33457c48de
In the typical sequence of parse-transform-render the most efficient place to transform coordinate values is deep within the parser immediately after values have been read (this would avoid a second pass over whatever structure is used to back geometries). To accomplish this transform during parsing, we could add back parser read options to pass the transform function around. Until then, a transform method on geometries is straightforward to implement. This means we do a second pass through coordinate structures to transform, but this is typically done only once immediately after parsing.
Included in this directory
-
ol.html - the web page used to run the test suite.
-
spec - includes the OpenLayers test/spec files.
-
expect-0.2.0 - Minimalistic BDD-style assertion framework. https://github.com/LearnBoost/expect.js/
-
jquery-1.9.1 - The Write Less, Do More, JavaScript Library. http://jquery.com/
-
mocha-1.8.1 - the fun, simple, flexible JavaScript test framework. http://visionmedia.github.com/mocha/
-
sinon-1.6.0 - Standalone test spies, stubs and mocks for JavaScript. http://sinonjs.org/
-
test-extensions.js - includes OpenLayers-specific extensions to the testing frameworks.
-
mocha-phantom - a PhantomJS script for headless testing with mocha. http://metaskills.net/mocha-phantomjs/
Run the test suite with PhantomJS
With PhantomJS installed, and assuming phantomjs is in the PATH:
$ phantomjs mocha-phantom.coffee ol.html
(Works with PhantomJS 1.6.1, untested with other versions.)
This command can also be run by doing ./build.py test at the root of ol3.
Make sure that the polvr web server is running (./build.py serve), otherwise
you will most likely see something like 0 specs, 0 failures in 0.001s..
Tip for TDD'ers: to make PhantomJS run the test suite continuously each time
a spec file is changed you can use nosier (http://pypi.python.org/pypi/nosier)
and do nosier -p test -p src "./build.py test".