Hi Brian,
Looks fine; cheers,
-Joe
On 2/1/2013 10:59 AM, Brian Burkhalter wrote:
Hello,
The assertion is due to a condition occurring which is contrary to the
specification. The specification is accurate but insufficiently forthcoming
that the %a conversion is unsupported for BigDecimal. This change clarifies the
situation by updating the conversion table which occurs earlier in the class
comments.
diff -r d2495b9984fa -r 4770eaa01d38 src/share/classes/java/util/Formatter.java
--- a/src/share/classes/java/util/Formatter.java Fri Feb 01 07:39:41
2013 +0800
+++ b/src/share/classes/java/util/Formatter.java Fri Feb 01 10:46:23
2013 -0800
@@ -351,7 +351,9 @@
* <tr><td valign="top">{@code 'a'}, {@code 'A'}
* <td valign="top"> floating point
* <td> The result is formatted as a hexadecimal floating-point number
with
- * a significand and an exponent
+ * a significand and an exponent. This conversion is <b>not</b> supported
+ * for the {@code BigDecimal} type despite the latter's being in the
+ * <i>floating point</i> argument category.
*
* <tr><td valign="top">{@code 't'}, {@code 'T'}
* <td valign="top"> date/time
Thanks,
Brian
