Skip to content

Commit

Permalink
refactor: split pandas system tests to new module (#548)
Browse files Browse the repository at this point in the history
Thank you for opening a Pull Request! Before submitting your PR, there are a few things you can do to make sure it goes smoothly:
- [ ] Make sure to open an issue as a [bug/issue](https://github.com/googleapis/python-bigquery/issues/new/choose) before writing your code!  That way we can discuss the change, evaluate designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)

Follow-up to #448

Towards #366
  • Loading branch information
tswast authored Mar 15, 2021
1 parent 0d7212c commit efdf1c6
Show file tree
Hide file tree
Showing 4 changed files with 969 additions and 918 deletions.
39 changes: 39 additions & 0 deletions tests/system/conftest.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,39 @@
# Copyright 2021 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import pytest

from . import helpers


@pytest.fixture(scope="session")
def bigquery_client():
from google.cloud import bigquery

return bigquery.Client()


@pytest.fixture(scope="session")
def bqstorage_client(bigquery_client):
from google.cloud import bigquery_storage

return bigquery_storage.BigQueryReadClient(credentials=bigquery_client._credentials)


@pytest.fixture
def dataset_id(bigquery_client):
dataset_id = f"bqsystem_{helpers.temp_suffix()}"
bigquery_client.create_dataset(dataset_id)
yield dataset_id
bigquery_client.delete_dataset(dataset_id, delete_contents=True)
94 changes: 94 additions & 0 deletions tests/system/helpers.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,94 @@
# Copyright 2021 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import datetime
import decimal
import uuid

import google.api_core.exceptions
import test_utils.retry

from google.cloud._helpers import UTC


_naive = datetime.datetime(2016, 12, 5, 12, 41, 9)
_naive_microseconds = datetime.datetime(2016, 12, 5, 12, 41, 9, 250000)
_stamp = "%s %s" % (_naive.date().isoformat(), _naive.time().isoformat())
_stamp_microseconds = _stamp + ".250000"
_zoned = _naive.replace(tzinfo=UTC)
_zoned_microseconds = _naive_microseconds.replace(tzinfo=UTC)
_numeric = decimal.Decimal("123456789.123456789")


# Examples of most data types to test with query() and DB-API.
STANDARD_SQL_EXAMPLES = [
("SELECT 1", 1),
("SELECT 1.3", 1.3),
("SELECT TRUE", True),
('SELECT "ABC"', "ABC"),
('SELECT CAST("foo" AS BYTES)', b"foo"),
('SELECT TIMESTAMP "%s"' % (_stamp,), _zoned),
('SELECT TIMESTAMP "%s"' % (_stamp_microseconds,), _zoned_microseconds,),
('SELECT DATETIME(TIMESTAMP "%s")' % (_stamp,), _naive),
('SELECT DATETIME(TIMESTAMP "%s")' % (_stamp_microseconds,), _naive_microseconds,),
('SELECT DATE(TIMESTAMP "%s")' % (_stamp,), _naive.date()),
('SELECT TIME(TIMESTAMP "%s")' % (_stamp,), _naive.time()),
('SELECT NUMERIC "%s"' % (_numeric,), _numeric),
("SELECT (1, 2)", {"_field_1": 1, "_field_2": 2}),
(
"SELECT ((1, 2), (3, 4), 5)",
{
"_field_1": {"_field_1": 1, "_field_2": 2},
"_field_2": {"_field_1": 3, "_field_2": 4},
"_field_3": 5,
},
),
("SELECT [1, 2, 3]", [1, 2, 3]),
(
"SELECT ([1, 2], 3, [4, 5])",
{"_field_1": [1, 2], "_field_2": 3, "_field_3": [4, 5]},
),
(
"SELECT [(1, 2, 3), (4, 5, 6)]",
[
{"_field_1": 1, "_field_2": 2, "_field_3": 3},
{"_field_1": 4, "_field_2": 5, "_field_3": 6},
],
),
(
"SELECT [([1, 2, 3], 4), ([5, 6], 7)]",
[{"_field_1": [1, 2, 3], "_field_2": 4}, {"_field_1": [5, 6], "_field_2": 7}],
),
("SELECT ARRAY(SELECT STRUCT([1, 2]))", [{"_field_1": [1, 2]}]),
("SELECT ST_GeogPoint(1, 2)", "POINT(1 2)"),
]


def temp_suffix():
now = datetime.datetime.now()
return f"{now.strftime('%Y%m%d%H%M%S')}_{uuid.uuid4().hex[:8]}"


def _rate_limit_exceeded(forbidden):
"""Predicate: pass only exceptions with 'rateLimitExceeded' as reason."""
return any(error["reason"] == "rateLimitExceeded" for error in forbidden._errors)


# We need to wait to stay within the rate limits.
# The alternative outcome is a 403 Forbidden response from upstream, which
# they return instead of the more appropriate 429.
# See https://cloud.google.com/bigquery/quota-policy
retry_403 = test_utils.retry.RetryErrors(
google.api_core.exceptions.Forbidden, error_predicate=_rate_limit_exceeded,
)
Loading

0 comments on commit efdf1c6

Please sign in to comment.