One-stop-shop for docs and test coverage of dbt projects.
Optimized for dbt 1.0, see full support matrix.
dbt-coverage is to dbt what coverage.py and interrogate are to Python.
It is a single CLI tool which checks your dbt project for missing documentation and tests.
Keeping documentation and tests close to the actual SQL code that generates the final model is one of the best design choices of dbt. It ensures documentation is actually useful and tests are actually used. But how do you make adding those a habit in your dbt project?
That is exactly where dbt-coverage comes in. It will
- Give you a better sense of the level of documentation and test coverage in your project;
- Help your CI/CD pipeline make sure new changes include documentation and tests;
- Let you quickly assess the documentation and tests of a new
dbtproject you get your hands on.
Still not convinced? Here are some more features:
- ✨ zero-config: just install it and run it, there is nothing to set up
- 🏁 minimal dependences: the only dependencies are
click(already installed withdbt) andtyper - 📦 very small: at ~480 SLOC, you can easily validate it works as advertised
The package was presented during Coalesce, the annual dbt conference, as a part of the talk From 100 spreadsheets to 100 data analysts: the story of dbt at Slido. Watch a demo in the video below.
pip install dbt-coverage
dbt-coverage comes with two basic commands: compute and compare. The
documentation for the individual commands can be shown by using the --help
option.
Compute coverage from target/catalog.json and target/manifest.json files
found in a dbt project, e.g.
jaffle_shop.
To choose between documentation and test coverage, pass doc or test as the CLI argument.
$ cd jaffle_shop
$ dbt run # Materialize models
$ dbt docs generate # Generate catalog.json and manifest.json
$ dbt-coverage compute doc --cov-report coverage-doc.json # Compute doc coverage, print it and write it to coverage-doc.json file
Coverage report
=====================================================================
jaffle_shop.customers 6/7 85.7%
jaffle_shop.orders 9/9 100.0%
jaffle_shop.raw_customers 0/3 0.0%
jaffle_shop.raw_orders 0/4 0.0%
jaffle_shop.raw_payments 0/4 0.0%
jaffle_shop.stg_customers 0/3 0.0%
jaffle_shop.stg_orders 0/4 0.0%
jaffle_shop.stg_payments 0/4 0.0%
=====================================================================
Total 15/38 39.5%
$ dbt-coverage compute test --cov-report coverage-test.json # Compute test coverage, print it and write it to coverage-test.json file
Coverage report
=====================================================================
jaffle_shop.customers 1/7 14.3%
jaffle_shop.orders 8/9 88.9%
jaffle_shop.raw_customers 0/3 0.0%
jaffle_shop.raw_orders 0/4 0.0%
jaffle_shop.raw_payments 0/4 0.0%
jaffle_shop.stg_customers 1/3 33.3%
jaffle_shop.stg_orders 2/4 50.0%
jaffle_shop.stg_payments 2/4 50.0%
=====================================================================
Total 14/38 36.8%You can also choose a subset of tables to compare using one or multiple --model-path-filter options.
$ cd jaffle_shop
$ dbt run # Materialize models
$ dbt docs generate # Generate catalog.json and manifest.json
$ dbt-coverage compute doc --cov-report coverage-doc.json --model-path-filter models/staging/ # Compute doc coverage for a subset of tables, print it and write it to coverage-doc.json file
Coverage report
======================================================
jaffle_shop.stg_customers 0/3 0.0%
jaffle_shop.stg_orders 0/4 0.0%
jaffle_shop.stg_payments 0/4 0.0%
======================================================
Total 0/11 0.0%
$ dbt-coverage compute doc --cov-report coverage-doc.json --model-path-filter models/orders.sql --model-path-filter models/staging/ # Compute doc coverage for a subset of tables, print it and write it to coverage-doc.json file
Coverage report
======================================================
jaffle_shop.orders 0/9 0.0%
jaffle_shop.stg_customers 0/3 0.0%
jaffle_shop.stg_orders 0/4 0.0%
jaffle_shop.stg_payments 0/4 0.0%
======================================================
Total 0/20 0.0%You can also choose to print the output in the Markdown table format by specifying the --cov-format option.
