[R] R_problem_buf in the S.h files & lists in .C (PR#185)

[p.dalgaard@biostat.ku.dk]

Duncan Temple Lang <duncan@rice.research.bell-labs.com> writes:

> I didn't see any mention of this in the material on CRAN.  The two S.h
> files define R_problem_buf as a real symbol rather than simply
> declaring its existence. As a result, a shared library that links two
> or more files that include this will get duplicate symbol complaints.
> Since it is not defined elsewhere in the code, someone has to define it.

[I'll file this as a bug report and take it off of R-help]

Yes. This game is commonly played by the 

#ifdef __MAIN__
#define extern
#endif

and including the .h files in main.c. However, R doesn use that trick
(not even for the declarations in Defn.h) for some reason that I don't
quite see. A quick workaround could be to define the buffer in
S_compat.c. 

> A secondly thing is the documentation for .C() and company. It omits
> what type list() objects have when passed to a C routine. It should
> read void *, which is actually a SEXP * in this case. (I know, I
> should be using .External but I am trying to keep the code for the
> different S's and R's together until I have to abandon things.)

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._