Turbopuffer open-sources ultra-fast text search tokenizer

Original: so @morgallant has optimized FTS tokenization throughput to 423 MiB/s and open-sourced it (https://github.com/turbopuffer/alyze)

Source: github.com ↗