browser: cache setTimeout and clearTimeout to isolate from global

[ilgooz]

tl;dr:
sinon stubs time related global objects for unit testing and this breaks all packages depended on process.nextTick. caching setTimeout and clearTimeout at bootstrap protects against mutations.

related issue:
sinonjs/fake-timers#66

[calvinmetcalf]

var setTimeout = global.setTimeout; would probably do the trick

[calvinmetcalf]

I think the update can be simplified to just 2 lines that cache them with the same name, a test would be very helpful as well

[ilgooz]

@calvinmetcalf of course. you mean window as global with var setTimeout = window.setTimeout syntax? otherwise global would only work with nodejs.

[calvinmetcalf]

browserify/webpack sets global to be window in the browser or self in a web worker so it's a shorthand to the global object in any of those envs

[calvinmetcalf]

actually though it wouldn't shock me if this was used in some env that didn't give global, so your way is probably good.

a comment explaining this is probably a good idea.

[ilgooz]

@calvinmetcalf js world does not shock me anymore :). is it good this way? I couldn't come up with a test case though any suggestions welcome

[calvinmetcalf]

maybe something like cached from whatever global is present so that test runners that stub it don't break things

[ilgooz]

@calvinmetcalf 👍

[calvinmetcalf]

test for you

describe('rename globals', function (t) {
        it('throws an eroror', function (done){
          var oldTimeout = setTimeout;
          setTimeout = function () {
            setTimeout = oldTimeout;
            assert.ok(false);
            done();
          }
          ourProcess.nextTick(function () {
            assert.ok(true);
            setTimeout = oldTimeout;
            done();
          });
        });
      });

[ilgooz]

right! I've added some tests also for clearTimeout

[calvinmetcalf]

0.11.4

[ilgooz]

thanks! @calvinmetcalf fyi, setting clearTimeout to original one inside nextTick for cleanup has a side effect that makes tests to do not complain when clearTimeout replaced with original one because of execution order inside drainQueue()

[calvinmetcalf]

oh i see fixed 53fdf64