mirror of
https://github.com/metabarcoding/obitools4.git
synced 2026-04-30 12:00:39 +00:00
25 lines
1.2 KiB
Markdown
25 lines
1.2 KiB
Markdown
|
# SKM File Reader for Super-Kmers
|
||
|
|
|
||
|
|
This Go package provides a binary file reader (`SkmReader`) for `.skm` files, which store *super-kmers* — compact representations of DNA sequences using 2-bit encoding.
|
||
|
|
|
||
|
|
## Core Functionality
|
||
|
|
|
||
|
|
- **Binary Format Parsing**: Reads structured data from `.skm` files, where each record contains:
|
||
|
|
- A 2-byte little-endian integer specifying the sequence length.
|
||
|
|
- Packed nucleotide data, where every byte encodes up to four bases (2 bits per base).
|
||
|
|
|
||
|
|
- **Decoding Logic**: Converts packed 2-bit codes (`00`, `01`, `10`, `11`) to nucleotide characters using the mapping:
|
||
|
|
`{ 'a', 'c', 'g', 't' }`.
|
||
|
|
|
||
|
|
- **Memory-Efficient Reading**: Uses buffered I/O (64 KiB buffer) for fast sequential access.
|
||
|
|
|
||
|
|
- **Streaming Interface**: `Next()` returns the next super-kmer as a struct with:
|
||
|
|
- `Sequence`: decoded nucleotide byte slice.
|
||
|
|
- `Start`, `End`: positional metadata (currently fixed to full length).
|
||
|
|
|
||
|
|
- **Resource Management**: Provides a clean `.Close()` method for file handle cleanup.
|
||
|
|
|
||
|
|
## Use Case
|
||
|
|
|
||
|
|
Designed for high-performance processing of large genomic datasets (e.g., in k-mer analysis or sequence indexing), where storage size and read speed are critical.
|