Make standards that can be exported/imported/validated via HDF5/JSON/XML/YAML.
mkstd uses pre-existing standards for validation and schema specification, such that files produced using mkstd can be used independently of mkstd. For example, a tool developer can use mkstd to create a standard for their tool's data, but users do not necessarily need mkstd installed to use the data. However, mkstd also provides importers and exporters, so intended use also involves an mkstd installation for convenience.
pip install mkstd
# For HDF5 support
pip install hdfdict@git+https://github.com/SiggiGue/hdfdictFor environments requiring numpy<2, replace mkstd with mkstd[numpyv1] above.
mkstd is intended to be used at two stages of data management. An example of these stages is provided in .
At this stage, the "user" is the person designing the data type and corresponding standard. For example, a tool developer who wants to standardize the data produced by their tool. The steps could look like:
- (with
mkstd) Design the data type as a Pydantic data model. Thanks to Pydantic, it behaves like a standard for your data, as a Python object. - (with
mkstd) Export the standard as e.g. XML and JSON schemas. - (TODO, with
mkstd) Generate documentation for the standard, based on the Pydantic data model docstrings.
At this stage, the "user" is someone who wants to use data generated by the tool, or import their own data into the tool.
- (with or without
mkstd) Reformat data to match the standard specified by the e.g.mkstd-generated XML or JSON schema. - (with or without
mkstd) Validate the data against the schema. - (with or without
mkstd) Import/export the reformatted data with the tool.
As written above, many uses of the standard produced by mkstd are intended to be possible without an mkstd installation. This is because the generated standards are in standardized schema formats. Below are the different formats supported by mkstd, and how to use/validate standards/data independently of mkstd.
The XSD format is used. Search the web for validate xml data against schema.
The official JSON schema format is used. Search the web for validate json data against schema.
There is no official YAML schema format, so YAML data is typically validated against JSON schemas. mkstd takes this approach too. Hence, tools that can validate YAML data against a JSON schema can be used, without an mkstd installation.
For example, the pajv tool can be used to validate YAML data against a JSON schema, without mkstd.
pajv validate -s output/mkstd_generated_schema.yaml -d output/data.yamlBy default, mkstd stores the schemas for YAML standards in YAML too.
There is currently no standard available for the specification of HDF5 schemas. Hence, the HDF5 files produced by mkstd can only be validated with mkstd.
There is a format for HDF5 that enables interconversion with JSON. This is out-of-scope.