style: blank between sections

docs: README

feat: hand-crafted simd line parser

depend: remove unused dependencies

optimize: simd bitmap

style: clippy

optimize: parallel graph building

optimize: parallel dedup

feat: more tunables

feat: tunable chunk size ratio

feat: external crate rayon-k-way-merge

style: cargo clippy

optimize: brand-new parallel merging algorithm

optimize(counting neighbors): reduce worst complexity from nlogn to n

optimize: remove unnecessary vector initialization

optimize: pre-allocated vectors

optimize: parallel vector concat

optimize: hand-made parallel line parser

optimize: more effecient neighbor counting algorithm

optimize: cargo wizard

optimize: sort using primitive type

optimize: voracious multi-threaded radix sort

optimize: introduce smallvec

ciallo~

optimize: lower level line parsing

optimize: use mmap to minimize copy between kernel and userspace.

optimize

feat: scope timers

style: cargo clippy

optimize: multi-threaded resolver.