justinzeus/scholarr
1
0
Fork 0
Code Issues 1 Pull requests 18 Projects Releases Packages Wiki Activity Actions

decompose application.py into enrichment, scholar processing, run completion

Browse source
This commit is contained in:
Justin Visser 2026-02-27 14:19:39 +01:00
parent 9a7fa5004e
commit 6379c97e90
28 changed files with 2982 additions and 2576 deletions
Download patch file Download diff file Expand all files Collapse all files

12
app/services/scholar/state_detection.py
View file

@ -38,18 +38,18 @@ def classify_block_or_captcha_reason(
) -> str | None:
if "accounts.google.com" in final_url and ("signin" in final_url or "servicelogin" in final_url):
return "blocked_accounts_redirect"
if status_code == 429:
return "blocked_http_429_rate_limited"
if status_code == 403:
if "recaptcha" in body_lowered or "captcha" in body_lowered or "sorry/index" in final_url:
return "blocked_http_403_captcha_challenge"
return "blocked_http_403_forbidden"
if "sorry/index" in final_url or "sorry/index" in body_lowered:
return "blocked_google_sorry_challenge"
if "our systems have detected" in body_lowered or "unusual traffic" in body_lowered:
return "blocked_unusual_traffic_detected"
if "automated queries" in body_lowered:
return "blocked_automated_queries_detected"
if status_code == 429:
return "blocked_http_429_rate_limited"
if status_code == 403:
if "recaptcha" in body_lowered or "captcha" in body_lowered:
return "blocked_http_403_captcha_challenge"
return "blocked_http_403_forbidden"
if "not a robot" in body_lowered:
return "blocked_not_a_robot_challenge"
if "recaptcha" in body_lowered: