifier": "PART_KEYWORD_NOT_RECOGNISED" "identifier": "PART_KEYWORD_NOT_RECOGNISED"
} }
possible_paths = ("channel", "c", "user")
if second__path not in possible_paths:
return {
"error": "second__path parameter must be one of: " + str(possible_paths),
"identifier": "PATH_PARAMETER_NOT_RECOGNISED"
}
if part == "playlists": if part == "playlists":
return [] return []
elif part == "latest": elif part == "latest":
return extract_channel_latest(ucid) return extract_channel_latest(ucid, second__path)
elif part == "videos": elif part == "videos":
return extract_channel_videos(ucid) return extract_channel_videos(ucid, second__path)
else: # part == "", so extract whole channel else: # part == "", so extract whole channel
return extract_channel(ucid) return extract_channel(ucid, second__path)
@cherrypy.expose @cherrypy.expose
@cherrypy.tools.json_out() @cherrypy.tools.json_out()