seq kmer

kr2r/src/bin/annotate.rs

@@ -235,7 +235,7 @@ pub fn run(args: Args) -> Result<()> {
 
     // 开始计时
     let start = Instant::now();
-    println!("start...");
+    println!("annotate start...");
     for chunk_file in chunk_files {
         println!("chunk_file {:?}", chunk_file);
         process_chunk_file(&args, chunk_file, &hash_files)?;

kr2r/src/bin/build_k2_db.rs

@@ -7,8 +7,8 @@ use kr2r::db::{
     write_config_to_file,
 };
 use kr2r::utils::{
-    create_partition_files, create_partition_writers, find_library_fna_files, format_bytes,
-    get_file_limit, read_id_to_taxon_map,
+    create_partition_files, create_partition_writers, find_library_fna_files, get_file_limit,
+    read_id_to_taxon_map,
 };
 use kr2r::IndexOptions;
 use std::fs::remove_file;
@@ -25,10 +25,9 @@ pub struct Args {
     #[clap(long, value_parser = parse_size, default_value = "1G", help = "Specifies the hash file capacity.\nAcceptable formats include numeric values followed by 'K', 'M', or 'G' (e.g., '1.5G', '250M', '1024K').\nNote: The specified capacity affects the index size, with a factor of 4 applied.\nFor example, specifying '1G' results in an index size of '4G'.\nDefault: 1G (capacity 1G = file size 4G)")]
     pub hash_capacity: usize,
 
-    /// chunk temp directory
-    #[clap(long)]
-    pub chunk_dir: PathBuf,
-
+    // chunk temp directory
+    // #[clap(long)]
+    // pub chunk_dir: PathBuf,
     /// 包含原始配置
     #[clap(flatten)]
     pub build: Build,
@@ -83,11 +82,9 @@ pub fn run(args: Args, required_capacity: usize) -> Result<(), Box<dyn std::erro
         panic!("Exceeds File Number Limit");
     }
 
-    let chunk_files = create_partition_files(partition, &args.chunk_dir, "chunk");
+    let chunk_files = create_partition_files(partition, &k2d_dir, "chunk");
     let mut writers = create_partition_writers(&chunk_files);
 
-    println!("chunk_size {}", format_bytes(chunk_size as f64));
-
     let fna_files = find_library_fna_files(args.build.database);
 
     for fna_file in &fna_files {

kr2r/src/bin/classify.rs

@@ -179,7 +179,6 @@ where
 
     let _ = read_parallel(
         reader,
-        args.num_threads as usize - 2,
         args.num_threads as usize,
         &meros,
         |seqs| {
@@ -341,7 +340,7 @@ pub fn run(args: Args) -> Result<()> {
     if hash_config.hash_capacity == 0 {
         panic!("`hash_capacity` can't be zero!");
     }
-    println!("start...");
+    println!("classify start...");
     let start = Instant::now();
     let meros = idx_opts.as_meros();
     let hash_files = find_and_sort_files(&args.k2d_dir, "hash", ".k2d")?;

kr2r/src/bin/estimate_capacity.rs

@@ -88,7 +88,6 @@ fn process_sequence(
     read_parallel(
         &mut reader,
         args.threads,
-        args.threads - 2,
         &meros,
         |record_set| {
             let mut minimizer_set = HashSet::new();
@@ -152,6 +151,8 @@ pub fn run(args: Args) -> usize {
         panic!("Error: No library.fna files found in the specified directory. Please ensure that the directory contains at least one library.fna file and try again.");
     }
 
+    println!("estimate start... ");
+
     for fna_file in fna_files {
         let args_clone = Args {
             database: source.clone(),

kr2r/src/bin/hashshard.rs

@@ -69,7 +69,7 @@ pub fn run(args: Args) -> IOResult<()> {
     let hash_config = HashConfig::from_hash_header(index_filename)?;
 
     let partition = (hash_config.capacity + args.hash_capacity - 1) / args.hash_capacity;
-    println!("start...");
+    println!("hashshard start...");
     // 开始计时
     let start = Instant::now();
 

kr2r/src/bin/kun.rs

@@ -22,13 +22,7 @@ struct BuildArgs {
     #[clap(long)]
     pub k2d_dir: Option<PathBuf>,
 
