bug#13835: [PATCH] Let reverse! accept arbitrary types as second argument (new_tail)

bug#13835: [PATCH] Let reverse! accept arbitrary types as second argument (new_tail)

[David Kastrup]

There is no reason to restrict the type of the second argument to proper
lists as it is added as last CDR to the list without interpretation.

This allows for stack-depth friendly usage (simplified from an actual
use case that blew up around my ears) like

SCM
my_tree_copy (SCM src)
{
  if (scm_is_pair (src))
    {
      SCM res = SCM_EOL;
      do
        {
          res = scm_cons (my_tree_copy (scm_car (src)), res);
          src = scm_cdr (src);
        }
      while (scm_is_pair (src));
      return scm_reverse_x (res, my_tree_copy (src));
    }
  return src;
}

* libguile/list.c (scm_reverse_x): remove typecheck for 2nd arg
---
 libguile/list.c |    2 --
 1 file changed, 2 deletions(-)

diff --git a/libguile/list.c b/libguile/list.c
index 6c8f8be..d30f9e8 100644
--- a/libguile/list.c
+++ b/libguile/list.c
@@ -377,8 +377,6 @@ SCM_DEFINE (scm_reverse_x, "reverse!", 1, 1, 0,
   SCM_VALIDATE_LIST (1, lst);
   if (SCM_UNBNDP (new_tail))
     new_tail = SCM_EOL;
-  else
-    SCM_VALIDATE_LIST (2, new_tail);
 
   while (!SCM_NULL_OR_NIL_P (lst))
     {
-- 
1.7.10.4

bug#13835: [PATCH] Let reverse! accept arbitrary types as second argument (new_tail)

[Mark H Weaver]

David Kastrup <dak@gnu.org> writes:
> There is no reason to restrict the type of the second argument to proper
> lists as it is added as last CDR to the list without interpretation.
>
> This allows for stack-depth friendly usage [...]

Looks good to me.  Applied, thanks!

    Mark