Use Py_ssize_t and don't check overflow in loop

Author: lysnikolaou

Doc/c-api/unicode.rst

@@ -307,7 +307,7 @@ These APIs can be used for fast direct character conversions:
    possible.  This function does not raise exceptions.
 
 
-.. c:function:: Py_ssize_t PyUnicode_ToLower(Py_UCS4 ch, Py_UCS4 *buffer, int size)
+.. c:function:: Py_ssize_t PyUnicode_ToLower(Py_UCS4 ch, Py_UCS4 *buffer, Py_ssize_t size)
 
    Convert *ch* to lower case, store result in *buffer*, which should be
    able to hold as many characters needed for *ch* to be lower cased
@@ -318,7 +318,7 @@ These APIs can be used for fast direct character conversions:
    .. versionadded:: next
 
 
-.. c:function:: Py_ssize_t PyUnicode_ToUpper(Py_UCS4 ch, Py_UCS4 *buffer, int size)
+.. c:function:: Py_ssize_t PyUnicode_ToUpper(Py_UCS4 ch, Py_UCS4 *buffer, Py_ssize_t size)
 
    Convert *ch* to upper case, store result in *buffer*, which should be
    able to hold as many characters needed for *ch* to be upper cased
@@ -329,7 +329,7 @@ These APIs can be used for fast direct character conversions:
    .. versionadded:: next
 
 
-.. c:function:: Py_ssize_t PyUnicode_ToTitle(Py_UCS4 ch, Py_UCS4 *buffer, int size)
+.. c:function:: Py_ssize_t PyUnicode_ToTitle(Py_UCS4 ch, Py_UCS4 *buffer, Py_ssize_t size)
 
    Convert *ch* to title case, store result in *buffer*, which should be
    able to hold as many characters needed for *ch* to be title cased
@@ -340,7 +340,7 @@ These APIs can be used for fast direct character conversions:
    .. versionadded:: next
 
 
-.. c:function:: Py_ssize_t PyUnicode_ToFolded(Py_UCS4 ch, Py_UCS4 *buffer, int size)
+.. c:function:: Py_ssize_t PyUnicode_ToFolded(Py_UCS4 ch, Py_UCS4 *buffer, Py_ssize_t size)
 
    Foldcase *ch*, store result in *buffer*, which should be
    able to hold as many characters needed for *ch* to be foldcased

Include/cpython/unicodeobject.h

@@ -733,28 +733,28 @@ PyAPI_FUNC(int) _PyUnicode_IsAlpha(
     Py_UCS4 ch       /* Unicode character */
     );
 
-PyAPI_FUNC(int) PyUnicode_ToLower(
+PyAPI_FUNC(Py_ssize_t) PyUnicode_ToLower(
     Py_UCS4 ch,     /* Unicode character */
     Py_UCS4 *res,   /* Output buffer */
-    int size        /* Buffer size */
+    Py_ssize_t size        /* Buffer size */
     );
 
-PyAPI_FUNC(int) PyUnicode_ToUpper(
+PyAPI_FUNC(Py_ssize_t) PyUnicode_ToUpper(
     Py_UCS4 ch,     /* Unicode character */
     Py_UCS4 *res,   /* Output buffer */
-    int size        /* Buffer size */
+    Py_ssize_t size        /* Buffer size */
     );
 
-PyAPI_FUNC(int) PyUnicode_ToTitle(
+PyAPI_FUNC(Py_ssize_t) PyUnicode_ToTitle(
     Py_UCS4 ch,     /* Unicode character */
     Py_UCS4 *res,   /* Output buffer */
-    int size        /* Buffer size */
+    Py_ssize_t size        /* Buffer size */
     );
 
-PyAPI_FUNC(int) PyUnicode_ToFolded(
+PyAPI_FUNC(Py_ssize_t) PyUnicode_ToFolded(
     Py_UCS4 ch,     /* Unicode character */
     Py_UCS4 *res,   /* Output buffer */
-    int size        /* Buffer size */
+    Py_ssize_t size        /* Buffer size */
     );
 
 

Modules/_testcapi/unicode.c

@@ -221,7 +221,7 @@ unicode_copycharacters(PyObject *self, PyObject *args)
 }
 
 static PyObject *
-unicode_case_operation(PyObject *str, int (*function)(Py_UCS4, Py_UCS4 *, int))
+unicode_case_operation(PyObject *str, Py_ssize_t (*function)(Py_UCS4, Py_UCS4 *, Py_ssize_t))
 {
     if (!PyUnicode_Check(str)) {
         PyErr_Format(PyExc_TypeError, "expect str type, got %T", str);
@@ -236,7 +236,7 @@ unicode_case_operation(PyObject *str, int (*function)(Py_UCS4, Py_UCS4 *, int))
     Py_UCS4 c = PyUnicode_READ_CHAR(str, 0);
 
