Crate kira_cdh_compat_kmer_indexer 

CD-HIT-compatible k-mer indexing (“CD-HIT-NG”) in modern Rust (edition 2024).

New in 0.2:

• O(1) rolling reverse-complement for canonicalization
• Optional postings compression (delta+varint)
• Optional PLR sampling for sub-log locate
• Fast ASCII→2-bit mapping (LUT) + optional AVX2 path behind simd_x86
• Property tests via proptest

Public API remains stable. The Range<u64> returned by KmerIndex::locate is opaque and must be passed unmodified to KmerIndex::postings.

See README for on-disk format v2.

Re-exports

pub use encode::canonical;

pub use encode::encode_kmer;

pub use encode::revcomp;

Modules

encode

K-mer encoding: 2-bit mapping, rolling reverse complement, canonicalization.

Structs

BuildConfig

Build-time configuration.

KmerIndex

K-mer index: query-time API for both owned and mmap-backed layouts.

KmerIndexWriter

Writer that serializes an in-memory KmerIndex to a .kdx file.

Posting

Posting entry: (read_id, pos).

Range

A (half-open) range bounded inclusively below and exclusively above (start..end).

ReaderOptions

Enums

Compression

LocateStrategy

Locate acceleration strategy.

Functions

build_kmer_index_async

build_kmer_index_sync

Build a k-mer index synchronously using rayon.