[Pythonmac-SIG] Python-Cocoa with XCode 3 - external modules

Daniel Lord daniellord at mac.com
Wed May 28 04:58:42 CEST 2008


On May 27, 2008, at 14:01 PM, Orestis Markou wrote:

> XCode by default will use the system supplied python. I think if you  
> want to use external modules you have to include them directly in  
> your project. Pyobjc-devel users will know more.


I sent Steve to the Xcode group for a reason:

There are two camps IMHO for PyObjC development: 1) Xcode-Interface  
Builder and 2) Interface Builder - text editor - py2app
He asked the question in a way that placed him in the first camp in my  
mind.

Using Xcode and using Xcode is quite different from just using  
Interface Builder and py2app alone. I have abandoned Xcode itself for  
PyObjC because most of its great features are useless with PyObjC  
(like debugging) but to each his own.

To the point: Xcode makes it pretty simple to add build phases to a  
target and one of the standard default phases is a file-copy phase  
which you could add to copy each resource you need (like an egg or  
other file). You can also add shell (or perl, python ruby) scripts as  
well to make pretty sophisticated builds and even save those as  
templates. The Xcode group would be best for advice for complex  
builds.  If Orestis is right that adding the resources to the project  
auto-copies them, then that's great--haven't tried it myself--I  
thought that only happened for standard resources Xcode knew about.  
Apple has an example of using shell scripts to build a universal build  
of OpenSSL mostly using configure, make, and lipo via shell scripting.  
There are quite a few environment variables you can use which are  
documented in the Xcode 3.0 User Guide.


More information about the Pythonmac-SIG mailing list