     Py_UCS4 buf[3];
-    int chars = function(c, buf, Py_ARRAY_LENGTH(buf));
+    Py_ssize_t chars = function(c, buf, Py_ARRAY_LENGTH(buf));
     if (chars < 0) {
         return NULL;
     }

Objects/unicodectype.c

@@ -198,99 +198,99 @@ Py_UCS4 _PyUnicode_ToLowercase(Py_UCS4 ch)
     return ch + ctype->lower;
 }
 
-int PyUnicode_ToLower(Py_UCS4 ch, Py_UCS4 *res, int size)
+Py_ssize_t PyUnicode_ToLower(Py_UCS4 ch, Py_UCS4 *res, Py_ssize_t size)
 {
     const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
 
     if (ctype->flags & EXTENDED_CASE_MASK) {
         int index = ctype->lower & 0xFFFF;
         int n = ctype->lower >> 24;
+        if (n > size) {
+            PyErr_SetString(PyExc_ValueError, "output buffer is too small");
+            return -1;
+        }
+
         int i;
-        for (i = 0; i < n; i++) {
-            if (i >= size) {
-                PyErr_SetString(PyExc_ValueError, "output buffer is too small");
-                return -1;
-            }
+        for (i = 0; i < n; i++)
             res[i] = _PyUnicode_ExtendedCase[index + i];
-        }
         return n;
     }
 
-    if (0 >= size) {
+    if (size < 1) {
         PyErr_SetString(PyExc_ValueError, "output buffer is too small");
         return -1;
     }
     res[0] = ch + ctype->lower;
     return 1;
 }
 
-int PyUnicode_ToTitle(Py_UCS4 ch, Py_UCS4 *res, int size)
+Py_ssize_t PyUnicode_ToTitle(Py_UCS4 ch, Py_UCS4 *res, Py_ssize_t size)
 {
     const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
 
     if (ctype->flags & EXTENDED_CASE_MASK) {
         int index = ctype->title & 0xFFFF;
         int n = ctype->title >> 24;
+        if (n > size) {
+            PyErr_SetString(PyExc_ValueError, "output buffer is too small");
+            return -1;
+        }
+
         int i;
-        for (i = 0; i < n; i++) {
-            if (i >= size) {
-                PyErr_SetString(PyExc_ValueError, "output buffer is too small");
-                return -1;
-            }
+        for (i = 0; i < n; i++)
             res[i] = _PyUnicode_ExtendedCase[index + i];
-        }
         return n;
     }
 
-    if (0 >= size) {
+    if (size < 1) {
         PyErr_SetString(PyExc_ValueError, "output buffer is too small");
         return -1;
     }
     res[0] = ch + ctype->title;
     return 1;
 }
 
-int PyUnicode_ToUpper(Py_UCS4 ch, Py_UCS4 *res, int size)
+Py_ssize_t PyUnicode_ToUpper(Py_UCS4 ch, Py_UCS4 *res, Py_ssize_t size)
 {
     const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
 
     if (ctype->flags & EXTENDED_CASE_MASK) {
         int index = ctype->upper & 0xFFFF;
         int n = ctype->upper >> 24;
+        if (n > size) {
+            PyErr_SetString(PyExc_ValueError, "output buffer is too small");
+            return -1;
+        }
+
         int i;
-        for (i = 0; i < n; i++) {
-            if (i >= size) {
-                PyErr_SetString(PyExc_ValueError, "output buffer is too small");
-                return -1;
-            }
+        for (i = 0; i < n; i++)
             res[i] = _PyUnicode_ExtendedCase[index + i];
-        }
         return n;
     }
 
-    if (0 >= size) {
+    if (size < 1) {
         PyErr_SetString(PyExc_ValueError, "output buffer is too small");
         return -1;
     }
     res[0] = ch + ctype->upper;
     return 1;
 }
 
-int PyUnicode_ToFolded(Py_UCS4 ch, Py_UCS4 *res, int size)
+Py_ssize_t PyUnicode_ToFolded(Py_UCS4 ch, Py_UCS4 *res, Py_ssize_t size)
 {
     const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
 
     if (ctype->flags & EXTENDED_CASE_MASK && (ctype->lower >> 20) & 7) {
         int index = (ctype->lower & 0xFFFF) + (ctype->lower >> 24);
         int n = (ctype->lower >> 20) & 7;
+        if (n > size) {
+            PyErr_SetString(PyExc_ValueError, "output buffer is too small");
+            return -1;
+        }
+
         int i;
-        for (i = 0; i < n; i++) {
-            if (i >= size) {
-                PyErr_SetString(PyExc_ValueError, "output buffer is too small");
-                return -1;
-            }
+        for (i = 0; i < n; i++)
             res[i] = _PyUnicode_ExtendedCase[index + i];
-        }
         return n;
     }