sherl/ythdd
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

added comments to v1 api, introduced failsafe for videos without related video feed (e.g. age restricted videos)

Browse Source
This commit is contained in:
sherl
2024-10-31 01:35:53 +01:00
parent c6734c3d99
commit 9aea3b2e55
1 changed files with 30 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files

66
ythdd_api_v1.py
View File

@@ -31,22 +31,34 @@ def stats():
def hot(data):
#print(data)
# if we are given not enough data to work with, return bad request.
# example:
# hot related videoId
# 0. 1. 2. [No. of argument]
if len(data) <= 2:
incrementBadRequests()
return 400, f'error: bad request. supply required arguments.', []
# check our first argument
match data[1]:
# if it can be handled by yt_dlp, use yt_dlp
case "video" | "channel" | "handle" | "playlist":
url_lookup = {'video': 'https://www.youtube.com/watch?v=', 'channel': 'https://www.youtube.com/channel/', 'handle': 'https://www.youtube.com/@', 'playlist': 'https://www.youtube.com/playlist?list='}
comment_count = ""
# require 3 arguments (if using "c" or "nc")
if len(data) <= 3:
incrementBadRequests()
return 400, f'error: bad request. supply required arguments.', []
# comment settings should be either Comments, No Comments, or Limited Comments
if data[2] not in ("c", "nc", "lc"):
incrementBadRequests()
return notImplemented(data)
# require 4 arguments (when using "lc" we need to know the number of comments to retrieve)
if data[2] == "lc" and len(data) <= 4:
incrementBadRequests()
return 400, f'error: bad request. limited comments (lc) requires an extra argument specifying amount of comments.', []
# check if the additional "lc" argument is a number
elif data[2] == "lc":
try:
comment_count = str(int(data[3]))
@@ -54,6 +66,7 @@ def hot(data):
incrementBadRequests()
return 400, f'error: bad request. {data[3]} is not a number.', []
videoId = data[4]
# if the user didn't choose Limited Comments, then the videoId must be the fourth (zero included) argument
else:
videoId = data[3]
@@ -61,19 +74,29 @@ def hot(data):
incrementBadRequests()
return 400, f'error: bad request. wrong videoId: {videoId} is {len(videoId)} characters long, but should be 11.', []
# assume we want to get the comments
getcomments = True
# unless we're given "nc"
if data[2] == "nc":
getcomments = False
# try to get the data
try:
started = time.time()
extracted_dict = ythdd_extractor.extract(url_lookup[data[1]] + videoId, getcomments=getcomments, maxcomments=comment_count)
extracted_dict["took"] = time.time() - started
return 200, "OK", extracted_dict
except Exception as e:
# shrink yt_dlp's unnecessarily long error message
# example: \u001b[0;31mERROR:\u001b[0m [youtube] videoIdABCD: Sign in to confirm your age. This video may be inappropriate for some users.
# TODO: perhaps implement proper error handling? so that we don't need this hacky solution
incrementBadRequests()
return 400, f'error: failed to get "{videoId}" ({data[2]})', {'error_msg': str(e)}
error_msg = str(e).replace("\u001b[0;31mERROR:\u001b[0m [youtube] " + videoId + ": ", "") # yt_dlp error message with color codes
return 400, f'error: failed to get "{videoId}" ({data[2]})', {'error_msg': error_msg}
# if some functionality is not supported or buried deep within yt_dlp, as is the case with "related videos" feed, use our own (naive) approach
case "related":
# no logic needed for additional arguments here
videoId = data[2]
if len(videoId) != 11: # videoId sanity check
incrementBadRequests()
@@ -81,10 +104,16 @@ def hot(data):
started = time.time()
try:
# try to actually get the data
extracted_related = ythdd_extractor.related('https://www.youtube.com/watch?v=' + videoId)
extracted_related['took'] = time.time() - started
return 200, "OK", extracted_related
except KeyError:
# instead of throwing error at age restricted videos for not having "related videos" feed,
# return an empty feed
return 200, "OK", {'secondaryResults': {'results': []}, 'took': time.time() - started}
except Exception as e:
# general exception handler
incrementBadRequests()
return 400, f'error: unknown error while parsing {videoId}: {e}', []
@@ -92,41 +121,6 @@ def hot(data):
incrementBadRequests()
return notImplemented(data)
'''
if data[1] not in ("video", "channel", "handle", "playlist", "related"):
incrementBadRequests()
return notImplemented(data)
if data[2] not in ("c", "nc", "lc"): # comments, no comments, limited comments
incrementBadRequests()
return notImplemented(data)
if data[2] == "lc":
if len(data) <= 4:
incrementBadRequests()
return 400, f'error: bad request. limited comments (lc) requires an extra argument specifying amount of comments.', []
try:
comment_count = str(int(data[3]))
except:
incrementBadRequests()
return 400, f'error: bad request. {data[3]} is not a number.', []
videoId = data[4]
else:
videoId = data[3]
try:
url_lookup = {'video': 'https://www.youtube.com/watch?v=', 'channel': 'https://www.youtube.com/channel/', 'handle': 'https://www.youtube.com/@', 'playlist': 'https://www.youtube.com/playlist?list='}
if data[2] == "nc":
getcomments = False
else:
getcomments = True
started = int(time.time())
extracted_dict = ythdd_extractor.extract(url_lookup[data[1]] + videoId, getcomments=getcomments, maxcomments=comment_count)
extracted_dict["took"] = int(time.time()) - started
return 200, "OK", extracted_dict
except Exception as e:
incrementBadRequests()
return 400, f'error: failed to get "{videoId}" ({data[2]}). {e}', []
'''
def lookup(data):
match data[0]:
case 'stats':