Use buffer protocol to support all byte-like objects

python · vstinner · Mar 12, 2025 · Feb 8, 2025 · Feb 8, 2025 · Feb 9, 2025
commit b0e3eb557a1a7c68ae38c5017290ed024016fbe9
diff --git a/Doc/library/stdtypes.rst b/Doc/library/stdtypes.rst
@@ -2746,7 +2746,7 @@ data and are closely related to string objects in a variety of other ways.
 
       .. versionchanged:: next
          :meth:`bytes.fromhex` now accepts ASCII :class:`bytes` and
-         :class:`bytearray` objects as input.
+         :term:`bytes-like objects <bytes-like object>` as input.
 
 
 
@@ -2837,7 +2837,7 @@ objects.
 
       .. versionchanged:: next
          :meth:`bytearray.fromhex` now accepts ASCII :class:`bytes` and
-         :class:`bytearray` objects as input.
+         :term:`bytes-like objects <bytes-like object>` as input.
 
    A reverse conversion function exists to transform a bytearray object into its
    hexadecimal representation.

diff --git a/Doc/whatsnew/3.14.rst b/Doc/whatsnew/3.14.rst
@@ -355,7 +355,7 @@ Other language changes
   (Contrubuted by Sergey B Kirpichev in :gh:`87790`.)
 
 * The :func:`bytes.fromhex` and :func:`bytearray.fromhex` methods now accept
-  ASCII :class:`bytes` and :class:`bytearray` objects.
+  ASCII :class:`bytes` and :term:`bytes-like objects <bytes-like object>`.
   (Contributed by Daniel Pope in :gh:`129349`.)
 
 * ``\B`` in :mod:`regular expression <re>` now matches empty input string.

diff --git a/Lib/test/test_bytes.py b/Lib/test/test_bytes.py
@@ -457,11 +457,16 @@ def test_fromhex(self):
             self.assertRaises(ValueError, self.type2test.fromhex, c)
 
         # Check that we can parse bytes and bytearray
-        self.assertEqual(self.type2test.fromhex(b' 012abc'), b'\x01\x2a\xbc')
-        self.assertEqual(
-            self.type2test.fromhex(bytearray(b' 012abc')),
-            b'\x01\x2a\xbc',
-        )
+        tests = [
+            ("bytes", bytes),
+            ("bytearray", bytearray),
+            ("memoryview", memoryview),
+            ("array.array", lambda bs: array.array('B', bs)),
+        ]
+        for name, factory in tests:
+            with self.subTest(name=name):
+                self.assertEqual(self.type2test.fromhex(factory(b' 1A 2B 30 ')), b)
+
         # Invalid bytes are rejected
         for u8 in b"\0\x1C\x1D\x1E\x1F\x85\xa0":
             b = bytes([30, 31, u8])

diff --git a/Misc/NEWS.d/next/Core_and_Builtins/2025-02-08-09-55-33.gh-issue-129349.PkcG-l.rst b/Misc/NEWS.d/next/Core_and_Builtins/2025-02-08-09-55-33.gh-issue-129349.PkcG-l.rst
@@ -1,2 +1,2 @@
 :meth:`bytes.fromhex` and :meth:`bytearray.fromhex` now accepts ASCII
-:class:`bytes`/:class:`bytearray` objects.
+:class:`bytes` and :term:`bytes-like objects <bytes-like object>`.
diff --git a/Objects/bytesobject.c b/Objects/bytesobject.c
@@ -2512,6 +2512,8 @@ _PyBytes_FromHex(PyObject *string, int use_bytearray)
     unsigned int t
6D40
op, bot;
     const Py_UCS1 *str, *start, *end;
     _PyBytesWriter writer;
+    Py_buffer view;
+    view.obj = NULL;
 
     _PyBytesWriter_Init(&writer);
     writer.use_bytearray = use_bytearray;
@@ -2534,15 +2536,13 @@ _PyBytes_FromHex(PyObject *string, int use_bytearray)
         }
 
         assert(PyUnicode_KIND(string) == PyUnicode_1BYTE_KIND);
-        str = start = PyUnicode_1BYTE_DATA(string);
-    }
-    else if (PyBytes_Check(string)) {
-        hexlen = PyBytes_GET_SIZE(string);
-        str = start = (Py_UCS1 *)PyBytes_AS_STRING(string);
+        str = PyUnicode_1BYTE_DATA(string);
     }
-    else if (PyByteArray_Check(string)) {
-        hexlen = PyByteArray_GET_SIZE(string);
-        str = start = (Py_UCS1 *)PyByteArray_AS_STRING(string);
+    else if (PyObject_CheckBuffer(string)) {
+        if (PyObject_GetBuffer(string, &view, PyBUF_SIMPLE) != 0)
+            return NULL;
+        hexlen = view.len;
+        str = view.buf;
     }
     else {
         PyErr_Format(PyExc_TypeError,
@@ -2554,8 +2554,9 @@ _PyBytes_FromHex(PyObject *string, int use_bytearray)
     /* This overestimates if there are spaces */
     buf = _PyBytesWriter_Alloc(&writer, hexlen / 2);
     if (buf == NULL)
-        return NULL;
+        goto release_buffer;
 
+    start = str;
     end = str + hexlen;
     while (str < end) {
         /* skip over spaces in the input */
@@ -2589,6 +2590,9 @@ _PyBytes_FromHex(PyObject *string, int use_bytearray)
         *buf++ = (unsigned char)((top << 4) + bot);
     }
 
+    if (view.obj != NULL) {
+       PyBuffer_Release(&view);
+    }
     return _PyBytesWriter_Finish(&writer, buf);
 
   error:
@@ -2601,6 +2605,11 @@ _PyBytes_FromHex(PyObject *string, int use_bytearray)
                      "fromhex() arg at position %zd", invalid_char);
     }
     _PyBytesWriter_Dealloc(&writer);
+
+  release_buffer:
+    if (view.obj != NULL) {
+        PyBuffer_Release(&view);
+    }
     return NULL;
 }