Re: [Chicken-users] Re: problem with callbacks

[Felix Winkelmann]

Am Thu, 22 Jan 2004 12:21:41 -0500 (EST) hat Daniel B. Faken 
<address@hidden> geschrieben:

> I observe this behaviour with Chicken 1.30 (under redhat 7.3):
> compiling with 'csc tst.scm' and running './tst':
>
> ---- (1) program as shown above
> gloriot2:~/cr/spu/viewpoint/htcam%./tst
> foo is #<hash-table>
> test_fn(50): before callback
> callback-val is 50
> Error: unprintable non-immediate object encountered
>
> tst.scm: 3    make-hash-table
> tst.scm: 5    make-hash-table
> tst.scm: 23   print
> tst.scm: 25   test_fn
> ##sys#gc
> tst.scm: 18   print
> tst.scm: 19   print
> in callback: foo is gloriot2:~/cr/spu/viewpoint/htcam%

Right, it doesn't see the lexical variable.

> --- (2) compiling program with '(scm_callback 3)' uncommented:
> gloriot2:~/cr/spu/viewpoint/htcam%csc tst.scm
> tst.c: In function `scm_callback':
> tst.c:68: `f_65' undeclared (first use in this function)
> tst.c:68: (Each undeclared identifier is reported only once
> tst.c:68: for each function it appears in.)
> *** Shell command terminated with exit status 1: gcc tst.c -o tst.o -g -c
> -DC_NO_PIC_NO_DLL -I /home/dfaken/lib/chicken/include

I can't reproduce the error here (using 1.31 and MSVC on Windows), so I'm
not really sure why the compiler complains. But, basically two things have
to be kept in mind:

1) callback do indeed not capture the lexical environment (that should be
  explicitly stated in the manual, of course)
2) #> ... <# should only appear at toplevel (I will also update the manual
  accordingly)


cheers,
felix