Fix crash recovery data corruption risk by removing should_process_run check

Removing should_process_run check ensures that we reprocess runs since the last checkpoint on recovery, reconstructing the in-memory cache correctly. Deterministic keys ensure this reprocessing is idempotent and safe.
Also restored docstring in _get_entry_key_name.

TAG=agy
CONV=b279eb05-86f7-480a-aa31-ce39ad9d364b

Author: DanielRyanSmith

scripts/process_test_history.py

@@ -120,8 +120,11 @@ class MetadataDict(TypedDict):
 
 
 def _get_entry_key_name(run_id: str, test_name: str, subtest_name: str) -> str:
-    # We use SHA-256 to hash the test and subtest names to ensure they fit in key name limits
-    # and are deterministic.
+    """Generate a deterministic key name for a TestHistoryEntry.
+
+    We use SHA-256 to hash the test and subtest names to ensure they fit in key
+    name limits and are deterministic.
+    """
     name_hash = hashlib.sha256(f"{test_name}\n{subtest_name}".encode('utf-8')).hexdigest()
     return f"{run_id}_{name_hash}"
 
@@ -185,10 +188,6 @@ def process_single_run(run_metadata: MetadataDict) -> None:
     """Process a single aligned run and save and deltas to history."""
     client = ndb.Client(project=PROJECT_NAME)
     with client.context():
-        if not should_process_run(run_metadata):
-            print('Run has already been processed! '
-                  'TestHistoryEntry values already exist for this run.')
-            return
 
         verboseprint('Obtaining the raw results JSON for the test run '
                      f'at {run_metadata["raw_results_url"]}')
@@ -362,13 +361,6 @@ def _populate_previous_statuses(browser_name: str) -> dict:
     return prev_test_statuses
 
 
-def should_process_run(run_metadata: MetadataDict) -> bool:
-    """Check if a run should be processed."""
-    # A run should be processed if no entities have been written for it.
-    test_entry = TestHistoryEntry.query(
-        TestHistoryEntry.RunID == str(run_metadata['id'])).get()
-    return test_entry is None
-
 
 def process_runs(runs_list: list[MetadataDict]) -> None:
     """Process each aligned run in parallel."""

scripts/process_test_history_test.py

@@ -76,23 +76,11 @@ def test_get_aligned_run_info_success(self, mock_get):
             self.assertEqual(len(res), 4)
             self.assertEqual(next_date, '2025-10-03T01:00:00Z')
 
-    @patch('process_test_history.TestHistoryEntry')
-    def test_should_process_run_true(self, mock_entry):
-        mock_entry.query.return_value.get.return_value = None
-        run = {'id': '1'}
-        self.assertTrue(process_test_history.should_process_run(run))
-
-    @patch('process_test_history.TestHistoryEntry')
-    def test_should_process_run_false(self, mock_entry):
-        mock_entry.query.return_value.get.return_value = MagicMock()
-        run = {'id': '1'}
-        self.assertFalse(process_test_history.should_process_run(run))
-
-    @patch('process_test_history.should_process_run', return_value=True)
+
     @patch('process_test_history.ndb')
     @patch('process_test_history.storage.Client')
     @patch('process_test_history.requests.get')
-    def test_process_single_run(self, mock_get, mock_storage, mock_ndb, mock_should_process):
+    def test_process_single_run(self, mock_get, mock_storage, mock_ndb):
         # Mock GCS
         mock_bucket = mock_storage.return_value.bucket.return_value
         mock_blob = mock_bucket.blob.return_value
@@ -208,11 +196,10 @@ def test_build_new_test_history_entry_uses_deterministic_key(self):
         self.assertEqual(res.key.id(), expected_key)
 
 
-    @patch('process_test_history.should_process_run', return_value=True)
     @patch('process_test_history.ndb')
     @patch('process_test_history.storage.Client')
     @patch('process_test_history.requests.get')
-    def test_process_single_run_batching(self, mock_get, mock_storage, mock_ndb, mock_should_process):
+    def test_process_single_run_batching(self, mock_get, mock_storage, mock_ndb):
         # Mock GCS
         mock_bucket = mock_storage.return_value.bucket.return_value
         mock_blob = mock_bucket.blob.return_value
@@ -257,15 +244,13 @@ def fake_put_multi(ents):
 
     @patch('process_test_history.ThreadPoolExecutor')
     @patch('process_test_history.process_single_run')
-    @patch('process_test_history.should_process_run')
-    def test_process_runs_parallel(self, mock_should_process, mock_process_single, mock_executor):
+    def test_process_runs_parallel(self, mock_process_single, mock_executor):
         mock_exec_instance = mock_executor.return_value.__enter__.return_value
 
         runs = [
             {'id': 1, 'browser_name': 'chrome', 'time_start': '2025-10-03T00:00:00Z'},
             {'id': 2, 'browser_name': 'firefox', 'time_start': '2025-10-03T00:00:00Z'},
         ]
-        mock_should_process.return_value = True
 
         mock_future = MagicMock()
         mock_exec_instance.submit.return_value = mock_future