This allows "obj.description!" to work on iOS when 'obj' is an AnyObject,
even though 'description' is an implicit property created by the importer.
Swift SVN r18149
This is a better solution to <rdar://problem/16899681> because the
runtime magic is limited to implementing the witnesses of this
conformance.
The type checker fixes are because we can end up using unchecked
optionals in more places, via bridging, than we could before.
Swift SVN r18120
Use TypeInfo::indexArray to get the bounds of the array instead of duplicating its logic incorrectly. Fixes <rdar://problem/16916422>.
Swift SVN r18115
This allows us to, for example, convert an array of unchecked
optionals of some bridged type T to NSArray. It's a runtime failure if
the unchecked optional contains nil.
This is most of <rdar://problem/16899681>, but I want to improve the
testing and look into the various FIXMEs.
Swift SVN r18106
This fixes a case where the Swift-variadic and C-varargs versions of
various initializers were superseding each other
<rdar://problem/16801456>.
It also uncovered some more cases where we weren't getting quite the
right semantics for factory-methods-as-initializers, which are also
fixed here.
Swift SVN r18010
The old ones were:
- print/println
- printAny
- printf
- Console
The new printing story is just print/println. Every object can be printed.
You can customize the way it is printed by adopting Printable protocol. Full
details in comments inside stdlib/core/OutputStream.swift.
Printing is not completely finished yet. We still have ReplPrintable, which
should be removed, string interpolation still uses String constructors, and
printing objects that don't conform to Printable will result in printing
mangled names.
Swift SVN r18001
On the swift side, add an asmname'd "getSummary" function to make it easier for the runtime to access 'reflect(x).summary' (and to make it easy to change this when we get a real printing solution). On the C++ side, implement -description by trying to find convertStringToNSString if Foundation is loaded, and falling back to just returning @"SwiftObject" if not. It's unlikely SwiftObject's methods will be exercised if Foundation isn't loaded.
Swift SVN r17954
Make ObjC method partial applications go through a native-to-foreign thunk so that we properly handle bridging conversions in the resulting function value. Partial applications through dynamic lookup are still broken because they apparently go through a different path.
Swift SVN r17108
Before we create a new initializer, check whether it would collide
with existing initializers. If it's better than the existing
initializer, mark the existing one as unavailable; if it's not better,
don't build it.
In support of this, we tweak a few things w.r.t. unavailble
declarations:
- An unavailable declaration is shadowed by an available one,
- Don't emit SIL unavailable, imported initializers
This fixes the last problem with <rdar://problem/16509024>.
Swift SVN r16611
This makes it so that we actually register the RawOptionSet conformance of imported structs and emit it into executables, fixing <rdar://problem/16295991>.
Swift SVN r16323
The Swift variant uses a function with default arguments. In
Objective-C NSBundle.h defines a set of macros that simulate default
arguments.
Implements <rdar://problem/15787627>.
Swift SVN r16226
Make these three types conform to the BridgedToObjectiveC
protocol, which is needed for array bridging. This is one part of
<rdar://problem/16533359>.
Because this must happen in the Foundation module, hack
swift_conformsToProtocol to look in the Foundation module for
conformances when it can't find them in the module corresponding to
the type. This is an egregious hack to an egregious hack, but it gets
us closer.
Swift SVN r15997
...since all imports are private by default, and we rely on "import Cocoa"
bringing in AppKit. (If you say "import AppKit" directly, you get an
immediate call to IRGenImportedModules, so everything works.)
Swift SVN r12617
It relies on block shimming infrastructure that's in the process of being disrupted by the NewString transition and that should be superseded in the long term by proper compiler support for blocks.
Swift SVN r11829
If Foundation wasn't ever actually loaded into the ASTContext, type
lowering would fail to find the NSString-to-String conversion functions.
Fix this by only bridging NSString when a module named "Foundation" can
be loaded.
This still isn't very precise, since the bridging functions could still
be missing, but it at least avoids a user-triggerable error.
<rdar://problem/15027448>
Swift SVN r9380
The name used for name lookup of subscript operators is "__subscript",
but when looking into an Objective-C module we instead need to look
for methods with the subscripting selectors, e.g.,
objectAtIndexedSubscript:/objectForKeyedSubscript:. Do so, and make
sure that deserializing the method first still creates the subscript
declaration.
Fixes the majority of <rdar://problem/14656624>. We can subscript 'id' now.
Swift SVN r8700
Loading Cocoa (Clang) triggers the loading of several other modules,
including Foundation (Clang). We want to make sure we bring in the adapter
module Foundation (Swift), so loading Cocoa currently triggers a walk of
all imported modules to check for an adapter. The problem is that
Foundation (Swift) contains some eagerly-deserialized decls, which refer
to decls in Foundation (Clang). These decls were not importing correctly
because we hadn't finished loading all the adapters -- specifically, we
hadn't yet loaded ObjectiveC (Swift). This meant SEL was loaded as
swift.COpaquePointer instead of ObjectiveC.ObjCSel.
Now, when Foundation (Swift) asks to load Foundation (Clang), we'll walk
all imported modules under Foundation (Clang) to check for an adapter
before returning, even if Foundation (Clang) is already in our system.
This probably still isn't the best way to deal with this (particularly
since the modules in the subtree will get visited twice), but it ensures
that we've loaded all necessary adapter modules before trying to import
any decls.
TLDR: Importing Cocoa now correctly imports the Swift overlay for the
ObjectiveC module, meaning SELs are correctly imported.
<rdar://problem/14759044>
Swift SVN r7293
Test that string bridging interacts successfully both with native ObjC NSString methods and with Swift subclasses that define String methods end-to-end.
Swift SVN r5367
Preserve the AbstractCC of a function symbol when emitting a SIL ConstantRefInst by adding a new StaticFunction variant to LoweredValue. This lets us avoid a bunch of bitcasting noise for static functions that aren't used as values, and will lets us emit C-to-Swift-ABI thunks on demand when C functions are used as values.
Swift SVN r4543
A Swift method in an extension cannot override a method nor can it be
overridden by a method. Diagnose this. We may remove this limitation
at a later time.
An [objc] method in an extension can override and can be overridden, so
specifically allow this case. It's useful in our Foundation bindings.
Swift SVN r4308
In the lit.site.cfg, if a MODULES_SDK is available, add a '%sdk' substitution and enable REQUIRES: sdk tests. Add some tests under test/Interpreter/SDK to test some basic ObjC interop features.
Swift SVN r4243
