Don’t stop early: Case-folding source code at memory speed
How a branch-free loop and byte-space arithmetic let GitHub case-fold every byte of code search at >45 GiB/s on a single core.
Read the full story on GitHub Blog.