-    /// chunk directory
-    #[clap(long)]
-    chunk_dir: PathBuf,
-
-    #[clap(long, value_parser = parse_size, default_value = "1G", help = "Specifies the hash file capacity.\nAcceptable formats include numeric values followed by 'K', 'M', or 'G' (e.g., '1.5G', '250M', '1024K').\nNote: The specified capacity affects the index size, with a factor of 4 applied.\nFor example, specifying '1G' results in an index size of '4G'.\nDefault: 1G (capacity 1G = file size 4G)")]
-    pub hash_capacity: usize,
-
+    // chunk_dir: PathBuf,
     #[clap(flatten)]
     pub build: Build,
 
@@ -117,8 +111,7 @@ impl From<BuildArgs> for build_k2_db::Args {
             build: item.build,
             k2d_dir: item.k2d_dir,
             taxo: item.taxo,
-            chunk_dir: item.chunk_dir,
-            hash_capacity: item.hash_capacity,
+            hash_capacity: parse_size("1G").unwrap(),
         }
     }
 }

kr2r/src/bin/resolve.rs

@@ -223,7 +223,7 @@ pub fn run(args: Args) -> Result<()> {
 
     // 开始计时
     let start = Instant::now();
-    println!("start...");
+    println!("resolve start...");
 
     for i in 0..partition {
         let sample_file = &sample_files[i];

kr2r/src/bin/splitr.rs

@@ -148,7 +148,6 @@ where
 
     read_parallel(
         reader,
-        args.num_threads as usize - 2,
         args.num_threads as usize,
         &meros,
         |seqs| {
@@ -268,7 +267,7 @@ pub fn run(args: Args) -> Result<()> {
     if hash_config.hash_capacity == 0 {
         panic!("`hash_capacity` can't be zero!");
     }
-    println!("start...");
+    println!("splitr start...");
     let file_num_limit = get_file_limit();
     if hash_config.partition >= file_num_limit {
         panic!("Exceeds File Number Limit");

kr2r/src/db.rs

@@ -252,14 +252,12 @@ pub fn convert_fna_to_k2_format<P: AsRef<Path>>(
     threads: u32,
 ) {
     let mut reader = FastaReader::from_path(fna_file, 1).unwrap();
-    let queue_len = (threads - 2) as usize;
     let value_bits = hash_config.value_bits;
     let cell_size = std::mem::size_of::<Slot<u32>>();
 
     read_parallel(
         &mut reader,
         threads as usize,
-        queue_len,
         &meros,
         |seqs| {
             let mut k2_cell_list = Vec::new();

seqkmer/src/fastq.rs

@@ -184,7 +184,7 @@ where
 {
     fn next(&mut self) -> Result<Option<Vec<Base<Vec<u8>>>>> {
         let seqs: Vec<Base<Vec<u8>>> = (0..self.batch_size)
-            .filter_map(|_| self.read_next().transpose()) // 将 Result<Option<_>, _> 转换为 Option<Result<_, _>>
+            .filter_map(|_| self.read_next().transpose())
             .collect::<Result<Vec<_>>>()?;
 
         Ok(Some(seqs).filter(|v| !v.is_empty()))

seqkmer/src/parallel.rs

@@ -41,7 +41,6 @@ pub fn create_reader(
 pub fn read_parallel<R, W, O, F, Out>(
     reader: &mut R,
     n_threads: usize,
-    buffer_len: usize,
     meros: &Meros,
     work: W,
     func: F,
@@ -54,7 +53,7 @@ where
     F: FnOnce(&mut ParallelResult<Option<O>>) -> Out + Send,
 {
     assert!(n_threads > 2);
-    assert!(n_threads <= buffer_len);
+    let buffer_len = n_threads + 2;
     let (sender, receiver) = bounded::<Vec<Base<Vec<u8>>>>(buffer_len);
     let (done_send, done_recv) = bounded::<Option<O>>(buffer_len);
     let receiver = Arc::new(receiver); // 使用 Arc 来共享 receiver