These tools convert the MITI schema between 3 different formats:

1. yaml. The reference YAML specification, designed to be validated with Cerberus.
2. flatfields. A specification based on a flat table of fields.
3. frictionless. A Frictionless Data data package schema.

The translation is not perfect, as the schema languages are not equivalent.

Installation

The top level of this repository comprises a Python package, which is the standard structure for a Python source repository. To install from source, after git clone and cd mititools:

pip install .

Generate alternative schema

mititools generate --source-dir path-to-yaml/

This creates:

1. flatfields/fields.tsv
2. flatfields/tables.tsv
3. frictionless/datapackage.json
4. cv/* controlled vocabulary tables
5. example_fd_package/

A snapshot of the most recently generated version of artifacts 1-4 are committed as part of this repository. Item 5 is not.

Validation with FD

Frictionless Data (FD) is an extensive and modular system of high-level data management tools, with consistent bindings in Python, JavaScript, and the bash command-line.

frictionless validate example_fd_package/datapackage.json

Output:

# -----
# valid: biospecimen.tsv
# -----
# -----
# valid: cell.tsv
# -----

...

Description with FD

frictionless describe example_fd_package/cell.tsv

Output:

# --------
# metadata: example_fd_package/cell.tsv
# --------

dialect:
  delimiter: "\t"
encoding: utf-8
format: tsv
hashing: md5
name: cell
path: example_fd_package/cell.tsv
profile: tabular-data-resource
schema:
  fields:
    - name: data_file_id
      type: any
    - name: filename
      type: any
    - name: file_format
      type: any
    - name: parent_id
      type: any
    - name: software_and_version
      type: any
    - name: commit_sha
      type: any
    - name: comment
      type: any
    - name: header_size
      type: any
    - name: object_classes_included
      type: any
    - name: object_classes_description
      type: any
    - name: fov_size_x
      type: any
    - name: fov_size_y
      type: any
    - name: pyramid
      type: any
    - name: z_stack
      type: any
    - name: t_series
      type: any
    - name: type
      type: any
    - name: physical_size_x
      type: any
    - name: physical_size_xunit
      type: any
    - name: physical_size_y
      type: any
    - name: physical_size_yunit
      type: any
    - name: physical_size_z
      type: any
    - name: physical_size_zunit
      type: any
    - name: cell_state_level
      type: any
scheme: file