From 35454de390b28c4becb90067c0020537d4e3b5de Mon Sep 17 00:00:00 2001
From: Patrick Marks <patrick@10xgenomics.com>
Date: Thu, 26 Sep 2024 11:30:29 +0200
Subject: [PATCH 1/2] make sure to exhaust any un-read bytes held within lz4
 context

---
 src/decoder.rs | 52 ++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 48 insertions(+), 4 deletions(-)

diff --git a/src/decoder.rs b/src/decoder.rs
index c13e3b1c3..9cf3e49fa 100644
--- a/src/decoder.rs
+++ b/src/decoder.rs
@@ -69,13 +69,13 @@ impl<R: Read> Read for Decoder<R> {
                     self.next
                 };
                 self.len = self.r.read(&mut self.buf[0..need])?;
-                if self.len == 0 {
-                    break;
-                }
+                // NOTE: we do not exit here if there was nothing read
+                // The lz4 context may still have more bytes to emit.
+
                 self.pos = 0;
                 self.next -= self.len;
             }
-            while (dst_offset < buf.len()) && (self.pos < self.len) {
+            while (dst_offset < buf.len()) && ((self.pos < self.len) || self.len == 0) {
                 let mut src_size = (self.len - self.pos) as size_t;
                 let mut dst_size = (buf.len() - dst_offset) as size_t;
                 let len = check_error(unsafe {
@@ -90,6 +90,14 @@ impl<R: Read> Read for Decoder<R> {
                 })?;
                 self.pos += src_size as usize;
                 dst_offset += dst_size as usize;
+
+                // We need to keep trying to read bytes from the decompressor
+                // until it is no longer emitting them, even after it 
+                // has finished reading bytes.
+                if dst_size == 0 && src_size == 0 {
+                    return Ok(dst_offset);
+                }
+
                 if len == 0 {
                     self.next = 0;
                     return Ok(dst_offset);
@@ -290,6 +298,42 @@ mod test {
         finish_decode(decoder);
     }
 
+    /// Ensure that we emit the full decompressed stream even if we're
+    /// using a very small output buffer.
+    #[test]
+    fn issue_45() {
+        // create an encoder
+        let mut enc = crate::EncoderBuilder::new()
+            .build(Vec::new()).unwrap();
+
+        // write 'a' 100 times to the encoder
+        let text: Vec<u8> = vec!['a' as u8; 100];
+        enc.write_all(&text[..]).unwrap();
+
+        // flush the encoder
+        enc.flush().unwrap();
+
+
+        // read from the decoder, buf_size bytes at a time
+        for buf_size in [5, 10, 15, 20, 25] {
+            let mut buf = vec![0; buf_size];
+
+            let mut total_bytes_read = 0;
+
+            // create a decoder wrapping the backing buffer
+            let mut dec = crate::Decoder::new(&enc.writer()[..]).unwrap();
+            while let Ok(n) = dec.read(&mut buf[..]) {
+                if n == 0 {
+                    break;
+                }
+                
+                total_bytes_read += n;
+            }
+
+            assert_eq!(total_bytes_read, text.len());
+        }
+    }
+
     fn random() -> StdRng {
         let seed: [u8; 32] = [
             157, 164, 190, 237, 231, 103, 60, 22, 197, 108, 51, 176, 30, 170, 155, 21, 163, 249,

From 4f59e42dc760b5adc79fa8c5fb775e66b599bdd3 Mon Sep 17 00:00:00 2001
From: Patrick Marks <patrick@10xgenomics.com>
Date: Thu, 26 Sep 2024 11:31:37 +0200
Subject: [PATCH 2/2] fmt

---
 src/decoder.rs | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/src/decoder.rs b/src/decoder.rs
index 9cf3e49fa..f35af43dc 100644
--- a/src/decoder.rs
+++ b/src/decoder.rs
@@ -92,7 +92,7 @@ impl<R: Read> Read for Decoder<R> {
                 dst_offset += dst_size as usize;
 
                 // We need to keep trying to read bytes from the decompressor
-                // until it is no longer emitting them, even after it 
+                // until it is no longer emitting them, even after it
                 // has finished reading bytes.
                 if dst_size == 0 && src_size == 0 {
                     return Ok(dst_offset);
@@ -303,8 +303,7 @@ mod test {
     #[test]
     fn issue_45() {
         // create an encoder
-        let mut enc = crate::EncoderBuilder::new()
-            .build(Vec::new()).unwrap();
+        let mut enc = crate::EncoderBuilder::new().build(Vec::new()).unwrap();
 
         // write 'a' 100 times to the encoder
         let text: Vec<u8> = vec!['a' as u8; 100];
@@ -313,7 +312,6 @@ mod test {
         // flush the encoder
         enc.flush().unwrap();
 
-
         // read from the decoder, buf_size bytes at a time
         for buf_size in [5, 10, 15, 20, 25] {
             let mut buf = vec![0; buf_size];
@@ -326,7 +324,7 @@ mod test {
                 if n == 0 {
                     break;
                 }
-                
+
                 total_bytes_read += n;
             }