[swift-evolution] [swift-build-dev] [swiftpm] Add proposal for C language support

[Daniel Dunbar]

> On Jan 3, 2016, at 6:34 AM, Maury Markowitz via swift-evolution <swift-evolution at swift.org> wrote:
> 
>> On Jan 2, 2016, at 9:00 AM, Daniel Dunbar via swift-build-dev <swift-build-dev at swift.org> wrote:
>> 
>> Happy 2016!
>> 
>> I am working on an initial proposal for adding support for C language targets to the Swift package manager, and am interested in feedback:
> 
> The idea of including C libraries/code in Swift projects *without* having to use Xcode strikes me as *very* valuable. Two questions:
> 
> a) for item (2) in "solution" I see the advantages of placing the bridging in a separate directory, but I want to point out that this may make the interactions between various devenvs and things like GitHub more annoying. Xcode's Bridging-Header solution has it's own problems, but may be easier to work with in complex projects. Generally I think more flexibility here might be valuable, even going so far as a ".bh" for the bridging files, allowing them to be placed anywhere.

We anticipate directly adding support for targets with both C and Swift code, at some point (which is where the bridging header comes in), but that is more involved and something I wanted to explicitly tackle separately. This proposal is very focused at adding support for building C source code that is a part of the project, and exposing the headers that declare the interfaces to that code. It is not really intended to tackle the problem of using C headers to expose *other* code.

> b) for (3), I'm facing this problem right now porting some very old C that has a main.c that includes main() as well as other more general code that the rest of the system uses. I solved this by changing main(), but that's not ideal, I would greatly prefer to use the original code verbatim. So for the 10% of cases where this is a problem, perhaps a compiler directive would be useful?

If you are integrating the code directly into your project I feel like it makes more sense to require some adaption in order to keep the conventions simple. There are likely a large number of adaptions necessary to integrate existing C projects into these conventions, over time I expect we will allow customization for common problems (and ones that are easy to define a syntax for customizing), but I'm not sure if this one would bubble up to that level of importance.

 - Daniel

> 
> 
> _______________________________________________
> swift-evolution mailing list
> swift-evolution at swift.org
> https://lists.swift.org/mailman/listinfo/swift-evolution