feat: various minor code optimizations

nh13 · nh13 · commit 4c6a22dca09d · 2025-03-13T21:09:37.000-07:00
* replace ahash with rustc-hash.  rustc-hash was faster than
  ahash and fxhash empirically.
* optimize how the read name, specifically the read number, is
  written.
* increase the buffer size by 4xfor the output writers to 65K.
* delay creating the read bases from bytes for the display in an
  panic until it is needed.  This occurred during every comparison
  of the read sample barcode with the expected sample barcode, so
  while not individually expenseive, expensive enough in aggregate.
* create the necessary bytes for each sample barcode only once, since
  this was being done every time we compared barcodes.

Miscellaneous changes:

* fix the usage for threads, where it incorrectly said the minimum
  number of threads was three, and should be five.  This also
  changes the README.
* update rust toolchain to 1.85 (from 1.65.0).  This requires minor
  code changes throughout based on clippy requirements and
  formatting.
diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -30,7 +30,6 @@ path = "src/bin/main.rs"
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
 [dependencies]
-ahash = "0.8.11"
 anyhow = "1.0.38"
 bstr = "1.0.1"
 clap = { version = "4.0.25", features = ["derive"] }
@@ -49,6 +48,7 @@ serde-aux = "4.1.2"
 seq_io = "0.3.1"
 thiserror = "1.0.37"
 proglog = {version = "0.3.0", features = ["pretty_counts"] }
+rustc-hash = "2.1.1"
 
 [dev-dependencies]
 csv = "1.1.6"
diff --git a/README.md b/README.md
@@ -161,7 +161,7 @@ Options:
           [default: 2]
 
   -t, --threads <THREADS>
-          The number of threads to use. Cannot be less than 3
+          The number of threads to use. Cannot be less than 5
 
           [default: 8]
 
@@ -176,10 +176,10 @@ Options:
           1. `too-few-bases`: there are too few bases or qualities to extract given the read structures.  For example, if a read is 8bp long but the read structure is `10B`, or if a read is empty and the read structure is `+T`.
 
   -h, --help
-          Print help information (use `-h` for a summary)
+          Print help (see a summary with '-h')
 
