Change UnicodeDecodeError objects so that the 'object' attribute

is a bytes object.

Add 'y' and 'y#' format specifiers that work like 's' and 's#'
but only accept bytes objects.

2 files changed, 60 insertions, 0 deletions

diff --git a/Python/getargs.c b/Python/getargs.c
index f7a6604..8331a18 100644
--- a/Python/getargs.c
+++ b/Python/getargs.c
@@ -819,6 +819,32 @@ convertsimple(PyObject *arg, const char **p_format, va_list *p_va, int flags,
 		break;
 	}
 
+	case 'y': {/* bytes */
+		if (*format == '#') {
+			void **p = (void **)va_arg(*p_va, char **);
+			FETCH_SIZE;
+			
+			if (PyBytes_Check(arg)) {
+				*p = PyBytes_AS_STRING(arg);
+				STORE_SIZE(PyBytes_GET_SIZE(arg));
+			}
+			else
+				return converterr("bytes", arg, msgbuf, bufsize);
+			format++;
+		} else {
+			char **p = va_arg(*p_va, char **);
+			
+			if (PyBytes_Check(arg))
+				*p = PyBytes_AS_STRING(arg);
+			else
+				return converterr("bytes", arg, msgbuf, bufsize);
+			if ((Py_ssize_t)strlen(*p) != PyBytes_Size(arg))
+				return converterr("bytes without null bytes",
+						  arg, msgbuf, bufsize);
+		}
+		break;
+	}
+
 	case 'z': {/* string, may be NULL (None) */
 		if (*format == '#') { /* any buffer-like object */
 			void **p = (void **)va_arg(*p_va, char **);
@@ -1595,6 +1621,7 @@ skipitem(const char **p_format, va_list *p_va, int flags)
 	
 	case 's': /* string */
 	case 'z': /* string or None */
+	case 'y': /* bytes */
 	case 'u': /* unicode string */
 	case 't': /* buffer, read-only */
 	case 'w': /* buffer, read-write */
diff --git a/Python/modsupport.c b/Python/modsupport.c
index af774f0..8f600dc 100644
--- a/Python/modsupport.c
+++ b/Python/modsupport.c
@@ -424,6 +424,39 @@ do_mkvalue(const char **p_format, va_list *p_va, int flags)
 			return v;
 		}
 
+		case 'y':
+		{
+			PyObject *v;
+			char *str = va_arg(*p_va, char *);
+			Py_ssize_t n;
+			if (**p_format == '#') {
+				++*p_format;
+				if (flags & FLAG_SIZE_T)
+					n = va_arg(*p_va, Py_ssize_t);
+				else
+					n = va_arg(*p_va, int);
+			}
+			else
+				n = -1;
+			if (str == NULL) {
+				v = Py_None;
+				Py_INCREF(v);
+			}
+			else {
+				if (n < 0) {
+					size_t m = strlen(str);
+					if (m > PY_SSIZE_T_MAX) {
+						PyErr_SetString(PyExc_OverflowError,
+							"string too long for Python bytes");
+						return NULL;
+					}
+					n = (Py_ssize_t)m;
+				}
+				v = PyBytes_FromStringAndSize(str, n);
+			}
+			return v;
+		}
+
 		case 'N':
 		case 'S':
 		case 'O':