Allow is, includes and get to work on relative stateOrName

[pdanpdan]

No description provided.

[nateabele]

Looks good. I'll merge this as soon as I have a chance to write some tests for it.

[timkindberg]

I could see this one being a little tricky like sref was at first, where if you use it in ng-class for example the dev will have to know its based off the state that loaded the template containing the element with the ng-class. Does this complicate things?

[nateabele]

@timkindberg This is for the service, so it doesn't directly affect anything that happens in a template. That said, developers still have to know what they're doing if they're gonna use it in a template with relative state references.

[timkindberg]

Ah yes, seeing as people use these services directly in templates quite often, I'll wager that we'll see this problem crop up eventually... I guess I'll just add a blurb in the docs...

[timkindberg]

Is this going in? I was just about to write the docs for it... but realized its not merged.

[nateabele]

Not until someone writes tests for it, which is a pretty low priority for me right now.

[dlukez]

I can write some tests for this. How is it meant to be used? Like so?

// state is 'about.person'
var about = $state.get('about'); // could keep state object from declaration?
$state.is('.person', undefined, { relative: about }); // true

Although I believe the above approach would fail in findState() if I was searching a parent using '^', because as I understand it there won't be a 'parent' property on the state config, which would lead to an error thrown at https://github.com/angular-ui/ui-router/blob/master/src/state.js#L151. E.g.

// state is 'about'
var person = $state.get('about.person');
$state.is('^', undefined, { relative: person }); // Path '^' not valid...

Or perhaps would it be stored from a previous $stateChangeSuccess event? This should work with the case above.

[nateabele]

@dlukez Your usage is correct, but I think we should also allow something like:

$state.is('.person', undefined, { relative: "about" });

...to be equivalent.

This should just be a matter of calling base = findState(base) inside of findState().

Regarding the Path '^' not valid example, I'd double-check that. I think it works correctly, but if not, I wouldn't worry about it for now, since we still need to handle resolving state parents in a consistent way, which is still a big outstanding item on the to-do list before ui.router's API is totally stable.

[dlukez]

I created a PR with the tests against pdanpdan's fork. Let me know if I should PR them against this repo instead.

[timkindberg]

Can we squash all those commits?

[christopherthielen]

I rebased this and merged it into https://github.com/christopherthielen/ui-router/commit/693ba51b66b4db1057ded3921db26a0f73395c8c

However, I think there may be a breaking change here. @nateabele please have a look at https://github.com/christopherthielen/ui-router/commit/693ba51b66b4db1057ded3921db26a0f73395c8c#diff-230fff7c0b3ed8f0d4f2443a246fc76aR1146 and let me know before I move further on this PR.

[christopherthielen]

Also, before we merge anything, the docs will need updating.

[nateabele]

@christopherthielen Made one comment with an alternate implementation that shouldn't break BC.