First published release of check-datapackage!
On November 27th, 2025, we published our second Python package to PyPI. This package forms the basis for ensuring that any metadata created or edited for a Data Package is correct and compliant with the Data Package standard. Since we are and will be working with and managing many Data Packages over the coming years, this is an important tool for us to have! Generally, this will be a helpful tool for anyone working with and managing Data Packages.
What’s check-datapackage?
As with all our packages and software tools, we have a dedicated website for check-datapackage. So, rather than repeat what is already in that website, this post gives a very quick overview of what this package does and why you might want to use it. It can be summarised by its tagline:
Ensure the compliance of your Data Package metadata
The “only” thing check-datapackage does is to check the content of a datapackage.json file against the Data Package standard. Nothing fancy. But we designed it to be configurable, so that if you have specific needs for your Data Package, you can adjust the checks accordingly. It’s possible to both add checks on top of the standard or ignore certain checks from the standard. For example, if you want to ensure that certain fields that aren’t required by the standard are always present in the metadata, you can set up the checks to enforce that.
For now, check-datapackage is only a few Python functions and classes that you can use within your own Python scripts. But in the future, we plan to develop a command-line interface (CLI) so that you can use it directly from your terminal without needing to write any code. Along with including a config file, we hope to incorporate check-datapackage into typical build tools and automated check workflows.
Why use it?
We wanted this package to be incredibly simple and focused. It also doesn’t include extra dependencies or features that you might not need. We wanted it lightweight and easy to use.
While there are a few tools that provide some type of checks of Data Packages, such as the frictionless-py package, we didn’t want all the extras that came with these packages. Nor are these tools easy to configure for our needs. In this regard, there were no tools available that fit ours needs. So, we built our own package that does exactly what we need. Hopefully, it will be useful for other people too!
Eventually, when we develop check-datapackage as a CLI, you could include it as a pre-commit hook or part of your continuous integration workflow so that every time you make changes to your Data Package metadata, it is automatically checked for compliance. That way, you will always know that your Data Package metadata lives up to the standard and your configuration.
Example use
We have a detailed guide on how to use check-datapackage. But we’ll briefly show how you might use check-datapackage. The main function of the package is check(), which takes as input the properties of a Data Package (i.e., the contents of the datapackage.json file) as a Python dictionary and checks it against the standard.
import check_datapackage as cdp
# Normally you'd read in the `datapackage.json` file, but we'll
# show the actual contents here as a Python dict. You can use
# the `read_json()` helper function to read in `datapackage.json`
properties = {
"name": "woolly-dormice",
"id": "123-abc-123",
"resources": [{
"name": "woolly-dormice-2015",
"path": "data.csv",
"schema": {"fields": [{
"name": "eye-colour",
"type": "string",
}]},
}],
}
cdp.check(properties)At a minimum, a Data Package needs to have a resources property. So in this case, there are no issues with the Data Package. But if you were to remove the resources property, which is required, and run the check again, there would be an issue:
del properties["resources"]
cdp.check(properties)If you want these checks to be treated as an error, you set the parameter error to True:
cdp.check(properties, error=True)If you want to exclude certain checks, you can do that by using the Config and Exclusion classes. For example, if you want to exclude all required checks, you can define the exclusion, add it to the configuration, and pass it to the check function like so:
exclusion_required = cdp.Exclusion(type="required")
config = cdp.Config(exclusions=[exclusion_required])
cdp.check(properties=package_properties, config=config)If you want the issues listed in a more human-friendly way, you can use the explain() function that takes the list of issues returned by check() and formats them nicely:
issues = cdp.check(properties)
cdp.explain(issues)There’s many other checks you can configure with check-datapackage, so be sure to check out the website for more information!