From f628197cfa07ce8215757d5916220f23539f6f1b Mon Sep 17 00:00:00 2001
From: noamteyssier <22600644+noamteyssier@users.noreply.github.com>
Date: Wed, 1 Jul 2026 08:31:41 -0700
Subject: [PATCH 1/4] tests: added testing for sequential reader and
 N-sequences

---
 src/cbq/write.rs | 48 ++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 48 insertions(+)

diff --git a/src/cbq/write.rs b/src/cbq/write.rs
index 6301f7d..8ff0b53 100644
--- a/src/cbq/write.rs
+++ b/src/cbq/write.rs
@@ -427,6 +427,54 @@ mod tests {
         Ok(())
     }
 
+    /// Sequences containing `N`s, which drive the Elias-Fano `npos` column.
+    /// Blocks with no `N`s at all never populate `z_npos`, so a suite built
+    /// only from `sample_sequences` (all ACGT) never exercises this path.
+    fn sample_sequences_with_n(n_seq: usize, seq_len: usize) -> Vec<Vec<u8>> {
+        const BASES: [u8; 4] = [b'A', b'C', b'G', b'T'];
+        (0..n_seq)
+            .map(|i| {
+                (0..seq_len)
+                    .map(|j| {
+                        if i % 5 == 0 {
+                            b'N'
+                        } else {
+                            BASES[(i as usize + j) % 4]
+                        }
+                    })
+                    .collect::<Vec<u8>>()
+            })
+            .collect()
+    }
+
+    /// Round-trips sequences containing `N`s through a `ColumnarBlockWriter`
+    /// and back through a `Reader`, exercising the Elias-Fano `npos`
+    /// decompression path in [`ColumnarBlock::decompress_columns`].
+    #[test]
+    fn test_roundtrip_sequences_with_n() -> Result<()> {
+        // Small block size so N-bearing sequences span multiple blocks.
+        let block_size = 256;
+        let mut writer = ColumnarBlockWriter::new(Vec::new(), header(block_size))?;
+
+        let seqs = sample_sequences_with_n(1024, 100);
+        assert!(
+            seqs.iter().any(|s| s.contains(&b'N')),
+            "test fixture must actually contain N's to exercise npos"
+        );
+        for seq in &seqs {
+            writer.push(record(seq))?;
+        }
+        writer.finish()?;
+
+        let read_back = read_all_sequences(writer.inner);
+        assert_eq!(
+            read_back, seqs,
+            "round-trip mismatch for N-bearing sequences"
+        );
+
+        Ok(())
+    }
+
     /// `ingest_completed` on a source with no completed blocks is a no-op for
     /// the global writer and preserves the source's incomplete block.
     #[test]

From e1ca0dd2bff541e24966fe451e79347374f8e622 Mon Sep 17 00:00:00 2001
From: noamteyssier <22600644+noamteyssier@users.noreply.github.com>
Date: Wed, 1 Jul 2026 08:31:57 -0700
Subject: [PATCH 2/4] fix: use reserve instead of resize

---
 src/cbq/core/block.rs | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/src/cbq/core/block.rs b/src/cbq/core/block.rs
index 96b75b9..c434a6b 100644
--- a/src/cbq/core/block.rs
+++ b/src/cbq/core/block.rs
@@ -432,6 +432,11 @@ impl ColumnarBlock {
     }
 
     /// Decompress all columns back to native representation
+    ///
+    /// Note: `resize` can be only be used with `copy_decode` if passing
+    /// as `&mut [T]`. Passing a resized `&mut Vec<T>` will lead to an
+    /// append operation, not an overwrite. If passing `&mut Vec<T>`, the
+    /// `Vec` will be resized automatically by `copy_decode`.
     pub fn decompress_columns(&mut self) -> Result<()> {
         // decompress sequence lengths
         {
@@ -450,7 +455,7 @@ impl ColumnarBlock {
 
         // decompress npos
         if !self.z_npos.is_empty() {
-            self.ef_bytes.resize(self.len_nef, 0);
+            self.ef_bytes.reserve(self.len_nef);
             copy_decode(self.z_npos.as_slice(), &mut self.ef_bytes)?;
 
             let ef = EliasFano::deserialize_from(self.ef_bytes.as_slice())?;

From c88f36b2f548fdd4d49be7aa86197f0a657d4dcc Mon Sep 17 00:00:00 2001
From: noamteyssier <22600644+noamteyssier@users.noreply.github.com>
Date: Wed, 1 Jul 2026 08:33:42 -0700
Subject: [PATCH 3/4] chore(semver): bump

---
 Cargo.toml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Cargo.toml b/Cargo.toml
index 8c10bdb..6e31503 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,6 +1,6 @@
 [package]
 name = "binseq"
-version = "0.9.2"
+version = "0.9.3"
 edition = "2024"
 description = "A high efficiency binary format for sequencing data"
 license = "MIT"

From bb749baa35b2f7e45792b18fd332f72548870955 Mon Sep 17 00:00:00 2001
From: noamteyssier <22600644+noamteyssier@users.noreply.github.com>
Date: Wed, 1 Jul 2026 08:41:02 -0700
Subject: [PATCH 4/4] chore: ensure len_nef is cleared

---
 src/cbq/core/block.rs | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/cbq/core/block.rs b/src/cbq/core/block.rs
index c434a6b..fbfe4fc 100644
--- a/src/cbq/core/block.rs
+++ b/src/cbq/core/block.rs
@@ -112,6 +112,7 @@ impl ColumnarBlock {
             self.num_records = 0;
             self.current_size = 0;
             self.num_npos = 0;
+            self.len_nef = 0;
         }
 
         // clear spans