nth-of-type support

[zgreen]

It would be great if Lost supported nth-of-type() grid spans in addition to the current nth-child() implementation.

Use case

article {
  lost-column: 1/3;
}

<section>
  <h2>Section heading</h2>
  <article></article>
  <article></article> <!-- This article gets no gutter, but should have one. -->
  <article></article> <!-- This article gets cleared, but should instead be the last in the row. -->
  <!-- etc. -->
</section>

Obviously, one can work around this by simply adding a wrapping element, but that solution may neither be semantic nor desired.

I'm happy to submit a PR, if you're interested. Thanks for all your work—this project is terrific.

[peterramsing]

👍🏻

[corysimmons]

Yeah I'd be interested in a PR for this. It'd need to be a global setting as well as on a per-element basis.

[zgreen]

Terrific, I'll put something together.

[Dylan-Chapman]

+1 Just ran into this

[peterramsing]

@zgreen Are you still up for working on a PR? (or maybe you already are) 😄 Any help would be epic!

[peterramsing]

@zgreen I just noticed that you forked and had some work going already on this. Let me know if you need anything.

[zgreen]

Yeah, it's just about there—work got in the way. Got it working w/ lost-column, tests passing. I made some changes so that you can pass the parameters optionally/in any order—have a look when I submit the PR. I need to implement it for lost-row, add a few more tests, and then it should be good to go. Hopefully can finish this off today.

[peterramsing]

Thanks! Don't worry about any sort of rush. Enjoy this last week before 2016.

[zgreen]

👍

[corysimmons]

It'd be really cool if all Lost's rules could accept values in any order. :^\

[peterramsing]

@corysimmons I'm looking into this, actually. I'm curious what it will take. Still really diving into the code right now to better understand it all as a whole. 😄

[peterramsing]

@corysimmons see #213 for follow up to your last comment.