This library is a set of utilities to work with Gherkin documents and AST.

The command-line tool can be used to format .feature files or translate .feature files into .feature.md files.

Example usage:

# Translate all `.feature` files to `.feature.md` files and delete the `.feature` files.
# See https://github.com/cucumber/common/blob/main/gherkin/MARKDOWN_WITH_GHERKIN.md
# Note that the globs must be quoted to prevent the shell from expanding the globs.
npx @cucumber/gherkin-utils format --move "features/**/*.feature" "features/**/*.feature.md"

More details:

npx @cucumber/gherkin-utils --help

This module can also be used as a library. It provides two main utilities, pretty and gherkinDocumentWalker.

This function takes a GherkinDocument as input and returns a pretty-printed representation in Gherkin or Markdown.

The GherkinDocumentWalker is a class for walking and filtering the AST produced by Gherkin after parsing a feature file. When running walkGherkinDocument on a GherkinDocument, it will produce a deep copy of the object.

It takes two arguments upon creation:

• filters: set of functions used to know if the walked elements are kept in the result. By default, all elements are kept.
• handlers: set of function that can be used to alter the produced elements.

Filtering keeps the meaning of the original GherkinDocument, which means:

• if a Background was present, it will always be in the Feature (or Rule)
• the kept scenarios will have the same steps and examples than the original

By default, all elements are accepted, which means that if you want to do filtering you should reject all other elements. To ease this, we also provide the rejectAllFilters.

Here's an example:

import { GherkinDocumentWalker, rejectAllFilters } from '@cucumber/gherkin-utils'

// Only keeps scenarios which name include 'magic'
const filter = new GherkinDocumentWalker({
  ...rejectAllFilters,
  ...{ acceptScenario: (scenario) => scenario.name.includes('magic') },
})

// Makes a list with all the scenario names
const allScenarioNames: string[] = []
const scenarioNameFinder = new GherkinDocumentWalker({}, {
  handleScenario: (scenario) => allScenarioNames.push(scenario.name),
})