google-pso-data-validator 7.5.0__tar.gz → 7.5.2__tar.gz

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: google-pso-data-validator
-Version: 7.5.0
+Version: 7.5.2
 Summary: A package to enable easy data validation
 Author: PSO DVT Engineering team
 Author-email: data-validator-eng@google.com
@@ -15,7 +15,7 @@ Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: Flask>=2.2
-Requires-Dist: fsspec>=2024.9.0
+Requires-Dist: fsspec!=2025.3.1
 Requires-Dist: google-api-python-client>=2.144.0
 Requires-Dist: google-cloud-bigquery>=3.25.0
 Requires-Dist: google-cloud-bigquery-storage>=2.26.0
@@ -42,7 +42,7 @@ Provides-Extra: develop
 Requires-Dist: black==22.3.0; extra == "develop"
 Requires-Dist: flake8; extra == "develop"
 Requires-Dist: freezegun; extra == "develop"
-Requires-Dist: pyfakefs==4.6.2; extra == "develop"
+Requires-Dist: pyfakefs; extra == "develop"
 Requires-Dist: pytest; extra == "develop"
 Requires-Dist: pytest-cov; extra == "develop"
 Requires-Dist: pytest-timeout; extra == "develop"

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/google_pso_data_validator.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: google-pso-data-validator
-Version: 7.5.0
+Version: 7.5.2
 Summary: A package to enable easy data validation
 Author: PSO DVT Engineering team
 Author-email: data-validator-eng@google.com
@@ -15,7 +15,7 @@ Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: Flask>=2.2
-Requires-Dist: fsspec>=2024.9.0
+Requires-Dist: fsspec!=2025.3.1
 Requires-Dist: google-api-python-client>=2.144.0
 Requires-Dist: google-cloud-bigquery>=3.25.0
 Requires-Dist: google-cloud-bigquery-storage>=2.26.0
@@ -42,7 +42,7 @@ Provides-Extra: develop
 Requires-Dist: black==22.3.0; extra == "develop"
 Requires-Dist: flake8; extra == "develop"
 Requires-Dist: freezegun; extra == "develop"
-Requires-Dist: pyfakefs==4.6.2; extra == "develop"
+Requires-Dist: pyfakefs; extra == "develop"
 Requires-Dist: pytest; extra == "develop"
 Requires-Dist: pytest-cov; extra == "develop"
 Requires-Dist: pytest-timeout; extra == "develop"

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/google_pso_data_validator.egg-info/requires.txt

@@ -1,5 +1,5 @@
 Flask>=2.2
-fsspec>=2024.9.0
+fsspec!=2025.3.1
 google-api-python-client>=2.144.0
 google-cloud-bigquery>=3.25.0
 google-cloud-bigquery-storage>=2.26.0
@@ -26,7 +26,7 @@ tabulate>=0.9.0
 black==22.3.0
 flake8
 freezegun
-pyfakefs==4.6.2
+pyfakefs
 pytest
 pytest-cov
 pytest-timeout

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/setup.py

@@ -18,7 +18,7 @@ import setuptools
 
 name = "google-pso-data-validator"
 description = "A package to enable easy data validation"
-version = "7.5.0"
+version = "7.5.2"
 release_status = "Development Status :: 4 - Beta"
 
 with open("README.md", "r") as fh:
@@ -26,7 +26,7 @@ with open("README.md", "r") as fh:
 
 dependencies = [
     "Flask>=2.2",  # Some versions of airflow such as 2.9.1 depend on flask<2.3 and >=2.2
-    "fsspec>=2024.9.0",
+    "fsspec!=2025.3.1",  # Version 2025.3.1 was breaking Python 3.8 support.
     "google-api-python-client>=2.144.0",
     "google-cloud-bigquery>=3.25.0",
     "google-cloud-bigquery-storage>=2.26.0",
@@ -54,7 +54,7 @@ extras_require = {
         "black==22.3.0",
         "flake8",
         "freezegun",
-        "pyfakefs==4.6.2",
+        "pyfakefs",
         "pytest",
         "pytest-cov",
         "pytest-timeout",

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/third_party/ibis/ibis_addon/operations.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-""" The Ibis Addons Operations are intended to help facilitate new expressions
+"""The Ibis Addons Operations are intended to help facilitate new expressions
 when required before they can be pushed upstream to Ibis.
 
