simplify PyUnicode_DecodeLatin1

Author: timfel

graalpython/com.oracle.graal.python.cext/src/unicodeobject.c

@@ -590,7 +590,12 @@ PyObject * PyUnicode_DecodeASCII(const char *s, Py_ssize_t size, const char *err
 }
 
 PyObject * PyUnicode_DecodeLatin(const char *s, Py_ssize_t size, const char *errors) {
-	return PyUnicode_Decode(s, size, "latin1", errors);
+    if (size < 0) {
+        PyErr_SetString(PyExc_SystemError,
+                        "Negative size passed to PyUnicode_New");
+        return NULL;
+    }
+    return to_sulong(polyglot_from_string_n(s, size, "latin1"));
 }
 
 UPCALL_ID(PyUnicode_Tailmatch);

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/objects/cext/hpy/GraalHPyContextFunctions.java

@@ -1472,6 +1472,7 @@ Object execute(Object[] arguments,
                 } catch (OverflowException | InteropException ex) {
                     throw CompilerDirectives.shouldNotReachHere(ex);
                 }
+                // TODO: TruffleString - when we have ISO-8859-1, we can just force the encoding and short-circuit the error reading etc
                 String errors = castToJavaStringNode.execute(callHPyFunction.call(context, GraalHPyNativeSymbol.POLYGLOT_FROM_STRING, arguments[3], "ascii"));
                 CodingErrorAction errorAction = CodecsModuleBuiltins.convertCodingErrorAction(errors);
                 String result = decode(errorAction, bytes);