The performance of String.encodeUTF8 may be improved as suggested in this email to core-libs-dev.
It proposed to copy the prefix known to be ascii from countPositives in bulk before entering the character at a time loop.
https://mail.openjdk.org/pipermail/core-libs-dev/2025-July/149417.html
It proposed to copy the prefix known to be ascii from countPositives in bulk before entering the character at a time loop.
https://mail.openjdk.org/pipermail/core-libs-dev/2025-July/149417.html
- links to
-
Review(master) openjdk/jdk/26597