chrome-coverage
v0.0.1
Published
Tiny, powerful data binding & web application framework
Downloads
8
Maintainers
schrotieReadme
Parse and evaluate coverage reports exported by Chrome browser
chrome-coverage
Chrome's coverage reports are in a rather raw/low-level JSON format. This small library evaluates these reports, provides a higher level format, generates a small HTML report of the coverage and generates Mocha tests that assert given coverage ratios to be met.
I developed this tool in conjunction with
my fork of
mocha-headless-chrome. Mocha headless Chrome allows you to run frontend mocha tests in the
console on a headless Chrome, and my fork allows you to export Chrome's
coverage report from such runs. You can use chrome-coverage without this,
though. Just manually export Chrome's coverage report from its developer tools
and use chrome-coverage to process it however you like.
The following example assumes that you use mocha-headless-chrome with coverage
export.
Example
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>Test</title>
<link rel="stylesheet" href="../node_modules/mocha/mocha.css" />
<script src="../node_modules/mocha/mocha.js"></script>
<script src="../node_modules/chai/chai.js"></script>
<script type="importmap">
{ "imports": {"quary": "../node_modules/quary/quary.mjs"}}
</script>
</head>
<body>
<div id="mocha"></div>
<div id="test"></div>
<script type="module">
import chromeCoverage from '../node_modules/chrome-coverage/chrome-coverage.min.mjs';
mocha.setup('bdd');
chai.should();
(async function() {
if(window.puppeteerStartCoverage) await window.puppeteerStartCoverage();
await loadTests();
chromeCoverage({getCoverage});
mocha.checkLeaks();
mocha.run();
})();
async function loadTests() {
const directoryResponse = await fetch('/test')
const directoryText = await directoryResponse.text();
const testScripts = directoryText.match(/([\w-]+)\.mjs/g);
return Promise.all(testScripts.map(file => import(`./${file}`)));
}
function getCoverage() {
if(window.puppeteerStopCoverage) {
return window.puppeteerStopCoverage({jsPath: 'jsCoverage.json'})
.then(({js}) => js);
}
else return fetch('../jsCoverage.json').then(res => res.text());
}
</script>
</body>
</html>You can run this from the console, e.g.
./node_modules/.bin/mocha-headless-chrome -f http://localhost:3000/test/test.htmlIt will then save the JS coverage report to jsCoverage.json and add a mocha
test to cover 100% of the code loaded in the tests.
You can also run this in the browser (e.g. by pointing your browser to the above
URL). Then it will load a previously saved jsCoverage.json, run the
abovementioned mocha tests and display a report of the coverage of all loaded
files. This report let's you click files and you'll see the actual uncovered
code - better than in Chrome where only whole lines may or may not be marked.
API
chrome-coverage provides three things:
- a method to transform Chrome's coverage report
- a method to generate Mocha tests and a coverage report
- a web component to render its coverage report
You can use these independently, though the Mocha test generator wraps it all together.
process
import {process} from './node_modules/chrome-coverage/chrome-coverage.mjs';
fetch('coverage.json').then(res => res.text()).then(process).then(doSomething);process has the following signature:
function process(coverage, {filter = defaultFilter, files = []} = {}) {/* ... */}coverage may be either the JSON string as in the example above, or the parsed
JSON. The optional options object may contain a filter function and/or an
expected files array.
filter
coverage - Chrome's coverage report - is an array of files with the respective
coverage information. process first filters that array by doing:
coverage.filter(filter)Thus the filter function will get one argument: the current element of the
coverage array. Usually you'll likely only care for the url of the files:
function filter({url}) { // .... Chrome's coverage report contains
everything that is loaded, e.g. the mocha library, your tests code ...
usually you only care about your source code's coverage and thus your filter
function should return true for your source code files.
By default process will filter out files from the src directory.
files
If your tests don't load some of your source files, Chrome will not see them and they'll not be included in any coverage report or test. In order to be sure not to miss some files, provide an array of filenames here. These will be checked against the actually covered files and added to the array if missing.
chrome-coverage
chrome-coverage if a web component, that you may add to you HTML:
<!DOCTYPE html>
<html><body>
<chrome-coverage></chrome-coverage>
<script type="module">
import {process} from '../node_modules/chrome-coverage/chrome-coverage.min.mjs';
const chromeCov = document.body.querySelector('chrome-coverage');
fetch('./jsCoverage.json')
.then(res => res.text())
.then(process)
.then(coverage => chromeCov.coverage = coverage);
</script>
</body></html>This will render a coverage report.
chromeCoverage
chrome-coverage's default export is a method that defines Mocha tests,
processes Chrome's report and renders the HTML report. The method has this
signature:
function describeCoverage({ratio = 1, uncoveredFiles = 0, getCoverage, files, filter}) { // ...You must at least pass getCoverage, everything else is optional.
getCoverage is a method that will be called without arguments and must return
Chrome's coverage report or a Promise that resolves to it.
Note that you should call chromeCoverage/describeCoverage after loading
all your tests. The coverage test it describes should be the last Mocha
test because the coverage generated by mocha-headless-chrome should only be
finalized after running all your other tests. If you generate your coverage
report manually, though, order does not matter.
files and filter are passed to process, see above.
ratio is the asserted coverage ratio. If you want at least 70% coverage, pass
{ratio: 0.7, getCoverage}.
uncoveredFiles is the number (count) of acceptable uncovered files. This only
applies if you also pass files. Otherwise chrome-coverage cannot know of
missing files and does not generate the respective test.