R devel - Mar 2011 - Standalone C++ application for processing R parser output(SEXP)

Hi All,

I am trying to write a source-to-source compiler for R. I am trying to
leverage the R parser code for the purpose. I am trying to transform the
SEXP returned from the parser into an AST for our own Ruby embedded Domain
specific language.

I tried using R CMD SHBIN to compile a C function that parses arbitrary R
expressions. But I think, the generated .so file can be used from within R
and not be called from other C or Ruby programs(I get linker errors).

My Idea  is to use the SEXP processing functions/MACROS (CAR, CDR, CADR,
etc..) from within C code and transform it to our AST format. I tried
linking to libR.a and other lib*.so's when I compile the C code using gcc
but, it doesn't work.

I read that R exposes only small number of functions for library/package
writers and the compiled *.so can only from within R.

Any ideas on what is wrong, or how I can go about it?

Appreciate any help.

Thanks
RJ

	[[alternative HTML version deleted]]

On 11-03-23 7:35 PM, Rob Anderson wrote:
> Hi All,
>
> I am trying to write a source-to-source compiler for R. I am trying to
> leverage the R parser code for the purpose. I am trying to transform the
> SEXP returned from the parser into an AST for our own Ruby embedded Domain
> specific language.
>
> I tried using R CMD SHBIN to compile a C function that parses arbitrary R
> expressions. But I think, the generated .so file can be used from within R
> and not be called from other C or Ruby programs(I get linker errors).
>
> My Idea  is to use the SEXP processing functions/MACROS (CAR, CDR, CADR,
> etc..) from within C code and transform it to our AST format. I tried
> linking to libR.a and other lib*.so's when I compile the C code using
gcc
> but, it doesn't work.
>
> I read that R exposes only small number of functions for library/package
> writers and the compiled *.so can only from within R.
>
> Any ideas on what is wrong, or how I can go about it?
I think you need to think of your program as a new front end for R, even 
if you're only using a few R functions.  See Chapter 8 in the Writing R 
Extensions manual.

Duncan Murdoch

On Wed, Mar 23, 2011 at 6:35 PM, Rob Anderson <rkjandra at gmail.com>
wrote:
> Hi All,
>
> I am trying to write a source-to-source compiler for R. I am trying to
> leverage the R parser code for the purpose. I am trying to transform the
> SEXP returned from the parser into an AST for our own Ruby embedded Domain
> specific language.
>
> I tried using R CMD SHBIN to compile a C function that parses arbitrary R
> expressions. But I think, the generated .so file can be used from within R
> and not be called from other C or Ruby programs(I get linker errors).
>
I hope I am not telling you what you already know.   There are working
examples of "C Standalone" programs that link with R mathlib. I
learned from the examples here

http://www.math.ncu.edu.tw/~chenwc/R_note/index.php?item=standalone

and

http://www.stat.berkeley.edu/classes/s243/rmath.html

I can't say for sure if these give you access to all of the R stuff
you want, but you do get access to quite a bit.

PJ



> My Idea ?is to use the SEXP processing functions/MACROS (CAR, CDR, CADR,
> etc..) from within C code and transform it to our AST format. I tried
> linking to libR.a and other lib*.so's when I compile the C code using
gcc
> but, it doesn't work.
>
> I read that R exposes only small number of functions for library/package
> writers and the compiled *.so can only from within R.
>
> Any ideas on what is wrong, or how I can go about it?
>
> Appreciate any help.
>
> Thanks
> RJ
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>


-- 
Paul E. Johnson
Professor, Political Science
1541 Lilac Lane, Room 504
University of Kansas