Is it possible for source code generated by a Run Script phase to be included automatically in the normal compile/link process?
I have a project with some .foo files, some regular .cpp/.h files, and some .h files that contain special symbols. The .foo files and the special .h files are run through preprocessors that generate additional .cpp files, which I want to feed into the build.
I don't want to use Build Rules to manage these, since those have to be defined on a per-project basis (and they can't catch the special .h files).
Instead I am using an external script which scans the source tree, identifying any files that need special treatment. That script then generates and executes a makefile, which (re)generates whatever additional .cpp files are necessary.
My problem is how to get these .cpp files into the subsequent build.
I've seen references to DERIVED_FILES_DIR, with the implication that source placed in there will somehow be sucked up into the build; but so far the only approach I've found is to explicitly add the intermediate files to the project.
I would prefer not to do that, as it means I have an extra "build then add newly-created .cpp to project" step whenever I add a .foo/special-.h to the project.
Is there any other way to get Xcode to incorporate this kind of dynamically generated source into the build, short of having my makefile doing the compiling as well then injecting the .o files with -filelist?
I'd rather not take that approach, since ideally this source would be built under the settings of the current build configuration rather than hard-coding some build flags into the makefile.