Merge pull request #706 from spectacles-ci/josh/eng-84-profiling-on-c…

…ore-does-not-work

Fix profiling and open it back up to all queries

spectacles/runner.py

@@ -399,9 +399,7 @@ async def validate_sql(
         )
 
         async with self.branch_manager(ref=ref):
-            await validator.search(
-                explores, fail_fast, chunk_size, profile=profile if fail_fast else False
-            )
+            await validator.search(explores, fail_fast, chunk_size, profile=profile)
 
         # Create dimension tests for the desired ref when explores errored
         if not fail_fast and incremental:

spectacles/validators/sql.py

@@ -16,7 +16,7 @@
 DEFAULT_CHUNK_SIZE = 500
 DEFAULT_QUERY_CONCURRENCY = 10
 DEFAULT_RUNTIME_THRESHOLD = 5
-ProfilerTableRow = Tuple[str, float, str, str]
+ProfilerTableRow = Tuple[str, str, float, str, str]
 
 
 @dataclass
@@ -62,7 +62,13 @@ def to_profiler_format(self) -> ProfilerTableRow:
                 "Query.explore_url cannot be None, "
                 "run Query.create to get an explore URL"
             )
-        return (self.explore.name, self.runtime, self.query_id, self.explore_url)
+        return (
+            self.explore.name,
+            self.dimensions[0].name if len(self.dimensions) == 1 else "*",
+            self.runtime,
+            self.query_id,
+            self.explore_url,
+        )
 
 
 def print_profile_results(queries: List[Query], runtime_threshold: int) -> None:
@@ -79,6 +85,7 @@ def print_profile_results(queries: List[Query], runtime_threshold: int) -> None:
             [query.to_profiler_format() for query in queries_by_runtime],
             headers=[
                 "Explore",
+                "Dimension(s)",
                 "Runtime (s)",
                 "Query ID",
                 "Explore From Here",
@@ -315,6 +322,7 @@ async def _get_query_results(
                             Union[CompletedQueryResult, ErrorQueryResult], query_result
                         )
                         query = self._task_to_query[task_id]
+                        query.runtime = query_result.runtime
                         if query_result.runtime > self.runtime_threshold:
                             self._long_running_queries.append(query)
                         if query_result.status == "complete":

tests/integration/test_runner.py

@@ -191,3 +191,11 @@ async def test_incremental_sql_with_diff_explores_and_invalid_existing_sql_shoul
     assert result["tested"][1]["explore"] == "users__fail"
     assert result["tested"][1]["status"] == "passed"
     assert len(result["errors"]) == 0
+
+
+async def test_validate_sql_with_query_profiler_should_work(
+    looker_client: LookerClient, caplog: pytest.LogCaptureFixture
+):
+    runner = Runner(looker_client, "eye_exam")
+    await runner.validate_sql(fail_fast=True, profile=True, runtime_threshold=0)
+    assert "Query profiler results" in caplog.text

tests/unit/test_query.py

@@ -70,4 +70,25 @@ def test_query_should_convert_to_profiler_format(
         query_id=query_id,
         explore_url=explore_url,
     )
-    assert query.to_profiler_format() == (explore.name, runtime, query_id, explore_url)
+    assert query.to_profiler_format() == (
+        explore.name,
+        "*",
+        runtime,
+        query_id,
+        explore_url,
+    )
+
+    query = Query(
+        explore=explore,
+        dimensions=(dimension,),
+        runtime=runtime,
+        query_id=query_id,
+        explore_url=explore_url,
+    )
+    assert query.to_profiler_format() == (
+        explore.name,
+        dimension.name,
+        runtime,
+        query_id,
+        explore_url,
+    )