Migration

Frictionless is a logical continuation of many existing packages created for Frictionless Data as though datapackage or tableschema. Although, most of these packages will be supported going forward, you can migrate to Frictionless, which is Python 3.8+, as it improves many aspects of working with data and metadata. This document also covers migration from one framework's version to another.

From v4 to v5

Since the initial Frictionless Framework release we'd been collecting feedback and analyzing both high-level users' needs and bug reports to identify shortcomings and areas that can be improved in the next version of the framework. Read about a new version of the framework and migration details in this blog:

• Welcome Frictionless Framework (v5)

From dataflows

Frictionless Framework provides the frictionless transform function for data transformation. It can be used to migrate from dataflows or datapackage-pipelines:

• Transforming Data
• Transform Steps

From goodtables

Frictionless Framework provides the frictionless validate function which is in high-level exactly the same as goodtables validate. Also frictionless describe is an improved version of goodtables init. You instead need to use the frictionless command instead of the goodtables command:

• Validating Data
• Validation Checks
• Validation Errors

From datapackage

Frictionless Framework has Package and Resource classes which is almost the same as datapackage has:

• Describing Data
• Extracting Data
• Package Class
• Resource Class

From tableschema

Frictionless Framework has Schema and Field classes which is almost the same as tableschema has:

• Describing Data
• Extracting Data
• Schema Class
• Tabular Fields

From tabulator

Frictionless has Resource class which is an equivalent of the tabulator's Stream class:

• Extracting Data
• Resource Class
• File Schemes
• File Formats