Fixed C decoder failing to decode byte strings longer than 65536 bytes

Fixes agronholm#208.

docs/versionhistory.rst

@@ -8,6 +8,8 @@ This library adheres to `Semantic Versioning <http://semver.org/>`_.
 **UNRELEASED**
 
 - Fixed use-after-free in the decoder's C version when prematurely encountering the end of stream
+- Fixed the C version of the decoder improperly raising ``CBORDecodeEOF`` when decoding a text
+  string longer than 65536 bytes
 
 **5.6.0** (2024-01-17)
 

source/decoder.c

@@ -597,22 +597,22 @@ decode_definite_long_bytestring(CBORDecoderObject *self, Py_ssize_t length)
     PyObject *buffer = NULL;
     Py_ssize_t left = length;
     while (left) {
-        Py_ssize_t chunk_length = length <= 65536 ? length : 65536;
+        Py_ssize_t chunk_length = left <= 65536 ? left : 65536;
         PyObject *chunk = fp_read_object(self, chunk_length);
         if (!chunk) {
-            break;
+            goto error;
         }
 
         if (!PyBytes_CheckExact(chunk)) {
             Py_DECREF(chunk);
-            break;
+            goto error;
         }
 
         if (buffer) {
             PyObject *new_buffer = PyByteArray_Concat(buffer, chunk);
             Py_DECREF(chunk);
             if (!new_buffer)
-                break;
+                goto error;
 
             if (new_buffer != buffer) {
                 Py_DECREF(buffer);
@@ -622,7 +622,7 @@ decode_definite_long_bytestring(CBORDecoderObject *self, Py_ssize_t length)
             buffer = PyByteArray_FromObject(chunk);
             Py_DECREF(chunk);
             if (!buffer)
-                break;
+                goto error;
         }
         left -= chunk_length;
     }
@@ -638,6 +638,9 @@ decode_definite_long_bytestring(CBORDecoderObject *self, Py_ssize_t length)
         }
     }
     return ret;
+error:
+    Py_XDECREF(buffer);
+    return NULL;
 }
 
 

tests/test_decoder.py

@@ -17,8 +17,6 @@
 from uuid import UUID
 
 import pytest
-from hypothesis import example, given
-from hypothesis.strategies import binary
 
 from cbor2 import FrozenDict
 
@@ -215,21 +213,16 @@ def test_special(impl, special_values):
 @pytest.mark.parametrize(
     "payload, expected",
     [
-        ("40", b""),
-        ("4401020304", b"\x01\x02\x03\x04"),
+        pytest.param("40", b"", id="blank"),
+        pytest.param("4401020304", b"\x01\x02\x03\x04", id="short"),
+        pytest.param("5a00011170" + "12" * 70000, b"\x12" * 70000, id="long"),
     ],
 )
 def test_binary(impl, payload, expected):
     decoded = impl.loads(unhexlify(payload))
     assert decoded == expected
 
 
-@given(binary(min_size=2**6, max_size=2**20))
-@example(b"\x12" * 65537)  # anything over 2**16 fails in C
-def test_binary_roundtrip(impl, expected):
-    assert expected == impl.loads(impl.dumps(expected)), "Binary string fails to round-trip"
-
-
 @pytest.mark.parametrize(
     "payload, expected",
     [