feat(commit-context): Do not create if older than 1 year (#54624)

## Objective:
Do not create a suspect commit if the resulting commit is over 1 year
old.

Author: NisanthanNanthakumar

src/sentry/integrations/utils/commit_context.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import logging
+from datetime import datetime, timedelta, timezone
 from typing import Any, List, Mapping, Sequence, Tuple
 
 import sentry_sdk
@@ -108,7 +109,14 @@ def find_commit_context_for_event(
                 },
             )
 
-        if commit_context:
+        # Only return suspect commits that are less than a year old
+        if commit_context and is_date_less_than_year(commit_context["committedDate"]):
             result.append((commit_context, code_mapping))
 
     return result, installation
+
+
+def is_date_less_than_year(date):
+    return datetime.strptime(date, "%Y-%m-%dT%H:%M:%SZ").replace(
+        tzinfo=timezone.utc
+    ) > datetime.now(tz=timezone.utc) - timedelta(days=365)

tests/sentry/tasks/test_commit_context.py

@@ -1,4 +1,5 @@
-from datetime import timedelta
+from datetime import datetime, timedelta
+from datetime import timezone as datetime_timezone
 from unittest.mock import Mock, patch
 
 import pytest
@@ -83,7 +84,9 @@ class TestCommitContext(TestCommitContextMixin):
         "sentry.integrations.github.GitHubIntegration.get_commit_context",
         return_value={
             "commitId": "asdfwreqr",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "admin@localhost",
@@ -141,11 +144,51 @@ def test_failed_to_fetch_commit_context_record(self, mock_get_commit_context, mo
             error_message="integration_failed",
         )
 
+    @patch("sentry.tasks.commit_context.logger")
     @patch(
         "sentry.integrations.github.GitHubIntegration.get_commit_context",
         return_value={
             "commitId": "asdfasdf",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (
+                datetime.now(tz=datetime_timezone.utc) - timedelta(days=370)
+            ).strftime("%Y-%m-%dT%H:%M:%SZ"),
+            "commitMessage": "placeholder commit message",
+            "commitAuthorName": "",
+            "commitAuthorEmail": "admin@localhost",
+        },
+    )
+    def test_found_commit_is_too_old(self, mock_get_commit_context, mock_logger):
+        with self.tasks():
+            assert not GroupOwner.objects.filter(group=self.event.group).exists()
+            event_frames = get_frame_paths(self.event)
+            process_commit_context(
+                event_id=self.event.event_id,
+                event_platform=self.event.platform,
+                event_frames=event_frames,
+                group_id=self.event.group_id,
+                project_id=self.event.project_id,
+            )
+
+        assert mock_logger.info.call_count == 1
+        mock_logger.info.assert_called_with(
+            "process_commit_context.find_commit_context",
+            extra={
+                "event": self.event.event_id,
+                "group": self.event.group_id,
+                "organization": self.event.group.project.organization_id,
+                "reason": "could_not_fetch_commit_context",
+                "code_mappings_count": 1,
+                "fallback": True,
+            },
+        )
+
+    @patch(
+        "sentry.integrations.github.GitHubIntegration.get_commit_context",
+        return_value={
+            "commitId": "asdfasdf",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "admin@localhost",
@@ -170,7 +213,9 @@ def test_no_matching_commit_in_db(self, mock_get_commit_context):
         "sentry.integrations.github.GitHubIntegration.get_commit_context",
         return_value={
             "commitId": "asdfwreqr",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "admin@localhost",
@@ -255,7 +300,9 @@ def test_no_inapp_frame_in_stacktrace(self, mock_process_suspect_commits):
         "sentry.integrations.github.GitHubIntegration.get_commit_context",
         return_value={
             "commitId": "somekey",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "[email protected]",
@@ -296,7 +343,9 @@ def test_commit_author_not_in_sentry(self, mock_get_commit_context):
         "sentry.integrations.github.GitHubIntegration.get_commit_context",
         return_value={
             "commitId": "somekey",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "[email protected]",
@@ -337,7 +386,9 @@ def test_commit_author_no_user(self, mock_get_commit_context, mock_get_users_for
         "sentry.integrations.github.GitHubIntegration.get_commit_context",
         return_value={
             "commitId": "somekey",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "[email protected]",
@@ -423,7 +474,9 @@ def after_return(self, status, retval, task_id, args, kwargs, einfo):
     Mock(
         return_value={
             "commitId": "asdfwreqr",
-            "committedDate": "2023-02-14T11:11Z",
+            "committedDate": (datetime.now(tz=datetime_timezone.utc) - timedelta(days=7)).strftime(
+                "%Y-%m-%dT%H:%M:%SZ"
+            ),
             "commitMessage": "placeholder commit message",
             "commitAuthorName": "",
             "commitAuthorEmail": "admin@localhost",

tests/sentry/tasks/test_post_process.py

@@ -1303,7 +1303,9 @@ def test_issue_owners_should_ratelimit(self, logger):
 class ProcessCommitsTestMixin(BasePostProgressGroupMixin):
     github_blame_return_value = {
         "commitId": "asdfwreqr",
-        "committedDate": "",
+        "committedDate": (datetime.now(timezone.utc) - timedelta(days=2)).strftime(
+            "%Y-%m-%dT%H:%M:%SZ"
+        ),
         "commitMessage": "placeholder commit message",
         "commitAuthorName": "",
         "commitAuthorEmail": "admin@localhost",