 Raw SQL Filters:
@@ -24,6 +24,7 @@ non-textual languages.
 """
 import datetime
 import dateutil
+import numpy as np
 import string
 
 import google.cloud.bigquery as bq
@@ -94,6 +95,11 @@ except Exception:
     SnowflakeExprTranslator = None
 
 
+# Cast of datetime64 NaT to int64 and then in seconds results in the value below.
+# We need to use this value in the datetime.date simulation of the datetime64 behaviour.
+NAT_INT64_MIN_IN_SECONDS = np.iinfo(np.int64).min // 1_000_000_000
+
+
 class BinaryLength(ops.Value):
     arg = rlz.one_of([rlz.value(dt.Binary)])
     output_dtype = dt.int32
@@ -534,6 +540,10 @@ def _bigquery_field_to_ibis_dtype(field):
 def string_to_epoch(ts: str) -> int:
     """Function to convert string timestamp to epoch seconds"""
     try:
+        if pd.isna(ts):
+            # Casting datetime64 to int64 uses the minimum possible int64 when it
+            # encounters NaT. Simulating the same here for when auto cast fails.
+            return NAT_INT64_MIN_IN_SECONDS
         parsed_ts = dateutil.parser.isoparse(ts).astimezone(dateutil.tz.UTC)
         return (
             parsed_ts - datetime.datetime(1970, 1, 1, tzinfo=datetime.timezone.utc)
@@ -547,12 +557,11 @@ def string_to_epoch(ts: str) -> int:
 
 @execute_node.register(ops.ExtractEpochSeconds, (datetime.datetime, pd.Series))
 def execute_epoch_seconds_new(op, data, **kwargs):
-    import numpy as np
-
     convert = getattr(data, "view", data.astype)
     try:
         series = convert(np.int64)
-        return (series // 1_000_000_000).astype(np.int32)
+        # We need int64 below because NaT overflows int32.
+        return (series // 1_000_000_000).astype(np.int64)
     except TypeError:
         # Catch 'TypeError' for large timestamps beyond max datetime64[ns] as per Issue #1053
         # Cast to string instead to work around datetime64[ns] limitation

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/third_party/ibis/ibis_postgres/client.py

@@ -82,9 +82,11 @@ def _metadata(self, query: str) -> sch.Schema:
     )
     with self.begin() as con:
         cur = con.exec_driver_sql(f"SELECT * FROM {query} t0 LIMIT 0")
+        # Table OID below is cast to bigint because OID is based on
+        # an unsigned integer which can overflow integer data type.
         qry_cols = [
             f"('{column.name}'::text, {column.type_code},"
-            + f"{column.table_oid if column.table_oid else 'NULL'}::int,"
+            + f"{column.table_oid if column.table_oid else 'NULL'}::bigint,"
             + f"{column.table_column if column.table_column else 'NULL'}::int, {idx})"
             for idx, column in enumerate(cur.cursor.description)
         ]
@@ -93,7 +95,7 @@ def _metadata(self, query: str) -> sch.Schema:
                                 THEN format_type(t0.type_code, NULL)
                                 ELSE format_type(t1.atttypid, t1.atttypmod) END AS type
                     FROM UNNEST(array[{','.join(qry_cols)}])
-                    AS t0(name text, type_code int, attrelid int, attnum int, col_ord int)
+                    AS t0(name text, type_code int, attrelid bigint, attnum int, col_ord int)
                     LEFT JOIN pg_attribute t1 USING (attrelid, attnum) ORDER BY col_ord"""
         )
     yield from ((col, _get_type(typestr)) for col, typestr in type_info)

{google_pso_data_validator-7.5.0 → google_pso_data_validator-7.5.2}/third_party/ibis/ibis_teradata/registry.py

@@ -106,6 +106,21 @@ def teradata_cast_decimal_to_string(compiled_arg, from_, to):
     return "TO_CHAR({},'TM9')".format(compiled_arg)
 
 
+@teradata_cast.register(str, dt.Int8, dt.String)
+@teradata_cast.register(str, dt.Int16, dt.String)
+@teradata_cast.register(str, dt.Int32, dt.String)
+def teradata_cast_integer_to_string(compiled_arg, from_, to):
+    # INTEGER in a string cannot be longer than 10 characters (+1 for sign).
+    # Included smaller integers here too to minimize overloads.
+    return "CAST({} AS VARCHAR(11))".format(compiled_arg)
+
+
+@teradata_cast.register(str, dt.Int64, dt.String)
+def teradata_cast_bigint_to_string(compiled_arg, from_, to):
+    # BIGINT in a string cannot be longer than 20 characters (+1 for sign).
+    return "CAST({} AS VARCHAR(21))".format(compiled_arg)
+
+
 @teradata_cast.register(str, dt.Time, dt.String)
 def teradata_cast_time_to_string(compiled_arg, from_, to):
     # Time always has a time zone associated with it in Teradata