NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI, #4605 most downloaded on PyPI
Data management framework for Python that provides functionality to describe, extract, validate, and transform tabular data
Latest stable
5.21.0
Last release 1 months ago
15 Oct 2026
Release timing varies
gaps range from 9 days to 13 months
Rarely documented
notes for 13 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
6 years old
311 releases · first in 2020
Drag across the ruler to compare, or tap a release to open it.
Nothing published for this version
Nothing published for this version
Nothing published for this version
- Support --path CLI argument
--path CLI argument (#829)Your coding agent can read these notes before it upgrades. Set up the MCP server →
Nothing published for this version
Added support for Package(innerpath) argument for unzipping a data package's descriptor
Package(innerpath) argument for unzipping a data package's descriptorNothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Support control/dialect as JSON in CLI
Nothing published for this version
Implemented describe_dialect and describe(path, type="dialect")
describe_dialect and describe(path, type="dialect")--dialect argument in CLINothing published for this version
Nothing published for this version
Implemented Schema.from_jsonschema
Schema.from_jsonschema (#797)Use field.constraints.maxLength for SQL's VARCHAR
field.constraints.maxLength for SQL's VARCHAR (#795)Nothing published for this version
- Implemented resource.to_view()
resource.to_view() (#781)Nothing published for this version
Nothing published for this version
Make fields[].arrayItem errors more granular
fields[].arrayItem errors more granular (#767)Added support for fields[].arrayItem
fields[].arrayItem (#750)Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
- Released frictionless@4 :tada:
frictionless@4 :tada:Nothing published for this version
Nothing published for this version
Updated loaders (#658) (BREAKING)
filelike loader to stream loadertext loader to buffer loaderImprove transform API (#657) (BREAKING)
transform_resource(resource) signaturetransform_package(package) signatureImproved resource/package import/export (#655) (BREAKING)
parser.write_row_stream APIresource.from/to APIpackage.from/to APIStorage APIsystem.create_storage APIPandasStorage into PandasParserSpssStorage into SpssParserImproved transformation steps (#650) (BREAKING)
Improved layout and data streams concepts (#648) (BREAKING)
data_stream to list_streamreadData to readListssample to fragment (sample now is raw lists)Reworked Check API and split labels/header (BREAKING)
Rebased on Detector class (BREAKING)
Detector class (BREAKING)
Detectorresource.infer omit empty objectsresource.read_*(size) argumentresource.labels propertyImproved checks/steps API (#621) (BREAKING)
validate(extra_checks=[...]) to validate(checks=[{"code": 'code', ...}])Updated describe/extract/transform/validate APIs (BREAKING)
validate_table (use validate_resource)Table and File classesdataflows pluginnopool by parallel (not parallel by default)report.tables to report.tasksreport.tasks[].resource (instead of plain path/scheme/format/etc)Introduced Layout class (BREAKING)
Query class and arguments/properties to Layoutheader options from Dialect to Layout- Updated transform API - Added transform(type) argument
transform(type) argumentUpdated describe API (BREAKING)
describe(source_type) argument to typeRemoved extract_table (use extract_resource with the same API)
extract_table (use extract_resource with the same API)extract(source_type) argument to typeNothing published for this version
Initial API/codebase improvements for v4 (BREAKING)
Package/Resource(source) notation (guess descriptor/path/etc)schema.infer -> Schema.from_sampleresource.inline -> resource.memorycompression_path -> innerpathcompression: no -> compression: ""Package/Resource.infer not to infer stats (use stats=True)Package/Resource.infer(only_sample) argumentResouce.from/to_zip (use Package.from/to_zip)Resouce.source (use Resource.data or Resource.fullpath)package/resource.infer(source) argument (use constructors)Make Resource independent from Table/File (#607) (BREAKING)
resource.read_sample() to resource.sampleresource.read_header() to resource.headerresource.read_stats() to resource.statsresource.to_table()resource.to_file()Nothing published for this version
Nothing published for this version
Nothing published for this version
Optimize Row/Header/Table and rename header errors (#601) (BREAKING)
resource/table.read_data(_stream) now includes a header row if presenterrors.ExtraHeaderError->ExtraLabelError (extra-label-error)errors.MissingHeaderError->MissingLabelError (missing-label-error)errors.BlankHeaderError->BlankLabelError (blank-label-error)errors.DuplicateHeaderError->DuplicateLabelError (duplicate-label-error)errors.NonMatchingHeaderError->IncorrectLabelError (incorrect-label-error)schema.read/write_data->read/write_cellsRenamed aws plugin to s3 (#594) (BREAKING)
$ pip install frictionless[aws] # before
$ pip install frictionless[s3] # after
One column per quarter.