Troglodite/scotus-watch
3
0
Fork 0
Code Issues 10 Pull Requests Actions Packages Projects Releases Wiki Activity

Process 404 returns correctly. Fixes: #20

Browse Source
This commit is contained in:
Christopher T. Johnson
2025-02-27 10:22:20 -05:00
parent a6316f27f6
commit 30dd9b2bcd
1 changed files with 4 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
workers.py
View File

@@ -129,12 +129,16 @@ def update_db(case_id: str, db: QSqlDatabase) -> int:
r = requests.get( r = requests.get(
f"https://www.supremecourt.gov/docket/docketfiles/html/public/{case_id}.html" f"https://www.supremecourt.gov/docket/docketfiles/html/public/{case_id}.html"
) )
if r.status_code == 404:
return -1
if r.status_code != 200: if r.status_code != 200:
print(r.status_code) print(r.status_code)
exit(1) exit(1)
bs = BeautifulSoup(r.text, "lxml") bs = BeautifulSoup(r.text, "lxml")
# #
# SCOTUS does not return 404 for page not found. # SCOTUS does not return 404 for page not found.
# Feb 27: SCOTUS is returning 404s but I don't trust them.
# #
title = bs.find("title") title = bs.find("title")
assert isinstance(title, Tag) and isinstance(title.string, str) assert isinstance(title, Tag) and isinstance(title.string, str)