From: Tim Peters <tim.peters@gmail.com>
Date: Sun, 16 Dec 2001 19:44:20 +0000 (+0000)
Subject: _PyEval_SliceIndex():  explain why a NULL argument is allowed (thanks
X-Git-Tag: v2.2.1c1~335
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=b519638d1ee6715d4c262d804408041d4caf6b5a;p=thirdparty%2FPython%2Fcpython.git

_PyEval_SliceIndex():  explain why a NULL argument is allowed (thanks
to Guido for the revelation).
---

diff --git a/Python/ceval.c b/Python/ceval.c
index f99e44a9aef5..91a74e43aea8 100644
--- a/Python/ceval.c
+++ b/Python/ceval.c
@@ -3332,14 +3332,14 @@ loop_subscript(PyObject *v, PyObject *w)
    Silently reduce values larger than INT_MAX to INT_MAX, and silently
    boost values less than -INT_MAX to 0.  Return 0 on error, 1 on success.
 */
-/* XXX If v is NULL, this goes out of its way to indicate success(!), but
-   XXX doesn't store into *pi.  Why isn't that an error, or at least v!=NULL
-   XXX an asserted precondition?
+/* Note:  If v is NULL, return success without storing into *pi.  This
+   is because_PyEval_SliceIndex() is called by apply_slice(), which can be
+   called by the SLICE opcode with v and/or w equal to NULL.
 */
 int
 _PyEval_SliceIndex(PyObject *v, int *pi)
 {
-	if (v != NULL) {  /* XXX why isn't this assert(v != NULL()? */
+	if (v != NULL) {
 		long x;
 		if (PyInt_Check(v)) {
 			x = PyInt_AsLong(v);