A different approach to the problem reported in

Patch #419651: Metrowerks on Mac adds 0x itself
C std says %#x and %#X conversion of 0 do not add the 0x/0X base marker.
Metrowerks apparently does.  Mark Favas reported the same bug under a
Compaq compiler on Tru64 Unix, but no other libc broken in this respect
is known (known to be OK under MSVC and gcc).
So just try the damn thing at runtime and see what the platform does.
Note that we've always had bugs here, but never knew it before because
a relevant test case didn't exist before 2.1.

2 files changed, 54 insertions, 37 deletions

diff --git a/Objects/stringobject.c b/Objects/stringobject.c
index 861cade..cb781ed 100644
--- a/Objects/stringobject.c
+++ b/Objects/stringobject.c
@@ -141,8 +141,8 @@ PyObject *PyString_Decode(const char *s,
 			  const char *errors)
 {
     PyObject *buffer = NULL, *str;
-    
-    if (encoding == NULL) 
+
+    if (encoding == NULL)
 	encoding = PyUnicode_GetDefaultEncoding();
 
     /* Decode via the codec registry */
@@ -169,7 +169,7 @@ PyObject *PyString_Decode(const char *s,
     }
     Py_DECREF(buffer);
     return str;
-    
+
  onError:
     Py_XDECREF(buffer);
     return NULL;
@@ -181,7 +181,7 @@ PyObject *PyString_Encode(const char *s,
 			  const char *errors)
 {
     PyObject *v, *str;
-    
+
     str = PyString_FromStringAndSize(s, size);
     if (str == NULL)
 	return NULL;
@@ -195,13 +195,13 @@ PyObject *PyString_AsEncodedString(PyObject *str,
 				   const char *errors)
 {
     PyObject *v;
-    
+
     if (!PyString_Check(str)) {
         PyErr_BadArgument();
         goto onError;
     }
 
-    if (encoding == NULL) 
+    if (encoding == NULL)
 	encoding = PyUnicode_GetDefaultEncoding();
 
     /* Encode via the codec registry */
@@ -224,7 +224,7 @@ PyObject *PyString_AsEncodedString(PyObject *str,
         goto onError;
     }
     return v;
-    
+
  onError:
     return NULL;
 }
@@ -272,7 +272,7 @@ PyString_AsString(register PyObject *op)
 }
 
 /* Internal API needed by PyString_AsStringAndSize(): */
-extern 
+extern
 PyObject *_PyUnicode_AsDefaultEncodedString(PyObject *unicode,
 					    const char *errors);
 
@@ -415,7 +415,7 @@ string_concat(register PyStringObject *a, register PyObject *bb)
 	if (!PyString_Check(bb)) {
 		if (PyUnicode_Check(bb))
 		    return PyUnicode_Concat((PyObject *)a, bb);
-		PyErr_Format(PyExc_TypeError, 
+		PyErr_Format(PyExc_TypeError,
 			     "cannot add type \"%.200s\" to string",
 			     bb->ob_type->tp_name);
 		return NULL;
@@ -908,7 +908,7 @@ string_find_internal(PyStringObject *self, PyObject *args, int dir)
 	int n, i = 0, last = INT_MAX;
 	PyObject *subobj;
 
-	if (!PyArg_ParseTuple(args, "O|O&O&:find/rfind/index/rindex", 
+	if (!PyArg_ParseTuple(args, "O|O&O&:find/rfind/index/rindex",
 		&subobj, _PyEval_SliceIndex, &i, _PyEval_SliceIndex, &last))
 		return -2;
 	if (PyString_Check(subobj)) {
@@ -941,14 +941,14 @@ string_find_internal(PyStringObject *self, PyObject *args, int dir)
 	}
 	else {
 		int j;
-	    
+
         	if (n == 0 && i <= last)
 			return (long)last;
 		for (j = last-n; j >= i; --j)
 			if (s[j] == sub[0] && memcmp(&s[j], sub, n) == 0)
 				return (long)j;
 	}
-	
+
 	return -1;
 }
 