This can be especially useful when using dbt-coverage in CI/CD pipelines.
$ cd jaffle_shop
$ dbt run # Materialize models
$ dbt docs generate # Generate catalog.json and manifest.json
$ dbt-coverage compute doc --model-path-filter models/staging/ --cov-format markdown
# Coverage report
| Model | Columns Covered | % |
|:------|----------------:|:-:|
| jaffle_shop.stg_customers | 0/3 | 0.0% |
| jaffle_shop.stg_orders | 0/4 | 0.0% |
| jaffle_shop.stg_payments | 0/4 | 0.0% |
| Total | 0/11 | 0.0% |To compute the coverages, dbt-coverage looks up the artefacts from the dbt run execution in the
./target/ folder in the current directory. You can specify a custom path via the --run-artifacts-dir
option.
$ dbt-coverage compute doc --run-artifacts-dir jaffle_shop/target --cov-report coverage-doc.json # Compute doc coverage from the artefacts located in jaffle_shop/target, print it and write it to coverage-doc.json file
Coverage report
================================================
jaffle_shop.customers 0/7 0.0%
jaffle_shop.orders 0/9 0.0%
jaffle_shop.raw_customers 0/3 0.0%
jaffle_shop.raw_orders 0/4 0.0%
jaffle_shop.raw_payments 0/4 0.0%
jaffle_shop.stg_customers 0/3 0.0%
jaffle_shop.stg_orders 0/4 0.0%
jaffle_shop.stg_payments 0/4 0.0%
================================================
Total 0/38 0.0%Compare two coverage.json files generated by the compute command. This is
useful to ensure that the coverage does not drop while making changes to the
project.
$ dbt-coverage compare coverage-after.json coverage-before.json
# Coverage delta summary
before after +/-
=============================================
Coverage 39.47% 38.46% -1.01%
=============================================
Tables 8 8 +0/+0
Columns 38 39 +1/+0
=============================================
Hits 15 15 +0/+0
Misses 23 24 +1/+0
=============================================
# New misses
=========================================================================
Catalog 15/38 (39.47%) -> 15/39 (38.46%)
=========================================================================
- jaffle_shop.customers 6/7 (85.71%) -> 6/8 (75.00%)
-- new_col -/- (-) -> 0/1 (0.00%)
=========================================================================$ cd my-dbt-project
$ dbt run # Materialize models
$ dbt docs generate # Generate catalog.json and manifest.json
$ dbt-coverage compute doc --cov-report before.json --cov-fail-under 0.5 # Fail if coverage is lower than 50%
# Make changes to the dbt project, e.g. add some columns to the DWH, document some columns, etc.
$ dbt run # Materialize the changed models
$ dbt docs generate # Generate catalog.json and manifest.json
$ dbt-coverage compute doc --cov-report after.json --cov-fail-compare before.json # Fail if the current coverage is lower than coverage in before.json
$ dbt-coverage compare after.json before.json # Generate a detailed coverage delta reportDifferent version of dbt-coverage support different versions of dbt. Here is
the support matrix.
dbt |
dbt-coverage |
|---|---|
| <0.20 | not tested |
| 0.20 - 0.21 | 0.1 |
| 1.0 - 1.8 | 0.2, 0.3 |
- https://github.com/mikaelene/dbt-test-coverage
- interrogate (docs coverage for Python)
- coverage.py (execution coverage for Python)
Clone this repo including submodules, create a virtual environment and install dependencies:
git clone --recurse-submodules [email protected]:slidoapp/dbt-coverage.git
cd dbt-coverage
pip install poetry
poetry shell
poetry install
pre-commit installTo run all integration tests locally, run:
toxLicensed under the MIT license (see LICENSE.md file for more details).
