Skip to main content

Working in Python

Frictionless Framework is a Python framework so it's naturally to be used within a Python programming environment.

Install#

To install the package please follow the Quick Start guide.

Import#

The package in Python is exposed as a single module called frictionless.

Core#

After installation you can import and use it to access the core functionality:

Python
import frictionless
frictionless.Schema
frictionless.Resource
frictionless.Package
# etc

Another option is to used named import:

Python
from frictionless import Schema, Resource, Package
Schema
Resource
Package
# etc

Frictionless has a reach API including high-level functions and classes. You can find a full list in the API Reference.

Plugins#

Frictionless ships with more than dozen builtin plugins. Usually you don't need to use them directly as it will be inferred and handled automatically, but, in some cases, you need to import their classes. For example:

Python
from frictionless.plugins.csv import CsvDialect
CsvDialect

Debug#

All the metadata classes inherit from the dict type so you can just print it:

Python
from pprint import pprint
from frictionless import Resource
resource = Resource('data/table.csv')
resource.infer()
pprint(resource)
# Prints metadata dict i.e. path, schema, etc

For tabular resource you can also use the resource.to_view() function for pretty-printing tables:

Python
from pprint import pprint
from frictionless import Resource
resource = Resource('data/table.csv')
pprint(resource.to_view())
# Prints a pretty-formatted data table
pprint(resource.read_rows())
# Of course, you can always just use `resource.read_rows()` or other reading functions:

Errors#

Frictionless uses a unifed error tree but there is only one exception frictionless.FrictionlessException so you need to catch this one and analyze an attached error:

Python
from pprint import pprint
from frictionless import Resource, FrictionlessException
try:
resource = Resource('bad.csv')
except FrictionlessException as exception:
pprint(exception.error)
# Prints the SchemaError metadata in this case

If you got some other exception using the Frictionless codebase it might be a bug so please report it to the Issue Tracker.

Hints#

Working with Frictionless in some modern Code Editor or interactive Notebook you can use code hints/completion capability. For example, there are many transform steps available, to find a right one just start typing:

Python
from frictionless import steps
steps.table_<TAB>
# Code completion will show the list of available table steps

The same works for class' and functions' arguments:

Python
from frictionless import Resource
Resource('data/table.csv, <TAB>
# Code completion will show the list of available arguments steps
Last updated on by roll