sqvid 0.1.5

SQVID: Simple sQl Validator of varIous Datasources

SQVID, the Simple sQl Validator of varIous Datasources is a framework for validating any type of data source that can be queried via SQL with the help of SQLAlchemy. It aims to be a simplified and extensible counterpart to validation of dbt models or data assertions of dataform that does not require you to use the full dbt or dataform and still ensure your data is automatically validated to be what you expect it to be. This allows SQVID to be used on all sorts of data sources: from CSVs and spreadsheets to massive databases.

You can easily use SQVID to serve as a “sanity check” of your processing pipeline or as a testing framework for your various ETL processes.

Installation

pip install sqvid

Example

Let us consider a database table called suppliers that would result from executing the following code snippet in a SQLite database called test_sqvid_db.

CREATE TABLE `suppliers` (
  `SupplierID` int NOT NULL,
  `SupplierName` varchar(255) DEFAULT NULL,
  `ContactName` varchar(255) DEFAULT NULL,
  `Address` varchar(255) DEFAULT NULL,
  `City` varchar(255) DEFAULT NULL,
  `PostalCode` varchar(255) DEFAULT NULL,
  `Country` varchar(255) DEFAULT NULL,
  `Phone` varchar(255) DEFAULT NULL
);

INSERT INTO `suppliers` (`SupplierID`, `SupplierName`, `ContactName`, `Address`, `City`, `PostalCode`, `Country`, `Phone`) VALUES
(1, "Exotic Liquid", "Charlotte Cooper", "49 Gilbert St.", "Londona", "EC1 4SD", "UK", "[phone]"),
(2, "New Orleans Cajun Delights", "Shelley Burke", "P.O. Box 78934", "New Orleans", "70117", "USA", "[phone]"),
(3, "Grandma Kelly's Homestead", "Regina Murphy", "707 Oxford Rd.", "Ann Arbor", "48104", "USA", "[phone]"),
(4, "Tokyo Traders", "Yoshi Nagase", "9-8 Sekimai Musashino-shi", "Tokyo", "100", "Japan", "[phone]"),
(5, "Cooperativa de Quesos 'Las Cabras'", "Antonio del Valle Saavedra", "Calle del Rosal 4", "Oviedo", "33007", "Spain", "(98) 598 76 54"),

In order to validate that this table contains the data we would expect it to, we can put together the following SQVID validation config:

[general]
sqla = "sqlite:///test_sqvid_db.sqlite"
db_name = 'test_sqvid_db'

[[test_sqvid_db.suppliers.SupplierID]]
validator = 'unique'

[[test_sqvid_db.suppliers.SupplierID]]
validator = 'in_range'
args = {min = 1, max = 256}

Note that the the validation config file is formated using TOML – you can find a very nice tutorial on this formatting language at LearnXinYMinutes.

The [general] section specifies SQLAlchemy connection string in sqla and the name of the DB that is going to have its data validated in db_name.

The other sections specify the various validations performed on this DB. In particular the table suppliers and data in its column SupplierID is being validated via two validators: the unique validator ensures that each value in this column occurs only once and the in_range validator checks whether all data points in this column fall withing the min and max range specified via parameters of the same name in args.

Once we save a validation config like this one into a file (say validate_suppliers.toml), SQVID validation tests can be invoked in the following way:

sqvid --config ./validate_suppliers.toml

This should provide output close to the following:

PASSED: Validation on [test_sqvid_db] suppliers.SupplierID of unique
PASSED: Validation on [test_sqvid_db] suppliers.SupplierID of in_range({'min': 1, 'max': 256})

Since all tests passed, sqvid would finish with exit code 0.

Failing validations

What happens when a SQVID validation test fails? We can easily see that by slightly changing the config file from the above:

[general]
sqla = "sqlite:///test_sqvid_db.sqlite"
db_name = 'test_sqvid_db'

[[test_sqvid_db.suppliers.SupplierID]]
validator = 'unique'

[[test_sqvid_db.suppliers.SupplierID]]
validator = 'in_range'
args = {min = 3, max = 256}

Note that the contents stayed the same, except for the final line where the min parameter has been set to 3. If we now save this file (to say ./validate_suppliers_fail.toml), we can again execute SQVID tests in a similar way:

sqvid --config ./validate_suppliers_fail.toml

The output should change to something like this:

PASSED: Validation on [test_sqvid_db] suppliers.SupplierID of unique
FAILED: Validation on [test_sqvid_db] suppliers.SupplierID of in_range({'min': 3, 'max': 256})
Offending 2 rows:
+--------------+------------------------------+--------------------+------------------+---------------+--------------+-----------+------------------+
|  SupplierID  |  SupplierName                |  ContactName       |  Address         |  City         |  PostalCode  |  Country  |  Phone           |
+--------------+------------------------------+--------------------+------------------+---------------+--------------+-----------+------------------+
|           1  |  Exotic Liquid               |  Charlotte Cooper  |  49 Gilbert St.  |  Londona      |  EC1 4SD     |  UK       |  [phone]  |
|           2  |  New Orleans Cajun Delights  |  Shelley Burke     |  P.O. Box 78934  |  New Orleans  |  70117       |  USA      |  [phone]  |
+--------------+------------------------------+--------------------+------------------+---------------+--------------+-----------+------------------+

As we would expect, the unique validation still passed while the in_range validation failed on the two rows which have their SupplierID outside of the [3, 256] range.

Since some tests failed, sqvid would finish with exit code 1.

Tests

As this project makes use of Poetry, after installing it the tests can be ran by executing the following from the project’s root directory:

poetry run pytest

They can also be ran with coverage:

poetry run pytest --cov=sqvid

License

Copyright 2019 Marek “mr.Shu” Suppa

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

http://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.