fix(core): fix git url regex matching taking too long (#3213)

Co-authored-by: Ralf Grubenmann <ralf.grubenmann@gmail.com>
SwissDataScienceCenter · Nov 17, 2022 · 8245ce3 · 8245ce3
1 parent 4dca1a4
commit 8245ce3
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 1 deletion.
diff --git a/renku/domain_model/git.py b/renku/domain_model/git.py
@@ -44,7 +44,7 @@
 
 _RE_PORT = r":(?P<port>\d+)"
 
-_RE_PATHNAME = r"(?P<path>(([\w\-\~\.]+)/)*?(((?P<owner>([\w\-\.]+/?)+)/)?(?P<name>[\w\-\.]+)(\.git)?)?)/*"
+_RE_PATHNAME = r"(?P<path>((\~[\w\-\.]+)/)*?(((?P<owner>[\w\-\./]+?)/)?(?P<name>[\w\-\.]+)(\.git)?)?)/*"
 
 _RE_PATHNAME_WITH_GITLAB = (
     r"(?P<path>((((gitlab/){0,1}|([\w\-\~\.]+/)*?)(?P<owner>([\w\-\.]+/)*[\w\-\.]+)/)?"

diff --git a/tests/core/models/test_git.py b/tests/core/models/test_git.py
@@ -18,6 +18,7 @@
 """Git regex tests."""
 
 import os
+import time
 
 import pytest
 
@@ -287,6 +288,15 @@
             "port": "1234",
             "env": "https://gitlab.example.com:1234/",
         },
+        {
+            "href": "https://gitlab.example.com/renku-test/test-2022-11-11-17-01-46.git",
+            "scheme": "https",
+            "hostname": "gitlab.example.com",
+            "name": "test-2022-11-11-17-01-46",
+            "path": "renku-test/test-2022-11-11-17-01-46.git",
+            "owner": "renku-test",
+            "env": "https://gitlab.example.com",
+        },
     ],
 )
 def test_valid_href(fields):
@@ -296,4 +306,8 @@ def test_valid_href(fields):
     if gitlab_env:
         os.environ["GITLAB_BASE_URL"] = gitlab_env
 
+    start = time.monotonic()
     assert GitURL(**fields) == GitURL.parse(fields["href"])
+    duration = time.monotonic() - start
+
+    assert duration < 1.0, "Something wrong with the GitUrl regexes, probably catastrophic backtracking"