mirror of
https://github.com/metabarcoding/obitools4.git
synced 2026-04-30 12:00:39 +00:00
Eric Coissac
8c7017a99d
- Update obioptions.Version from "Release 4.4.29" to "/v/ Release v5" - Update version.txt from 4.29 → .30 (automated by Makefile)
24 lines
1.4 KiB
Markdown
24 lines
1.4 KiB
Markdown
# SKM File Format Specification
|
||
|
||
This Go package implements a binary format for storing *super-kmers*—compact representations of DNA sequences used in bioinformatics. The tests validate reading/writing, padding behavior, and file size correctness.
|
||
|
||
## Core Functionalities
|
||
|
||
- **SuperKmer Structure**: Each super-kmer stores a DNA sequence (as bytes), likely padded to 4-base boundaries for efficient storage.
|
||
- **SkmWriter**: Serializes super-kmers into a binary file. Each entry writes:
|
||
- A 2-byte little-endian length (number of bases),
|
||
- Then `ceil(length/4)` bytes encoding nucleotides in 2 bits each (A=0, C=1, G=2, T=3).
|
||
- **SkmReader**: Parses the binary format back into memory. Returns `(SuperKmer, bool)` via `Next()`, with EOF signaled by `ok = false`.
|
||
- **Case Handling**: Writes preserve original case; reads normalize to lowercase (via `| 0x20` in tests), ensuring robust comparison.
|
||
|
||
## Test Coverage
|
||
|
||
- **Round-trip integrity**: Verifies exact sequence recovery after write/read.
|
||
- **Empty file handling**: Confirms reader returns `ok = false` immediately on empty files.
|
||
- **Variable-length padding**: Validates correct encoding/decoding for sequences of length 1–5.
|
||
- **Size validation**: Confirms file size = `2 + ceil(L/4)` bytes for a sequence of length *L*.
|
||
|
||
## Use Case
|
||
|
||
Efficient, lossless storage and retrieval of super-kmers for downstream genomic analysis (e.g., assembly or alignment acceleration).
|