BQT is a CLI tool designed to facilitate unit testing for BigQuery queries. It enables users to define mock data and anticipated outputs for their BigQuery queries, providing a simulated environment for accurate testing. BQT ensures that the actual query outputs align with the mocked outputs.
BQT does not need access to BigQuery cloud resources. It uses a small and fast simulator which you can run either locally or on your CI.
Usage example:
bqt test --tests tests_folder
test_folder
: is a folder with json
files defining tests.
bqt output tells you if there are failing tests:
/bqt test --tests tests_data
Parsing tests in: tests_data
Detected test: tests_data/test1.json
Detected test: tests_data/test2.json
Detected test: tests_data/test3.json
Detected test: tests_data/test4.json
Detected test: tests_data/test5.json
Parsed Tests: 5
Running Tests...
Running Test: simple_test : tests_data/test1.json
✅ Test Success: simple_test : tests_data/test1.json
Running Test: simple_test : tests_data/test2.json
✅ Test Success: simple_test : tests_data/test2.json
Running Test: simple_test : tests_data/test3.json
------Unexpected data-------
column2 : something2
v : 2
-------------
✅ Test Success: simple_test : tests_data/test3.json
Running Test: simple_test : tests_data/test4.json
------Unexpected data-------
column2 : something2
v : 2
-------------
------Missing Dataa-------
column2 : something5
v : 5
-------------
Error: Expected data is missing..
❌ Test Failed: simple_test : tests_data/test4.json
Running Test: simple_test : tests_data/test5.json
------Unexpected data-------
column2 : something
v : 1
-------------
------Unexpected data-------
column2 : something2
v : 2
-------------
------Missing Dataa-------
column2 : something5
v : 5
-------------
Error: Expected data is missing..
❌ Test Failed: simple_test : tests_data/test5.json
2023/09/23 16:22:26 Some tests failed
Test files
- BQT uses tests defined in json you can see examples in this repo's
tests_data
folder
sample test:
{
"name": "simple_test",
"file": "tests_data/test1.sql",
"mocks": {
"`dataset`.`table`": {
"filepath": "tests_data/test1_in1.csv",
"types": {
"c1": "int64"
}
}
},
"output": {
"filepath": "tests_data/out.csv",
"types": {
"column1": "string"
}
}
}
this test:
tests_data/test1.sql
this is the query being tested
dataset
.table
this table in the SQL query is mocked with data defined in tests_data/test1_in1.csv
- the output of this query has to match the data defined in
tests_data/out.csv
Installation
Nix
You can use the nix flake to build/run/develop this project:
nix profile install github:dav009/bqt
bqt --help
You can also use the flake to build/run:
nix build github:dav009/bqt
nix run github:dav009/bqt
Other
go install github.com/dav009/bqt/cmd/bqt@latest
Download from releases
- Download binary from Github releases
Details
- bqt runs a BQ simulator to run queries
- bqt does some simple string replacement to find table names and replace table references with mock data