Overloaded augmented assignments

[japaric]

Implements overload-able augmented/compound assignments, like a += b via the AddAssign trait, as specified in RFC 953

r? @nikomatsakis

[bors]

☔ The latest upstream changes (presumably #28358) made this pull request unmergeable. Please resolve the merge conflicts.

[nikomatsakis]

Nit: can we newtype this bool? I think I prefer something like:

struct IsAssign(bool)

and then in the call foo(..., IsAssign(true)), though IsAssign::Yes is also ok.

[nikomatsakis]

This looks good. r+ modulo nits -- can you also add a test that you can't implement AddAssign and friends without the suitable feature gate? (Another related question -- do we really want two distinct feature gates? Seems like the traits and associated syntax live or die together?)

Also: I think it may be worth waiting to land this until the beta is released, just in case it introduces accidental breakage.

[japaric]

@nikomatsakis I've addressed all your comments (I think)

do we really want two distinct feature gates? Seems like the traits and associated syntax live or die together?

I did try using the same feature name for the language feature and for the traits, but it didn't work. What I observed is that when I added #![feature(augmented_assignment)] to some source file, the language feature was enabled, i.e. a += b didn't raised an error; but implementing the OpAssign trait still raised the "unstable library feature 'augmented_assignments'" error. Which seems to indicate there's a bug when a language feature and a library feature have the same name: #![feature(foo)] enables the language feature foo, but not the library feature foo when it should enable both.

Also: I think it may be worth waiting to land this until the beta is released, just in case it introduces accidental breakage.

Sounds good to me

[bors]

☔ The latest upstream changes (presumably #28336) made this pull request unmergeable. Please resolve the merge conflicts.

[nikomatsakis]

@japaric ok feel free to r=me once rebased

[japaric]

@bors r=nmatsakis

[bors]

📌 Commit f5569ec has been approved by nmatsakis

[bors]

⌛ Testing commit f5569ec with merge c61bf60...

[bors]

💔 Test failed - auto-linux-64-opt

[bluss]

@bors retry

[bors]

⌛ Testing commit f5569ec with merge 295d51f...

[bors]

💔 Test failed - auto-linux-64-opt

[alexcrichton]

@bors: retry

On Sat, Sep 19, 2015 at 4:08 AM, bors notifications@github.com wrote:

[image: 💔] Test failed - auto-linux-64-opt
http://buildbot.rust-lang.org/builders/auto-linux-64-opt/builds/6468

—
Reply to this email directly or view it on GitHub
#28345 (comment).

[bors]

⌛ Testing commit f5569ec with merge 783c3fc...

[bors]

☀️ Test successful - auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-x-android-t, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-win-gnu-32-nopt-t, auto-win-gnu-32-opt, auto-win-gnu-64-nopt-t, auto-win-gnu-64-opt, auto-win-msvc-32-opt, auto-win-msvc-64-opt