xsv

Indexing enables constant-time row slicing and accelerates operations like statistics; the README demonstrates slicing 10 rows from a 3-million record file is instantaneous after creating an index.

Joins use a simple hash index for speed, supporting inner, outer, and cross joins on large files in seconds, as shown in the tour with worldcitiespop.csv and countrynames.csv.

Commands are built to be piped together without performance loss, allowing complex workflows like filtering, selecting, and sampling in a single chain, as emphasized in the philosophy.

Statistics and frequency commands leverage parallelism when an index is present, cutting analysis time significantly; the tour shows stats running in 8 seconds vs. 12 seconds without indexing.