Merge pull request #79 from ArcInstitute/integrate-cbq

Integrate cbq

Author: noamteyssier

Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "binseq"
 version = "0.8.3"
-edition = "2021"
+edition = "2024"
 description = "A high efficiency binary format for sequencing data"
 license = "MIT"
 authors = ["Noam Teyssier <noam.teyssier@arcinstitute.org>"]
@@ -13,13 +13,15 @@ keywords = ["bioinformatics", "nucleotide", "sequencing", "genomics", "fastq"]
 [dependencies]
 anyhow = "1.0.100"
 auto_impl = "1.3.0"
-bitnuc = "0.3.2"
-bytemuck = "1.24.0"
+bitnuc = "0.4.0"
+bytemuck = { version = "1.24.0", features = ["derive", "extern_crate_alloc"] }
 byteorder = "1.5.0"
-itoa = "1.0.15"
+itoa = "1.0.17"
+memchr = "2.7.6"
 memmap2 = "0.9.9"
 num_cpus = "1.17.0"
 rand = { version = "0.9.2", features = ["small_rng"] }
+sucds = "0.8.3"
 thiserror = "2.0.17"
 zstd = { version = "0.13.3", features = ["zstdmt"] }
 
@@ -28,8 +30,6 @@ nucgen = "0.2.0"
 niffler = "3.0.0"
 seq_io = "0.3.4"
 parking_lot = "0.12.5"
-itoa = "1.0.15"
-memchr = "2.7.6"
 
 [lints.clippy]
 pedantic = { level = "warn", priority = -1 }

README.md

@@ -10,12 +10,13 @@
 BINSEQ is a binary file format family designed for efficient storage and processing of DNA sequences.
 They make use of two-bit encoding for nucleotides and are optimized for high-performance parallel processing.
 
-BINSEQ currently has two flavors:
+BINSEQ currently has three variants:
 
 1. **BQ**: (`*.bq`) files are for _fixed-length_ records **without** quality scores.
 2. **VBQ**: (`*.vbq`) files are for _variable-length_ records **with optional** quality scores and headers.
+3. **CBQ**: (`*.cbq`) files are for _columnar variable-length_ records **with optional** quality scores and headers.
 
-Both flavors support both single and paired sequences.
+All variants support both single and paired sequences.
 
 ## Getting Started
 

examples/example.rs

@@ -1,5 +1,5 @@
 use std::fs::File;
-use std::io::{stdout, BufWriter, Write};
+use std::io::{BufWriter, Write, stdout};
 use std::sync::Arc;
 
 use anyhow::Result;

examples/parallel_processing.rs

@@ -2,12 +2,12 @@ use std::{
     fs::File,
     io::BufWriter,
     sync::{
-        atomic::{AtomicUsize, Ordering},
         Arc,
+        atomic::{AtomicUsize, Ordering},
     },
 };
 
-use anyhow::{bail, Result};
+use anyhow::{Result, bail};
 use binseq::{
     bq::{self, BinseqHeaderBuilder},
     context::SeqCtx,

examples/parallel_range.rs

@@ -1,6 +1,6 @@
 use binseq::{BinseqReader, BinseqRecord, ParallelProcessor, ParallelReader, Result};
-use std::sync::atomic::{AtomicUsize, Ordering};
 use std::sync::Arc;
+use std::sync::atomic::{AtomicUsize, Ordering};
 
 #[derive(Clone)]
 struct RangeProcessor {

examples/read_write.rs

@@ -3,10 +3,10 @@ use std::{
     io::{BufReader, BufWriter},
 };
 
-use anyhow::{bail, Result};
+use anyhow::{Result, bail};
 use binseq::{
-    bq::{BinseqHeaderBuilder, BinseqWriterBuilder, MmapReader},
     BinseqRecord,
+    bq::{BinseqHeaderBuilder, BinseqWriterBuilder, MmapReader},
 };
 use seq_io::fastq::{Reader, Record};
 

src/bq/reader.rs

@@ -19,8 +19,8 @@ use memmap2::Mmap;
 
 use super::header::{BinseqHeader, SIZE_HEADER};
 use crate::{
-    error::{ReadError, Result},
     BinseqRecord, Error, ParallelProcessor, ParallelReader,
+    error::{ReadError, Result},
 };
 
 /// A reference to a binary sequence record in a memory-mapped file
@@ -692,10 +692,10 @@ impl<R: Read> StreamReader<R> {
     /// * The data format is invalid
     pub fn next_record(&mut self) -> Option<Result<RefRecord<'_>>> {
         // Ensure header is read
-        if self.header.is_none() {
-            if let Some(e) = self.read_header().err() {
-                return Some(Err(e));
-            }
+        if self.header.is_none()
+            && let Some(e) = self.read_header().err()
+        {
+            return Some(Err(e));
         }
 
         let config = self

src/bq/writer.rs

@@ -10,12 +10,12 @@
 use std::io::{BufWriter, Write};
 
 use byteorder::{LittleEndian, WriteBytesExt};
-use rand::{rngs::SmallRng, SeedableRng};
+use rand::{SeedableRng, rngs::SmallRng};
 
 use super::BinseqHeader;
 use crate::{
-    error::{Result, WriteError},
     Policy, RNG_SEED,
+    error::{Result, WriteError},
 };
 
 /// Writes a single flag value to a writer in little-endian format