Fix single-column-line? behavior for last elements

[beatriz]

I noticed that the :single-column-line? rule is not currently respected when the single-column element is the last one in a map. This PR fixes that.

[weavejester]

Please correct me if I'm wrong, but wouldn't we want to check for nil after skipping the whitespace and commas? i.e.

  (and (let [zloc (skip-whitespace-and-commas (z/right* zloc) z/right*)]
         (or (nil? zloc) (line-break? zloc)))
       (line-break? (skip-whitespace-and-commas (z/left* zloc) z/left*)))

Otherwise a single column line ending in a comma or whitespace wouldn't trigger this rule.

[beatriz]

Yes, you are correct! Fixed it.

[weavejester]

Thanks! Could you use the code I suggested in my previous comment? It makes it a little more concise and constrains the scope of the let clause.

[beatriz]

Of course!

[beatriz]

This is ready to review, when you have the chance 🙂

[weavejester]

Ideally we should have some invalid indentation so that there's a change in the two expression strings, otherwise we can't distinguish between it working and it doing nothing. So perhaps:

    (is (reformats-to?
         ["{:key-first 1"
          " :key2 3"
          " :key3"
          " (fn [a b c d e]"
          "   (+ a b c d e))}"]
         ["{:key-first 1"
          " :key2      3"
          " :key3"
          " (fn [a b c d e]"
          "   (+ a b c d e))}"]
         {:align-map-columns? true})))

[beatriz]

Makes total sense, I've adjusted the test.

[weavejester]

This all looks good! Can you squash down your commits and give them an appropriate commit message like:

Fix single-column-line? behavior for last elements