From a04975571cad46c5e4d21ed53ac0f49089cb1c99 Mon Sep 17 00:00:00 2001
From: Antoine Lambert <anlambert@softwareheritage.org>
Date: Tue, 26 Sep 2023 13:42:37 +0200
Subject: [PATCH] gitweb: Remove invalid use of str.rstrip
rstrip is not a method to remove a string suffix so use another
way to extract gitweb project name.
It fixes the computation of some gitweb origin URLs.
Related to swh/infra/sysadm-environment#5050.
---
swh/lister/gitweb/lister.py | 10 +++-------
swh/lister/gitweb/tests/test_lister.py | 4 ++++
2 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/swh/lister/gitweb/lister.py b/swh/lister/gitweb/lister.py
index 8d0d81b8..1710ba79 100644
--- a/swh/lister/gitweb/lister.py
+++ b/swh/lister/gitweb/lister.py
@@ -167,13 +167,9 @@ def try_to_determine_git_repository(repository_url: str) -> Optional[str]:
"""
result = None
parsed_url = urlparse(repository_url)
- params = parse_qs(parsed_url.query).get("p")
- if params:
- repo = params[0]
- if repo and repo.endswith(";a=summary"):
- repo = repo.rstrip(";a=summary")
-
- result = f"git://{parsed_url.netloc}/{repo}"
+ repo = parse_qs(parsed_url.query, separator=";").get("p")
+ if repo:
+ result = f"git://{parsed_url.netloc}/{repo[0]}"
return result
diff --git a/swh/lister/gitweb/tests/test_lister.py b/swh/lister/gitweb/tests/test_lister.py
index 1c83e7dc..dbb5ecbf 100644
--- a/swh/lister/gitweb/tests/test_lister.py
+++ b/swh/lister/gitweb/tests/test_lister.py
@@ -138,6 +138,10 @@ def test_lister_gitweb_get_origin_from_repo_failing(
"https://git.shadowcat.co.uk?p=File-Slurp.git;a=summary",
"git://git.shadowcat.co.uk/File-Slurp.git",
),
+ (
+ "https://git.example.org?p=baaaa;a=summary",
+ "git://git.example.org/baaaa",
+ ),
("https://domain.org/foobar", None),
],
)
--
GitLab