RFR: 8347009: Speed ​​up parseInt and parseLong [v3]

[j3graham]

On Sat, 4 Jan 2025 12:09:04 GMT, Shaojin Wen <swen at openjdk.org> wrote:

>> This is an optimization for decimal Integer.parseInt and Long.parseLong, which improves performance by about 10%. The optimization includes:
>> 1. Improve performance by parsing 2 numbers at a time, which has performance improvements for numbers with length >= 3.
>> 2. It uses charAt(0) for the first number. Assuming that the optimization can eliminate boundary checks, this will be more friendly to parsing numbers with length 1.
>> 3. It removes the reliance on the Character.digit method and eliminates the reliance on the CharacterDataLatin1#DIGITS cache array, which avoids performance degradation caused by cache misses.
>
> Shaojin Wen has updated the pull request incrementally with one additional commit since the last revision:
> 
>   use String::value

Some other thoughts that might help here; except for leading zeros, 9 (or 18 for a long) or fewer digits cannot overflow. So something like this might work:
 - read first char and check if it is a sign character.
 - skip leading zeros
 - loop over up to 9 digits, performing the calculation without needing to check for overflow. I think it would be enough to do this 1 digit at a time.
 - If there are any digits left, take only the next digit, and incorporate it into the result while doing the overflow check.
 - If there are still more digits left it is too big to fit.
 
 This approach could also be taken for the more general parse methods.

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

PR Comment: https://git.openjdk.org/jdk/pull/22919#issuecomment-2571374257