Any flaw to never create ObjC files, but always create ObjC ++ files?

Question

By default, Xcode creates both an .h and .m file when requesting a new ObjC class. Everything works fine until you need to reference any C ++ file elsewhere in your project and do not run it in a .h or .m file.

At that moment, when the ObjC compiler is completely confused and throwing mountains of parsing errors, and you, the user (that is: me), are even more confused until it hits me: of course, I have to make this file instead this objc ++ file.

Possible options:

• tell Xcode that this particular file, even if it is a .m file, is really an ObjC ++ file or
• rename this file to .mm.

The first option is not very pleasant for me, because this file is really ObjC ++, regardless of what, according to the project, is.

The second option doesn’t work, because it twists the Git repository, which then forgets that another .m file was used earlier, which is really the history of this “new” .mm file.

So, from now on, I decided to always rename any .m file that Xcode creates for me .mm first, after creating, so that I don't lose the story.

So far, it worked well for me, but I have this slight concern in my head that there might be some kind of angular case when I really want to have an ObjC file, not an ObjC ++ file.

What would these corner cases be? Does anyone know of any ObjC ++ file that happens to NOT contain a C ++ link, but somehow strangled the ObjC compiler, simply thanks to the .mm file?

And if there are no flaws, why not just give up .m forever and stick with .mm instead?

Answer 1

There are no shortcomings, although two methods ( .cxx_construct and .cxx_destruct ) were created because of you, but they are used only to create and destroy C ++ objects when creating / deallocating an instance. If your class does not have C ++ members, these functions do nothing and add only very low overhead. Otherwise, you still have C functions created for your Objective-C methods, not C ++.

Answer 2

C ++ analysis is much slower than ObjC parsing, so ObjC ++ files have significantly longer compilation times. I'm not sure that this overhead will apply to ObjC ++ files that do not contain C ++, but it will make some sense, which is harder to parse because the compiler needs to look for C ++ constructs.

In addition, the C ++ type system has several slightly different rules from C, which also apply to the ObjC / C code in the C ++ file. I do not remember the details, but it will not be harmful; it may just take a few extra throws.

Answer 3

Create an Objective-C ++ file template so that you get a .mm file instead of a .m file when you create a new file. Create a copy of the Apple Objective-C class templates and rename the .m files to .mm. For more information on creating Xcode 4 file templates, see the following article:

Creating Custom Xcode 4 File Templates

Answer 4

Incompatibility of languages ​​aside, one of the reasons to avoid the project completely .mm is that you may be tempted to start sections of your methods in C ++, which will lead to a project written in a (relatively obscure) hybrid of two languages, and there will be understand only those who know both. (I did it before)

A good way to avoid cluttering your obj-c headers with C ++ is to declare instance variables in your implementation file (which is allowed with xcode 4.2 / clang 3.0, possibly earlier). For instance:

 @implementation MyClass { std::vector<int> myVector; } 

This helps to maintain common ground between objective-c and C ++ with minimal values.

Answer 5

I use the .mm file exclusively for my iOS code and have never had any problems. Yes, my compilations are a bit slower since a clean compiler takes 15 seconds versus 10 seconds. At least on iMac this is not essential.