danger-plugin-tslint

Build Status npm version semantic-release Greenkeeper badge

Danger plugin for TSLint

Usage

Setup TSLint

This Danger plugin requires that you output the TSLint results as a JSON file before running danger on CI.

One way to do this is to use TSLint's JSON formatter and tee.

Given a package.json with a "lint" script:

{
  "scripts": {
    "lint": "tslint 'src/**/*.{ts,tsx}'"
  }
}

Running yarn run lint --silent -- --format json will only output the JSON results, which are piped into tee and written to disk in the reports/lint-results.json file.

# ci-script.sh

mkdir -p reports/
yarn run lint --silent -- --format json | tee reports/lint-results.json

yarn run danger

In this example, may also want to add the reports/ directory to your .gitignore file, as this file does not need to be checked into source control.

Setup Danger

Install:

yarn add danger-plugin-tslint --dev

At a glance:

// dangerfile.js
import path from 'path'
import tslint from 'danger-plugin-tslint'

// Handle TSLint results in `reports/lint-results.json` and leave a Danger comment on the PR
tslint({
  lintResultsJsonPath: path.resolve(__dirname, 'reports', 'lint-results.json'),
})

By default tslint() will use the defaultResultHandler in src/resultHandlers.ts. If you want to supply a custom result handler, which also requires you to call Danger functions like fail() and message() , you can do that too:

// dangerfile.js
import path from 'path'
import tslint from 'danger-plugin-tslint'

tslint({
  lintResultsJsonPath: path.resolve(__dirname, 'reports', 'lint-results.json'),
  handleResults: (results) => {
    if (results.length > 0) {
      const formattedResults = formatResults(results)
      fail(`TSLint failed\n\n${formattedResults}`)
    } else {
      message('👍 TSLint passed')
    }
  }
})

See the documentation for detailed information (and also check out src/index.ts).

Changelog

See the GitHub release history.

Development

Install Yarn, and install the dependencies - yarn install.

Run the Jest test suite with yarn test.

This project uses semantic-release for automated NPM package publishing.

The main caveat: instead of running git commit, run yarn commit and follow the prompts to input a conventional changelog message via commitizen.

:heart: