JSHint JUnit Reporter

A JSHint output report that returns results compatible with JUnit XML. This makes it possible to integrate the results into any reporting framework that accepts that format. I have found this particularly useful in continuous integration scenarios (Bamboo, Jenkins, etc.)

The entire JSHint run is considered a test suite and each file with failures is a test case. A failure node is added to each test case indicating the number of linting errors for that test case. The body of that node enumerates the messages.

<?xml version="1.0" encoding="utf-8"?>
<testsuite name="jshint" tests="1" failures="3" errors="0" >
	<testcase name="test.js">
		<failure message="3 JSHint Failures" >
1. line 3, char 5: Missing &quot;use strict&quot; statement.
2. line 4, char 2: Missing semicolon.
3. line 1, char 1: &apos;x&apos; is not defined.
		</failure>
	</testcase>
</testsuite>

Installation

Download the file directly on install using NPM:

npm install jshint-junit-reporter

Usage

Pass the path to reporter.js to the JSHint --reporter option like this:

jshint --reporter=reporter.js reporter.js

You can also use this plugin with the grunt-contrib-jshint plugin that support the reporter option using something like this in your options object:

options: {
	reporter: require("jshint-junit-reporter"),
	reporterOutput: "junit-output.xml"
}

Note: To use this option you should be on grunt-contrib-jshint >0.5.3.

Bamboo Integration

In order to hook this into Bamboo I have a 3 step test-plan:

1. Checkout Code
2. Build Script
3. Read JUnit XML Files

Build Script

grunt jshint

---

I also fill in the Working Sub-directory field with the path to my repo, which my case is grunt.

JUnit Parser Configuration

grunt/junit-output.xml

Again, change this to whatever works for your naming convention. With the newer verison of grunt you can just reference the single file and it should work fine!

Results

Limitations

The reporter API provided by JSHint provides access only to failure information. Therefore, the resulting XML will only list test cases for files that contained failures.

The number of tests containing linting failures will be reflected in the "tests" attribute of the testsuite element. The total number of failures in those files will be reflected in the "failures" attribute.

In the happy case that there are no failures an empty test case will be created for you.