valid8 interprets rules in a human-readable format (YAML) and applies those rules to a directory structure.

This package requires Python3.6+.

Clone the repository and jump in the directory, then:

$ pip install -r requirements.txt
$ pip install .

valid8 supports two subcommands:

• apply for most uses
• lint to check the validity of the rules file

Usage:

$ valid8 apply [--directory directory] rules.yml
$ valid8 lint rules.yml

Example rules.yml with one rule:

- rulename: count_lines_in_specified_files
  filters:
    path_list:
      - "Makefile"
      - "setup.py"
      - valid8
  actions:
    exists: true
    scripts:
      - wc -l ${FILEPATH}

Other examples can be found in the examples/ folder.

Each rule is described by two different concepts:

• Filters select the files to which the rule will be applied.
• Actions run checks on the files selected by the filters.

Each rule is considered valid if all actions taken on the files selected by the filters returned a 0 exit code (success).

N.B. Compatible pattern matching expressions: anything compatible with glob

The find filter is also available as syntactic sugar for the path, path_list and paths_from_file filters.

Using find with a direct argument calls path

  filters:
    find: find_this_file.txt

Using find with a YAML list calls path_list

  filters:
    find:
      - find_this_file.txt
      - also_these_files.*

Using find with a YAML mapping and the key file calls path_from_file

  filters:
    find:
      file: file_with_paths.txt

These modes are not combinable. Only use one per rule.

N.B. Only available in action.match and action.scripts

In some rules, it's useful to use information about the filtered files to make a determination on an action. The substitutions keys below are available for use in the parameter of the match and scripts action.

For example, the following will check that every file matching predictions/*/predictions.csv has a corresponding JSON file with the original directory name as its filename.

- rulename: predictions_file
  filters:
    path: "predictions/*/predictions.csv"
  actions:
    match: "pipelines/{DIR_NAME}.json"

Examples are based on the file path a/b/c.txt

1. Add the code in filters/ or actions/ respectively.
2. Each filter or action must be a function and must include the context parameter in the signature.
3. In the filters/ or actions/ __init__.py, add an import to add the new function under the filters or actions module. Name must be unique

License

The license is documented in the LICENSE file and on the NIST website.

Versions and releases:

See

• the repository tags for all releases. Gitlab link Github link
• the CHANGELOG file for a history of the releases.
• the __version__ field in valid8/__init__.py.

Contact:

Please send any issues, questions, or comments to [email protected]