Running tests directly and as part of a suite

[josh]

I want to run my test file alone - node test/unit/test_404.js
but also with the test suite - node test/unit/runner.js

to achieve that I added an if/else statement in each test file to check if the code is running directly with the node executable or if it's required by other file.
if it was run directly, I run my server before running the test.
if the file was required by the test suite runner, I just export the test so the runner will call it (module.exports = test).

Here are 2 version of one of the test files - http://hastebin.com/fovofufome.js 
They are doing the same thing, except that in version 2 I moved the runServerBeforeTest function to it's own helper file.
I found the first version more readable but it's more verbose and I need to repeate that in each test file.
Any opinions or even different approaches for testing?

btw, I deliberately uses node's built-in asserts and not using any test framework. I would like to feel pain before adding libraries and so far I am happy with this 'leas' approach.

Pasting my code here as well:

// test_404.js version 1

// core modules
var assert = require('assert');

// npm package
var request = require('request');


// GET /not-exist should return 404 - not found
function test(config) {
  console.log('GET /not-exist');

  request({
      uri: "http://localhost:" + config.webSitePort + "/not-exist",
      followRedirect: false 
    }, 
    function (err, res, body) {
      if (err) {
        console.log('Error in ' + __filename + '. ' + err);
      } else {
        assert.equal(res.statusCode, 404)
        // process.exit();
      };
    });
};

// The following code let us run this test file by itself or as part of the runner.js
// if this file is by itself with 'node test_foo.js' we need to run the server before calling the test
// else - expose the test function since the runner.js already run the server
if (module === require.main) {
  runServerBeforeTest();
} else {
  module.exports = test;
};

function runServerBeforeTest() {
  // my modules
  var config = require('../../config/test.js');
  var app = require('../../app.js');

  app.init(config, function (err, msg) {
    if (err) {
      console.log('Error during application init: ' + err);
    } else {
      console.log(msg);
      app.start(test(config));
    }
  });
};


// test_404.js version 2

// core modules
var assert = require('assert');

// npm package
var request = require('request');


// GET /not-exist should return 404 - not found
function test(config) {
  console.log('GET /not-exist');

  request({
      uri: "http://localhost:" + config.webSitePort + "/not-exist",
      followRedirect: false 
    }, 
    function (err, res, body) {
      if (err) {
        console.log('Error in ' + __filename + '. ' + err);
      } else {
        assert.equal(res.statusCode, 404)
        // process.exit();
      };
    });
};

// The following code let us run this test file by itself or as part of the runner.js
// if this file is by itself with 'node test_foo.js' we need to run the server before calling the test
// else - expose the test function since the runner.js already run the server
if (module === require.main) {
  runServerBeforeTest();
  require(__dirname + '/helper.js').runServerBeforeTest(test);
} else {
  module.exports = test;
};

// helper.js - after my server is listening, run a given test

module.exports.runServerBeforeTest = runServerBeforeTest;

function runServerBeforeTest(test) {
  // my modules
  var config = require('../../config/test.js');
  var app = require('../../app.js');

  app.init(config, function (err, msg) {
    if (err) {
      console.log('Error during application init: ' + err);
    } else {
      console.log(msg);
      app.start(test(config));
    }
  });
};


// runner.js - after my server is listening - run all test*.js files 

// my modules
var config = require('../../config/test.js');
var app = require('../../app.js');
var helper = require(__dirname + '/helper.js');

function runTests() {
  console.log('running tests:');

  testFiles = [];
  var test = null;
  testFiles = require("fs").readdirSync(__dirname).filter(function(file) {
    return (/^test/.test(file)) 
  });
  
  testFiles.forEach(function(file) {
    test = require("./" + file);
    if (typeof(test) === 'function') {
      test(config);
    }
  });
};

app.init(config, function (err, msg) {
  if (err) {
    console.log('Error during application init: ' + err);
  } else {
    console.log(msg);
    app.start(runTests(config));
  }
});

[Daniel Rinehart]

I would suggest modifying your runner.js so that it can take optional command line options to just run a single test or directory. That way you avoid any duplication of setup code or logic in individual tests. You just execute "node test/unit/runner.js test/unit/test_404.js" to run an individual test or "node test/unit/runner.js" to run them all.

-- Daniel R. <dan...@neophi.com> [http://danielr.neophi.com/]

--
--
Job Board: http://jobs.nodejs.org/
Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nod...@googlegroups.com
To unsubscribe from this group, send email to
nodejs+un...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en
 
 
 

[josh]

Daniel, your suggestion will make my test files shorter and DRY, but on the other hand running my tests might be less intuitive than before -
node unit/test/test_404.js vs node unit/test/runner.js unit/test/test_404.js.
I want to make it easy as possible to run the tests, so at the moment I prefer to have some verbosity in the test file itelf.

Benjamin, Joe looks interesting but I am not sure if I want to introduce another dependency to my project, unless it really add a lot of value.
in addition it's written in coffeescript, which will introduce some difficulties for me and future developers on this code base.
I guess to solve this issue I can create a copy of joe in javascript, but not sure if it adds a ton of value that justify it. we'll see about that.
at this point I prefer to stick to one language. of course this is very subjective topic and I might change my mind in the future.

Thanks for those suggestions guys. I really appreciate it!

[Jamison Dance]

Just put the call to runner in the npm test command, so people have an example to look at how to run the tests.