container used in unit tests doesn't normalize model names

[kbullaughey]

I ran into a problem writing unit tests that I detailed in a StackOverflow question:

http://stackoverflow.com/questions/28130262/how-do-i-use-a-multi-word-model-name-when-using-ember-cli

I found a workaround that might suggest at the root cause. In my moduleForModel callbacks, I override the container's normalizeFullName method to dasherize strings:

setup: function(container) {
  container.normalizeFullName = function(fullName) { return fullName.dasherize(); }
}

The function I overrode is the default Ember implementation, which just returns fullName. This isn't necessary in in my actual app because the implementation looks like this:

function (fullName) {
  if (resolver.normalize) {
    return resolver.normalize(fullName);
  } else {
    Ember.deprecate('The Resolver should now provide a \'normalize\' function', false);
    return fullName;
  }
}

In my case I'm using the ActiveModelAdapter, which expects incoming json roots to be underscored. This is then converted to a camelCase name that is passed around in Ember Data, and then re-normalized to a dasherized name during resolution on the container. But in unit tests, it's never converted to the dasherized name and a lookup using the camelCase name fails.

[rwjblue]

This should be resolved in 0.2.0.

[kbullaughey]

Great. Thanks. I upgraded and I can confirm that I no longer have the problem.

[rwjblue]

Awesome, glad to hear it!