   -V, --version
-          Print version information
+          Print version
 ```
 <!-- end usage -->
 
diff --git a/rust-toolchain.toml b/rust-toolchain.toml
@@ -1,3 +1,3 @@
 [toolchain]
-channel = "1.85"
+channel = "1.85.0"
 components = ["rustfmt", "clippy"]
diff --git a/src/bin/commands/demux.rs b/src/bin/commands/demux.rs
@@ -93,6 +93,7 @@ impl ReadSet {
     const SPACE: u8 = b' ';
     const COLON: u8 = b':';
     const PLUS: u8 = b'+';
+    const READ_NUMBERS: &[u8] = b"12345678";
 
     /// Produces an iterator over references to the template segments stored in this ``ReadSet``.
     fn template_segments(&self) -> SegmentIter {
@@ -213,7 +214,12 @@ impl ReadSet {
             None => {
                 // If no pre-existing comment, assume the read is a passing filter, non-control
                 // read and generate a comment for it (sample barcode is added below).
-                write!(writer, "{}:N:0:", read_num)?;
+                if read_num < Self::READ_NUMBERS.len() {
+                    writer.write_all(&[Self::READ_NUMBERS[read_num - 1]])?;
+                    write!(writer, ":N:0:")?;
+                } else {
+                    write!(writer, "{}:N:0:", read_num)?;
+                }
             }
             Some(chars) => {
                 // Else check it's a 4-part name... fix the read number at the front and
@@ -239,7 +245,11 @@ impl ReadSet {
                         &chars[first_colon_idx + 1..chars.len()]
                     };
 
-                    write!(writer, "{}:", read_num)?;
+                    if read_num < Self::READ_NUMBERS.len() {
+                        writer.write_all(&[Self::READ_NUMBERS[read_num - 1], b':'])?;
+                    } else {
+                        write!(writer, "{}:", read_num)?;
+                    }
                     writer.write_all(remainder)?;
 
                     if *remainder.last().unwrap() != Self::COLON {
@@ -619,7 +629,7 @@ pub(crate) struct Demux {
     #[clap(long, short = 'd', default_value = "2")]
     min_mismatch_delta: usize,
 
-    /// The number of threads to use. Cannot be less than 3.
+    /// The number of threads to use. Cannot be less than 5.
     #[clap(long, short = 't', default_value = "8")]
     threads: usize,
 
@@ -666,9 +676,12 @@ impl Demux {
                 read_structures.iter().map(|s| s.segments_by_type(*output_type).count()).sum();
 
             for idx in 1..=segment_count {
-                output_type_writers.push(BufWriter::new(File::create(
-                    output_dir.join(format!("{}.{}{}.fq.gz", prefix, file_type_code, idx)),
-                )?));
+                output_type_writers.push(BufWriter::with_capacity(
+                    65_536usize,
+                    File::create(
+                        output_dir.join(format!("{}.{}{}.fq.gz", prefix, file_type_code, idx)),
+                    )?,
+                ));
             }
 
             match output_type {
@@ -1189,6 +1202,7 @@ mod tests {
             skip_reasons: vec![],
         };
         let demux_result = demux_inputs.execute();
+        #[allow(clippy::permissions_set_readonly_false)]
         permissions.set_readonly(false);
         fs::set_permissions(tmp.path(), permissions).unwrap();
         demux_result.unwrap();
@@ -1963,7 +1977,7 @@ mod tests {
             vec!["AAAAAAA", &SAMPLE1_BARCODE[0..7]], // barcode too short
             vec!["CCCCCCC", SAMPLE1_BARCODE],        // barcode the correct length
             vec!["", SAMPLE1_BARCODE],               // template basese too short
-            vec!["G", SAMPLE1_BARCODE],              // barcode the correct length
+            vec!["G", SAMPLE1_BARCODE],
         ];
 
         let input_files = vec![
@@ -1999,7 +2013,7 @@ mod tests {
             vec!["AAAAAAA", &SAMPLE1_BARCODE[0..7]], // barcode too short
             vec!["CCCCCCC", SAMPLE1_BARCODE],        // barcode the correct length
             vec!["", SAMPLE1_BARCODE],               // template basese too short
-            vec!["G", SAMPLE1_BARCODE],              // barcode the correct length
+            vec!["G", SAMPLE1_BARCODE],
         ];
 
         let input_files = vec![
diff --git a/src/lib/barcode_matching.rs b/src/lib/barcode_matching.rs
@@ -6,8 +6,7 @@ use crate::encode;
 use super::byte_is_nocall;
 use super::samples::Sample;
 use crate::bitenc::BitEnc;
-use ahash::HashMap as AHashMap;
-use ahash::HashMapExt;
+use rustc_hash::FxHashMap;
 
 const STARTING_CACHE_SIZE: usize = 1_000_000;
 
@@ -41,7 +40,7 @@ pub struct BarcodeMatcher {
     /// If true will attempt to use the cache when matching.
     use_cache: bool,
     /// Caching struct for storing results of previous matches
-    cache: AHashMap<Vec<u8>, BarcodeMatch>,
+    cache: FxHashMap<Vec<u8>, BarcodeMatch>,
 }
 
 impl BarcodeMatcher {
@@ -81,7 +80,7 @@ impl BarcodeMatcher {
             max_mismatches,
             min_mismatch_delta,
             use_cache,
-            cache: AHashMap::with_capacity(STARTING_CACHE_SIZE),
+            cache: FxHashMap::with_capacity_and_hasher(STARTING_CACHE_SIZE, Default::default()),
         }
     }
 
@@ -94,9 +93,7 @@ impl BarcodeMatcher {
     ) -> u8 {
         if observed_bases.nr_symbols() != expected_bases.nr_symbols() {
             let observed_string = decode(observed_bases);
-            assert_eq!(
-                observed_bases.nr_symbols(),
-                expected_bases.nr_symbols(),
+            panic!(
                 "Read barcode ({}) length ({}) differs from expected barcode ({}) length ({}) for sample {}",
                 observed_string,
                 observed_bases.nr_symbols(),
@@ -196,6 +193,7 @@ mod tests {
     fn barcode_to_sample(barcode: &str, idx: usize) -> Sample {
         Sample {
             barcode: barcode.to_string(),
+            barcode_bytes: barcode.as_bytes().to_vec(),
             sample_id: format!("sample_{idx}").to_string(),
             ordinal: idx,
         }
diff --git a/src/lib/mod.rs b/src/lib/mod.rs
@@ -53,11 +53,7 @@ pub fn encode(bases: &[u8]) -> BitEnc {
             IUPAC_MASKS[b'N' as usize]
         } else {
             let value = base.to_ascii_uppercase() as usize;
-            if value < 256 {
-                IUPAC_MASKS[value]
-            } else {
-                0
-            }
+            if value < 256 { IUPAC_MASKS[value] } else { 0 }
         };
         vec.push(bit);
     }
diff --git a/src/lib/samples.rs b/src/lib/samples.rs
@@ -19,6 +19,9 @@ pub struct Sample {
     pub sample_id: String,
     /// DNA barcode associated with the sample
     pub barcode: String,
+    /// DNA barcode as a byte
+    #[serde(skip_deserializing)]
+    pub barcode_bytes: Vec<u8>,
     /// index of the sample in the [`SampleGroup`] object, used for syncing indices across
     /// different structs
     #[serde(skip_deserializing)]
@@ -53,7 +56,8 @@ impl Sample {
             barcode.as_bytes().iter().all(|&b| is_valid_iupac(b)),
             "All sample barcode bases must be one of A, C, G, T, U, R, Y, S, W, K, M, D, V, H, B, N"
         );
-        Self { sample_id: name, barcode, ordinal }
+        let barcode_bytes = barcode.as_bytes().to_vec();
+        Self { sample_id: name, barcode, barcode_bytes, ordinal }
     }
 
     /// Returns the header line expected by serde when deserializing
@@ -294,11 +298,7 @@ mod tests {
         let barcode = "GATTACA".to_owned();
         let ordinal = 0;
         let sample = Sample::new(ordinal, name.clone(), barcode.clone());
-        assert_eq!(
-            Sample { sample_id: name, barcode, ordinal },
-            sample,
-            "Sample differed from expectation"
-        );
+        assert_eq!(Sample::new(ordinal, name, barcode), sample, "Sample differed from expectation");
     }
 
     // ############################################################################################
