feat: search statistics (#1616)

chalmerlowe · gcf-owl-bot[bot] · web-flow · commit b930e4673b0d · 2023-09-02T05:11:13.000-04:00
* experimental tweaks * feat: adds two search statistics classes and property * removes several personal debugging sentinels * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * adds tests * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * cleans up conflict * adds comment * adds some type hints, adds a test for SearchReasons * cleans up some comments * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * Update tests/unit/job/test_query_stats.py * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * updated type checks to be isinstance checks per linter * update linting * Update tests/unit/job/test_query_stats.py * Update tests/unit/job/test_query_stats.py * experiments with some tests that are failing * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * Fix linting * update package verification approach * update pandas installed version constant * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * remove unused package * set pragma no cover * adds controls to skip testing if pandas exceeds 2.0 * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * adds pragma no cover to a simple check * add checks against pandas 2.0 on system test * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * experiments with some tests that are failing * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * resolves merge conflict * resolves merge conflict * resolve conflicts * resolve merge conflicts * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * updates due to faulty confict resolution * adds docstrings to two classes * corrects formatting * Update tests/unit/job/test_query_stats.py * Update tests/unit/job/test_query_stats.py * updates default values and corrects mypy errors * corrects linting * Update google/cloud/bigquery/job/query.py --------- Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com>
diff --git a/google/cloud/bigquery/job/query.py b/google/cloud/bigquery/job/query.py
@@ -198,6 +198,59 @@ def from_api_repr(cls, stats: Dict[str, str]) -> "DmlStats":
         return cls(*args)
 
 
+class IndexUnusedReason(typing.NamedTuple):
+    """Reason about why no search index was used in the search query (or sub-query).
+
+    https://cloud.google.com/bigquery/docs/reference/rest/v2/Job#indexunusedreason
+    """
+
+    code: Optional[str] = None
+    """Specifies the high-level reason for the scenario when no search index was used.
+    """
+
+    message: Optional[str] = None
+    """Free form human-readable reason for the scenario when no search index was used.
+    """
+
+    baseTable: Optional[TableReference] = None
+    """Specifies the base table involved in the reason that no search index was used.
+    """
+
+    indexName: Optional[str] = None
+    """Specifies the name of the unused search index, if available."""
+
+    @classmethod
+    def from_api_repr(cls, reason):
+        code = reason.get("code")
+        message = reason.get("message")
+        baseTable = reason.get("baseTable")
+        indexName = reason.get("indexName")
+
+        return cls(code, message, baseTable, indexName)
+
+
+class SearchStats(typing.NamedTuple):
+    """Statistics related to Search Queries. Populated as part of JobStatistics2.
+
+    https://cloud.google.com/bigquery/docs/reference/rest/v2/Job#searchstatistics
+    """
+
+    mode: Optional[str] = None
+    """Indicates the type of search index usage in the entire search query."""
+
+    reason: List[IndexUnusedReason] = []
+    """Reason about why no search index was used in the search query (or sub-query)"""
+
+    @classmethod
+    def from_api_repr(cls, stats: Dict[str, Any]):
+        mode = stats.get("indexUsageMode", None)
+        reason = [
+            IndexUnusedReason.from_api_repr(r)
+            for r in stats.get("indexUnusedReasons", [])
+        ]
+        return cls(mode, reason)
+
+
 class ScriptOptions:
     """Options controlling the execution of scripts.
 
@@ -724,7 +777,6 @@ def to_api_repr(self) -> dict:
             Dict: A dictionary in the format used by the BigQuery API.
         """
         resource = copy.deepcopy(self._properties)
-
         # Query parameters have an addition property associated with them
         # to indicate if the query is using named or positional parameters.
         query_parameters = resource["query"].get("queryParameters")
@@ -858,6 +910,15 @@ def priority(self):
         """
         return self.configuration.priority
 