@@ -1364,7 +1364,7 @@ string_translate(PyStringObject *self, PyObject *args)
 		tablen = PyString_GET_SIZE(tableobj);
 	}
 	else if (PyUnicode_Check(tableobj)) {
-		/* Unicode .translate() does not support the deletechars 
+		/* Unicode .translate() does not support the deletechars
 		   parameter; instead a mapping to None will cause characters
 		   to be deleted. */
 		if (delobj != NULL) {
@@ -1460,7 +1460,7 @@ string_translate(PyStringObject *self, PyObject *args)
   found, or -1 if not found.  If len of PAT is greater than length of
   MEM, the function returns -1.
 */
-static int 
+static int
 mymemfind(const char *mem, int len, const char *pat, int pat_len)
 {
 	register int ii;
@@ -1483,7 +1483,7 @@ mymemfind(const char *mem, int len, const char *pat, int pat_len)
    meaning mem=1111 and pat==11 returns 2.
            mem=11111 and pat==11 also return 2.
  */
-static int 
+static int
 mymemcnt(const char *mem, int len, const char *pat, int pat_len)
 {
 	register int offset = 0;
@@ -1605,7 +1605,7 @@ string_replace(PyStringObject *self, PyObject *args)
 		sub_len = PyString_GET_SIZE(subobj);
 	}
 	else if (PyUnicode_Check(subobj))
-		return PyUnicode_Replace((PyObject *)self, 
+		return PyUnicode_Replace((PyObject *)self,
 					 subobj, replobj, count);
 	else if (PyObject_AsCharBuffer(subobj, &sub, &sub_len))
 		return NULL;
@@ -1615,7 +1615,7 @@ string_replace(PyStringObject *self, PyObject *args)
 		repl_len = PyString_GET_SIZE(replobj);
 	}
 	else if (PyUnicode_Check(replobj))
-		return PyUnicode_Replace((PyObject *)self, 
+		return PyUnicode_Replace((PyObject *)self,
 					 subobj, replobj, count);
 	else if (PyObject_AsCharBuffer(replobj, &repl, &repl_len))
 		return NULL;
@@ -1669,7 +1669,7 @@ string_startswith(PyStringObject *self, PyObject *args)
 	}
 	else if (PyUnicode_Check(subobj)) {
 	    	int rc;
-		rc = PyUnicode_Tailmatch((PyObject *)self, 
+		rc = PyUnicode_Tailmatch((PyObject *)self,
 					  subobj, start, end, -1);
 		if (rc == -1)
 			return NULL;
@@ -1727,7 +1727,7 @@ string_endswith(PyStringObject *self, PyObject *args)
 	}
 	else if (PyUnicode_Check(subobj)) {
 	    	int rc;
-		rc = PyUnicode_Tailmatch((PyObject *)self, 
+		rc = PyUnicode_Tailmatch((PyObject *)self,
 					  subobj, start, end, +1);
 		if (rc == -1)
 			return NULL;
@@ -1829,9 +1829,9 @@ string_expandtabs(PyStringObject *self, PyObject *args)
     return u;
 }
 
-static 
-PyObject *pad(PyStringObject *self, 
-	      int left, 
+static
+PyObject *pad(PyStringObject *self,
+	      int left,
 	      int right,
 	      char fill)
 {
@@ -1847,13 +1847,13 @@ PyObject *pad(PyStringObject *self,
         return (PyObject *)self;
     }
 
-    u = PyString_FromStringAndSize(NULL, 
+    u = PyString_FromStringAndSize(NULL,
 				   left + PyString_GET_SIZE(self) + right);
     if (u) {
         if (left)
             memset(PyString_AS_STRING(u), fill, left);
-        memcpy(PyString_AS_STRING(u) + left, 
-	       PyString_AS_STRING(self), 
+        memcpy(PyString_AS_STRING(u) + left,
+	       PyString_AS_STRING(self),
 	       PyString_GET_SIZE(self));
         if (right)
             memset(PyString_AS_STRING(u) + left + PyString_GET_SIZE(self),
@@ -2308,7 +2308,7 @@ string_splitlines(PyStringObject *self, PyObject *args)
 #undef SPLIT_APPEND
 
