As it's described in the Configuration section it's possible to provide custom inquiries in
.github/frictionless.yaml file. An Inquiry is a Frictionless Framework concept used to define a validation job.
The simplest way to customize an inquiry is to set validation paths:
tasks: - path: data/table.csv - path: data/table.xls
A task can have many parameters accepted by the Resource class so you can provide scheme, format, etc:
tasks: - path: data/table.csv.zip scheme: file format: csv encoging: utf-8 compression: zip
By default, Fricitonless tries to infer all these parameters.
Let's provide a dialect and a schema as it's described in Resource:
tasks: - path: table.csv dialect: headerRows: [1,2] csv: separator: ; schema: fields: - name: currency type: string - name: rate type: number groupChar: ',' constraints: maximum: 100
You can also provide dialect, or schema as a file path e.g.
It's possible to configure how the validation happens as it's in Checklist:
tasks: - path: table.csv checklsit: pickErrors: ['#header'] skipErrors: ['#row'] checks: - type: ascii-value - type: row-constraint formula: id>1
Please consult with Validation Checks for available checks.
It's possible to validate a data resource:
tasks: - resource: data/dataresource.json
Read more about Resource in the Frictionless Framework docs.
It's possible to validate a data package:
tasks: - package: data/datapackage.json
Read more about Package in the Frictionless Framework docs.