+    @property
+    def search_stats(self) -> Optional[SearchStats]:
+        """Returns a SearchStats object."""
+
+        stats = self._job_statistics().get("searchStatistics")
+        if stats is not None:
+            return SearchStats.from_api_repr(stats)
+        return None
+
     @property
     def query(self):
         """str: The query text used in this query job.
diff --git a/tests/unit/job/test_query.py b/tests/unit/job/test_query.py
@@ -911,6 +911,28 @@ def test_dml_stats(self):
         assert isinstance(job.dml_stats, DmlStats)
         assert job.dml_stats.inserted_row_count == 35
 
+    def test_search_stats(self):
+        from google.cloud.bigquery.job.query import SearchStats
+
+        client = _make_client(project=self.PROJECT)
+        job = self._make_one(self.JOB_ID, self.QUERY, client)
+        assert job.search_stats is None
+
+        statistics = job._properties["statistics"] = {}
+        assert job.search_stats is None
+
+        query_stats = statistics["query"] = {}
+        assert job.search_stats is None
+
+        query_stats["searchStatistics"] = {
+            "indexUsageMode": "INDEX_USAGE_MODE_UNSPECIFIED",
+            "indexUnusedReasons": [],
+        }
+        # job.search_stats is a daisy-chain of calls and gets:
+        # job.search_stats << job._job_statistics << job._properties
+        assert isinstance(job.search_stats, SearchStats)
+        assert job.search_stats.mode == "INDEX_USAGE_MODE_UNSPECIFIED"
+
     def test_result(self):
         from google.cloud.bigquery.table import RowIterator
 
diff --git a/tests/unit/job/test_query_stats.py b/tests/unit/job/test_query_stats.py
@@ -108,6 +108,75 @@ def test_from_api_repr_full_stats(self):
         assert result.updated_row_count == 4
 
 
+class TestSearchStatistics:
+    @staticmethod
+    def _get_target_class():
+        from google.cloud.bigquery.job.query import SearchStats
+
+        return SearchStats
+
+    def _make_one(self, *args, **kwargs):
+        return self._get_target_class()(*args, **kwargs)
+
+    def test_ctor_defaults(self):
+        search_stats = self._make_one()
+        assert search_stats.mode is None
+        assert search_stats.reason == []
+
+    def test_from_api_repr_unspecified(self):
+        klass = self._get_target_class()
+        result = klass.from_api_repr(
+            {"indexUsageMode": "INDEX_USAGE_MODE_UNSPECIFIED", "indexUnusedReasons": []}
+        )
+
+        assert isinstance(result, klass)
+        assert result.mode == "INDEX_USAGE_MODE_UNSPECIFIED"
+        assert result.reason == []
+
+
+class TestIndexUnusedReason:
+    @staticmethod
+    def _get_target_class():
+        from google.cloud.bigquery.job.query import IndexUnusedReason
+
+        return IndexUnusedReason
+
+    def _make_one(self, *args, **kwargs):
+        return self._get_target_class()(*args, **kwargs)
+
+    def test_ctor_defaults(self):
+        search_reason = self._make_one()
+        assert search_reason.code is None
+        assert search_reason.message is None
+        assert search_reason.baseTable is None
+        assert search_reason.indexName is None
+
+    def test_from_api_repr_unspecified(self):
+        klass = self._get_target_class()
+        result = klass.from_api_repr(
+            {
+                "code": "INDEX_CONFIG_NOT_AVAILABLE",
+                "message": "There is no search index...",
+                "baseTable": {
+                    "projectId": "bigquery-public-data",
+                    "datasetId": "usa_names",
+                    "tableId": "usa_1910_current",
+                },
+                "indexName": None,
+            }
+        )
+
+        assert isinstance(result, klass)
+        assert result.code == "INDEX_CONFIG_NOT_AVAILABLE"
+        assert result.message == "There is no search index..."
+        assert result.baseTable == {
+            "projectId": "bigquery-public-data",
+            "datasetId": "usa_names",
+            "tableId": "usa_1910_current",
+        }
+        assert result.indexName is None
+
+
 class TestQueryPlanEntryStep(_Base):
     KIND = "KIND"
     SUBSTEPS = ("SUB1", "SUB2")
