[SR-1009] Add methods to an Array containing a specific type

[swift-ci]

Previous ID	SR-1009
Radar	rdar://problem/17821143
Original Reporter	mattdipasquale (JIRA User)
Type	New Feature
Status	Resolved
Resolution	Done

Additional Detail from JIRA

Votes	0
Component/s	Compiler
Labels	New Feature
Assignee	@slavapestov
Priority	Medium

md5: b6b1e530310ef2059d8fb2110e717e5f

is duplicated by:

• SR-1447 Concrete same-type requirements

Issue Description:

For example, if I'd like to activate the constraints inside of constraints: [NSLayoutConstraint], instead of writing NSLayoutConstraint.activateConstraints(constraints), I'd like to write constraints.activate(). However, the compiler should only allow calling activate() on an array if the contained type implements activate(). Note: Apple's documentation says: "Typically, using [the activateConstraints()] method is more efficient than activating each constraint individually." So, constraints.activate() would be faster than constraints.map { $0.active = true }.

[belkadan]

The syntax for this would be

extension Array where Element: NSLayoutConstraint {
  func activate() { … }
}

except we haven't implemented that yet. @DougGregor, do you remember where we are with that?

[belkadan]

Actually, the version with : (as opposed to ==) might already work.

[DougGregor]

We've been doing some of the infrastructural work that would enable this feature, but nobody is actively working on the implementation of this feature (which we refer to as "concrete same-type requirements" on https://lists.swift.org/pipermail/swift-evolution/Week-of-Mon-20160229/011666.html)

[swift-ci]

Comment by Matthew Mario Di Pasquale (JIRA)

Cool! :) Glad this is already on the radar. Thank you for the updates, @belkadan & @DougGregor, and thank you for all your legendary contributions to Swift. :) I'm honored to be of assistance.

[karwa]

@belkadan@DougGregor Does this look like a reasonable fit for Swift 3? Has that infrastructure worked progressed to a point where somebody could pick up implementing this feature?

[slavapestov]

Unfortunately there's still a fair bit of work remaining before we can tackle this, I'm afraid. SILFunction::ContextGenericParams probably has to go away and be replaced with an ArrayRef<Substitution>.

[slavapestov]

An update: we're getting closer. SILFunction::ContextGenericParams is gone, and so is the canonical mangling signature stuff. Next step is to remove the archetype from AbstractTypeParamDecl; I'm working on that now.

[slavapestov]

Here it is: #5126

[slavapestov]

Merged.