[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: WO4: WOApplication.h/interpreter.h (re)definition of 'trace'



Right. Right. I've hit this myself when building my Java Wrappers. Lame.
Ok.  If you modify WOApplication.h by placing this line

#undef trace

near the top, say, below these imports...
#import <Foundation/NSDate.h>
#import <Foundation/NSString.h>
#import <Foundation/NSLock.h>


That problem should safely go away.

dave



Michael G Xenakis wrote:
> 
> Sorry I wasn't more informative in the subject... here's the story.
> 
> Trying to build Java Wrapper, which wraps a WebObjectsFramework project.
>  Now, somewhere w/in the whole build process - and I'm hoping someone in
> WobJ-land can give me an idea where - /Library/JDK/include/interpreter.h is
> being parsed including the following lines:
> #ifdef TRACING
>   extern int trace;
>   extern int tracem;
>   extern void trace_method(struct execenv*, struct methodblock*, int, int);
>   enum { TRACE_METHOD_ENTER, TRACE_METHOD_RETURN,
> TRACE_METHOD_NATIVE_RETURN };
> # define TRACE_METHOD(ee, mb, args_size, type) \
>       if (tracem) trace_method(ee, mb, args_size, type); else
> 
> #else
> # define trace  0
> # define tracem 0
> # define TRACE_METHOD(ee, mb, args_size, type)
> #endif
> 
> ...the result being that 'trace' becomes defined as '0'.  Now, later on in
> the build when the compiler attempts to parse WOApplication.h, it comes
> across the following line(s):
> - (void)trace:(BOOL)aFlag;
>     /* Trace all (stderr) */
> ...the compiler substitues the definition for trace and the build fails w/
> the following error:
> parse error before '0'
> 
> So, I'm sure there's some quirky lil' situation here, which - I'm assuming
> - has been discovered previously, of which the solution will require some
> funky lil' tweaks, technically speaking.  But what?
> 
> Also, before anyone asks, I have the symcjit compiler set aside and have
> yet to obtain a 'non-buggy' one as discussed previously on this list under
> "JIT happens?"  If this is the issue, is there a solution which allows
> continued usage of the interpreter?
> 
> All thanks in advance, at present, and following.
> mx.
> 
> Michael G Xenakis                       PLATINUM technology, inc
> michael.xenakis@platinum.com            800.365.7528 x43007
> 
> "Anyone who isn't confused here doesn't really understand what's going on."
> -?.?