Wow, I didn't think we'd ever see a PR for this. Nice job, I'll review soon.

The rule is simply that a dot accessor on a new line closes all implicit calls

All of em?

What happens in cases like this?

function arg, ->
  jQuery selector
    .fadeIn 'slow'

@jashkenas I had the same question, but then I re-read the last example :

$ 'body'
.click (e) ->
$ '.box'
.fadeIn 'fast'
.addClass '.active'
.css 'marginRight', '10px'

@jashkenas Only implicit calls, not function bodies or object bodies. So in your case the "open" function body prevents closing of the first function call. I agree explaining it in words is tougher than showing by examples / the algorithm itself, but it should be easy to grasp.

+1. The current trends with promises makes using coffeescript somewhat pointless when chaining as we need to wrap with parenthesis or assign temp variables.

Chaining like this would be highly pleasing, seems like one of the few remaining JS syntax constructs CS is failing to improve on. I could imagine it would be more fitting for redux though considering backwards compat.

I could imagine it would be more fitting for redux though considering backwards compat.

If we were using semver we would be on version 7.0.1. There are loads of backwards incompatible changes (the changes to string literals I worked on for example) on master. I also hope that #2887 will get merged in before 1.7.0 (more of a big change rather than backwards incompatible). The main point is I couldn't find the code that changes behavior in production, any examples? Redux is not tangible for me right now.

I totally agree about promises.

@xixixao I simply imagine backwards compatibility would be the only practical obstacle for this feature. I've followed this chaining syntax debate for a year or so and the only other reason that comes to mind is that the big CS contributors just don't like it, or care. I'd imagine their opinions may have changed, though. Heres to hoping :D

Here we go...

👍

👍 👍

This looks great! When can we expect to see a release containing this change? It'll definitely work wonderfully for promises, as well as a bunch of other chainy interfaces (jQuery's for example).

I think I have some cases in my code-base that relied on the old behavior. Can someone perhaps make some tool to find those cases, or possibly even issue a warning with the file/line and a notice about what changed when CoffeeScript encounter those cases in the next 2-3 releases, until everyone are aware of it?

I can see this causing some havoc... we at least need an easy way to let people find those cases before they upgrade.

@shesek I'd simply compile your codebase with both versions (you can do this now with last release and master) and diff the outputted JS. I'd be interested to see the use cases.

glad to see method chaining finally happening. Thanks!