From 5161ab0f2d08d705cb6b42a0542720dfa789f3b2 Mon Sep 17 00:00:00 2001
From: Mark Juggurnauth-Thomas <mbthomas@meta.com>
Date: Thu, 19 Oct 2023 09:22:09 +0100
Subject: [PATCH] Add a test demonstrating underread of underlying buffers

When reading from a decompressor that was constructed using `Decoder::with_buffer`, the decoder may not consume all of the input by the time it returns all of the decompressed output.  This means when you call `.finish()` on the decoder to get the underlying stream back, it is not pointing after the end of the compressed data.

This commit adds a test that demonstrates the issue.
---
 assets/zeros64.zst | Bin 0 -> 21 bytes
 tests/issue_251.rs |  38 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 38 insertions(+)
 create mode 100644 assets/zeros64.zst
 create mode 100644 tests/issue_251.rs

diff --git a/assets/zeros64.zst b/assets/zeros64.zst
new file mode 100644
index 0000000000000000000000000000000000000000..37cb6753c20948e0013f7f3223570491f0fcaf45
GIT binary patch
literal 21
ccmdPcs{dET!IgnQfPsOLVKRf5q}C31067c<uK)l5

literal 0
HcmV?d00001

diff --git a/tests/issue_251.rs b/tests/issue_251.rs
new file mode 100644
index 00000000..37ad1588
--- /dev/null
+++ b/tests/issue_251.rs
@@ -0,0 +1,38 @@
+use std::io::Read;
+
+#[test]
+fn test_issue_251() {
+    // This is 64 compressed zero bytes.
+    let compressed_data = include_bytes!("../assets/zeros64.zst");
+    let decompressed_size = 64;
+
+    // Construct a decompressor using `with_buffer`.  This should be ok as
+    // `Cursor` is `BufRead`.
+    let reader = std::io::Cursor::new(compressed_data);
+    let mut decomp = zstd::Decoder::with_buffer(reader).unwrap();
+
+    // Count how many bytes we decompress.
+    let mut total = 0;
+
+    // Decompress four bytes at a time (this is necessary to trigger underrun
+    // behaviour).
+    for _ in 0..(decompressed_size / 4) {
+        let mut buf = [0u8; 4];
+        let count = decomp.read(&mut buf).unwrap();
+        total += count;
+    }
+
+    // Finish reading and get the buffer back.
+    let reader = decomp.finish();
+
+    // The cursor should now be at the end of the compressed data.
+    println!("We read {total}/{decompressed_size} decompressed bytes");
+    println!(
+        "The underlying cursor is at position {} of {} compressed bytes",
+        reader.position(),
+        compressed_data.len()
+    );
+
+    assert_eq!(total, decompressed_size);
+    assert_eq!(reader.position() as usize, compressed_data.len());
+}