 
-static PyMethodDef 
+static PyMethodDef
 string_methods[] = {
 	/* Counterparts of the obsolete stropmodule functions; except
 	   string.maketrans(). */
@@ -2494,7 +2494,7 @@ formatfloat(char *buf, size_t buflen, int flags,
 	/* worst case length calc to ensure no buffer overrun:
 	     fmt = %#.<prec>g
 	     buf = '-' + [0-9]*prec + '.' + 'e+' + (longest exp
-	        for any double rep.) 
+	        for any double rep.)
 	     len = 1 + prec + 1 + 2 + 5 = 9 + prec
 	   If prec=0 the effective precision is 1 (the leading digit is
 	   always given), therefore increase by one to 10+prec. */
@@ -2517,7 +2517,7 @@ formatfloat(char *buf, size_t buflen, int flags,
  *     The string starting at *pbuf is of the form
  *         "-"? ("0x" | "0X")? digit+
  *     "0x"/"0X" are present only for x and X conversions, with F_ALT
- *         set in flags.  The case of hex digits will be correct, 
+ *         set in flags.  The case of hex digits will be correct,
  *     There will be at least prec digits, zero-filled on the left if
  *         necessary to get that many.
  * val		object to be converted
@@ -2673,9 +2673,15 @@ formatint(char *buf, size_t buflen, int flags,
 	/* When converting 0 under %#x or %#X, C leaves off the base marker,
 	 * but we want it (for consistency with other %#x conversions, and
 	 * for consistency with Python's hex() function).
+	 * BUG 28-Apr-2001 tim:  At least two platform Cs (Metrowerks &
+	 * Compaq Tru64) violate the std by converting 0 w/ leading 0x anyway.
+	 * So add it only if the platform didn't already.
 	 */
-	if (x == 0 && (flags & F_ALT) && (type == 'x' || type == 'X')) {
-		assert(buf[1] != type);  /* else this C *is* adding 0x/0X */
+	if (x == 0 &&
+	   (flags & F_ALT) &&
+	   (type == 'x' || type == 'X') &&
+	    buf[1] != (char)type)  /* this last always true under std C */
+		{
 		memmove(buf+2, buf, strlen(buf) + 1);
 		buf[0] = '0';
 		buf[1] = (char)type;
@@ -2768,7 +2774,7 @@ PyString_Format(PyObject *format, PyObject *args)
 			int len;
 			char formatbuf[FORMATBUFLEN]; /* For format{float,int,char}() */
 			char *fmt_start = fmt;
-			
+
 			fmt++;
 			if (*fmt == '(') {
 				char *keystart;
@@ -2778,7 +2784,7 @@ PyString_Format(PyObject *format, PyObject *args)
 
 				if (dict == NULL) {
 					PyErr_SetString(PyExc_TypeError,
-						 "format requires a mapping"); 
+						 "format requires a mapping");
 					goto error;
 				}
 				++fmt;
@@ -3125,7 +3131,7 @@ PyString_Format(PyObject *format, PyObject *args)
 	Py_DECREF(v);
 	Py_DECREF(args);
 	return w;
-	
+
  error:
 	Py_DECREF(result);
 	if (args_owned) {
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 1d72c0d..e52d628 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -4737,6 +4737,7 @@ formatint(Py_UNICODE *buf,
        + 1 + 1 = 24*/
     char fmt[64]; /* plenty big enough! */
     long x;
+    int use_native_c_format = 1;
 
     x = PyInt_AsLong(v);
     if (x == -1 && PyErr_Occurred())
@@ -4753,11 +4754,21 @@ formatint(Py_UNICODE *buf,
     /* When converting 0 under %#x or %#X, C leaves off the base marker,
      * but we want it (for consistency with other %#x conversions, and
      * for consistency with Python's hex() function).
+     * BUG 28-Apr-2001 tim:  At least two platform Cs (Metrowerks &
+     * Compaq Tru64) violate the std by converting 0 w/ leading 0x anyway.
+     * So add it only if the platform doesn't already.
      */
-    if (x == 0 && (flags & F_ALT) && (type == 'x' || type == 'X'))
-        sprintf(fmt, "0%c%%%s.%dl%c", type, "#", prec, type);
-    else
-        sprintf(fmt, "%%%s.%dl%c", (flags & F_ALT) ? "#" : "", prec, type);
+    if (x == 0 && (flags & F_ALT) && (type == 'x' || type == 'X')) {
+        /* Only way to know what the platform does is to try it. */
+        sprintf(fmt, type == 'x' ? "%#x" : "%#X", 0);
+        if (fmt[1] != (char)type) {
+            /* Supply our own leading 0x/0X -- needed under std C */
+            use_native_c_format = 0;
+            sprintf(fmt, "0%c%%#.%dl%c", type, prec, type);
+        }
+    }
+    if (use_native_c_format)
+         sprintf(fmt, "%%%s.%dl%c", (flags & F_ALT) ? "#" : "", prec, type);
     return usprintf(buf, fmt, x);
 }