Address hash problem

Author: armenzg

src/sentry/tasks/commits.py

@@ -33,7 +33,7 @@
 from sentry.users.services.user.service import user_service
 from sentry.utils.cache import cache
 from sentry.utils.email import MessageBuilder
-from sentry.utils.hashlib import md5_text
+from sentry.utils.hashlib import hash_values
 from sentry.utils.http import absolute_uri
 
 logger = logging.getLogger(__name__)
@@ -94,9 +94,10 @@ def get_github_compare_commits_cache_key(
     start_sha: str | None,
     end_sha: str,
 ) -> str:
-    digest = md5_text(
-        organization_id, repository_id, provider or "", start_sha or "", end_sha
-    ).hexdigest()
+    digest = hash_values(
+        [organization_id, repository_id, provider or "", start_sha or "", end_sha],
+        seed="fetch-commits:compare-commits",
+    )
     return f"fetch-commits:compare-commits:v1:{digest}"
 
 

tests/sentry/tasks/test_commits.py

@@ -13,7 +13,11 @@
 from sentry.models.releaseheadcommit import ReleaseHeadCommit
 from sentry.models.repository import Repository
 from sentry.silo.base import SiloMode
-from sentry.tasks.commits import fetch_commits, handle_invalid_identity
+from sentry.tasks.commits import (
+    fetch_commits,
+    get_github_compare_commits_cache_key,
+    handle_invalid_identity,
+)
 from sentry.testutils.asserts import assert_slo_metric
 from sentry.testutils.cases import TestCase
 from sentry.testutils.silo import assume_test_silo_mode, control_silo_test
@@ -30,6 +34,14 @@ def _github_compare_commits_result(self, repo_name: str, end_sha: str) -> list[d
             {"id": end_sha, "repository": repo_name},
         ]
 
+    def test_github_compare_commits_cache_key_avoids_ambiguous_id_collisions(
+        self, mock_record: MagicMock
+    ) -> None:
+        key_one = get_github_compare_commits_cache_key(1, 23, "integrations:github", "a", "b")
+        key_two = get_github_compare_commits_cache_key(12, 3, "integrations:github", "a", "b")
+
+        assert key_one != key_two
+
     def _test_simple_action(self, user, org):
         repo = Repository.objects.create(name="example", provider="dummy", organization_id=org.id)
         release = Release.objects.create(organization_id=org.id, version="abcabcabc")