scm_round_number using scm_round

scm_round_number using scm_round

[Kevin Ryde]

[-- Attachment #1: Type: text/plain, Size: 300 bytes --]

        * numbers.c (scm_round_number): For inum and big, just return x.  For
        real, use scm_round for problems with 2^54-1 etc covered there.

The problem that afflicted scm_round arises in the same algorithm done
with SCMs in scm_round_number.  I think it can call scm_round to share
code.


[-- Attachment #2: numbers.c.round-number.diff --]
[-- Type: text/plain, Size: 832 bytes --]

--- numbers.c.~1.243.~	2004-05-12 07:19:25.000000000 +1000
+++ numbers.c	2004-05-17 07:36:30.000000000 +1000
@@ -5007,6 +5007,15 @@
 	    "round towards the even one.")
 #define FUNC_NAME s_scm_round_number
 {
+  if (SCM_INUMP (x) || SCM_BIGP (x))
+    return x;
+  else if (SCM_REALP (x))
+    return scm_make_real (scm_round (SCM_REAL_VALUE (x)));
+  else
+    {
+      /* OPTIMIZE-ME: Fraction case could be done more efficiently by a
+         single quotient+remainder division then examining to see which way
+         the rounding should go.  */
   SCM plus_half = scm_sum (x, exactly_one_half);
   SCM result = scm_floor (plus_half);
   /* Adjust so that the scm_round is towards even.  */
@@ -5015,6 +5024,7 @@
     return scm_difference (result, SCM_MAKINUM (1));
   else
     return result;
+    }
 }
 #undef FUNC_NAME
 

[-- Attachment #3: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://mail.gnu.org/mailman/listinfo/guile-devel