R devel - Jul 2004 - Wrong object type produced - LANGSXP should be LISTSXP (PR#7055)

Full_Name: David Bauer
Version: 1.9
OS: Linux
Submission from: (NULL) (160.91.245.8)


In the file gram.y, the xxsubscript function generates a LANGSXP with another
LANGSXP as its CDR.  I believe that this is a mistake and that the second
LANGSXP should be a LISTSXP.  The inputs a1, a3 are parameters to the subscript
function (a2), and as such they should be in a dotted-pair list.

David Bauer


--- gram.y.orig      2003-11-15 05:40:35.000000000 -0500
+++ gram.y        2004-07-06 13:45:10.000000000 -0400
@@ -731,11 +731,11 @@
                                                                               

 static SEXP xxsubscript(SEXP a1, SEXP a2, SEXP a3)
 {
     SEXP ans;
     if (GenerateCode)
-       PROTECT(ans = LCONS(a2, LCONS(a1, CDR(a3))));
+       PROTECT(ans = LCONS(a2, CONS(a1, CDR(a3))));
     else
        PROTECT(ans = R_NilValue);
     UNPROTECT_PTR(a3);
     UNPROTECT_PTR(a1);
     return ans;

bauerda@ieee.org writes:
> Full_Name: David Bauer
> Version: 1.9
> OS: Linux
> Submission from: (NULL) (160.91.245.8)
> 
> 
> In the file gram.y, the xxsubscript function generates a LANGSXP with
another
> LANGSXP as its CDR.  I believe that this is a mistake and that the second
> LANGSXP should be a LISTSXP.  The inputs a1, a3 are parameters to the
subscript
> function (a2), and as such they should be in a dotted-pair list.
Hmmm. Probably true in principle (as far as I can see, corresponding
logic for funcalls does make the argument list an ordinary pairlist,
not a language object). However, is it not a victimless crime? I can't
think of a way to make it matter at the R level. It's the sort of thing
that you tend not to want to fix if it isn't broken...

Notice that 

z <- quote(x[2])
mode(z[-1])

returns call, but the same is true of any function call (and that's
kind of weird, but a consequence of a general rule that [-indexing
returns an object of the same mode as the original).

[snip...] 
> -       PROTECT(ans = LCONS(a2, LCONS(a1, CDR(a3))));
> +       PROTECT(ans = LCONS(a2, CONS(a1, CDR(a3))));

-- 
   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