<i18n dev> RFR: 8358880: Performance of parsing with DecimalFormat can be improved [v3]

[Justin Lu]

On Thu, 12 Jun 2025 15:26:55 GMT, Johannes Graham <duke at openjdk.org> wrote:

>> src/java.base/share/classes/jdk/internal/math/FloatingDecimal.java line 1841:
>> 
>>> 1839: 
>>> 1840:     static ASCIIToBinaryConverter readDoubleSignlessDigits(int decExp, char[] digits, int length) {
>>> 1841:         if (decExp < MIN_DECIMAL_EXPONENT) {
>> 
>> Is this check needed? I think `ASCIIToBinaryConverter` will return the proper zero value when `doubleValue()` is invoked. 
>> 
>> 
>>   if (decExponent < MIN_DECIMAL_EXPONENT - 1) {
>>       return (isNegative) ? -0.0 : 0.0;
>> 
>> 
>> And if this explicit check is a shortcut, I don't think we would need one for an edge case.
>
> Unfortunately some check is required (a test fails), but I now realize what I had was wrong. The issue is that on line 1084 (https://github.com/openjdk/jdk/pull/25644/files#diff-79e6fd549b5ec5e7f49658581beddcb07fcbb0c09ae8e1117c385b66514da6d2R1084)) exp can overflow and become positive again. I've updated the check to avoid the overflow.

Ah got it, I see your point. We would have goten underflow in `ASCIIToBinaryConverter.doubleValue()` for some extreme cases without a check. 

Is there a specific example you have that requires the switch to the newer check? Adding a comment along those lines might be helpful. Actually, I thought DigitList caps `decimalAt` to Integer.MIN/MAX, so then the first check you had would have been fine. (Maybe I am missing something?)

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/25644#discussion_r2143781479