gh-89653: Use int type for Unicode kind (#92704)

vstinner · web-flow · commit f62ad4f2c421 · 2022-05-13T12:41:05.000+02:00
Use the same type that PyUnicode_FromKindAndData() kind parameter
type (public C API): int.
diff --git a/Include/cpython/unicodeobject.h b/Include/cpython/unicodeobject.h
@@ -507,7 +507,7 @@ PyAPI_FUNC(Py_UCS4) _PyUnicode_FindMaxChar (
 typedef struct {
     PyObject *buffer;
     void *data;
-    enum PyUnicode_Kind kind;
+    int kind;
     Py_UCS4 maxchar;
     Py_ssize_t size;
     Py_ssize_t pos;
@@ -566,7 +566,7 @@ _PyUnicodeWriter_PrepareInternal(_PyUnicodeWriter *writer,
    macro instead. */
 PyAPI_FUNC(int)
 _PyUnicodeWriter_PrepareKindInternal(_PyUnicodeWriter *writer,
-                                     enum PyUnicode_Kind kind);
+                                     int kind);
 
 /* Append a Unicode character.
    Return 0 on success, raise an exception and return -1 on error. */
diff --git a/Modules/_csv.c b/Modules/_csv.c
@@ -868,7 +868,7 @@ Reader_iternext(ReaderObj *self)
     PyObject *fields = NULL;
     Py_UCS4 c;
     Py_ssize_t pos, linelen;
-    unsigned int kind;
+    int kind;
     const void *data;
     PyObject *lineobj;
 
@@ -1066,7 +1066,7 @@ join_reset(WriterObj *self)
  * record length.
  */
 static Py_ssize_t
-join_append_data(WriterObj *self, unsigned int field_kind, const void *field_data,
+join_append_data(WriterObj *self, int field_kind, const void *field_data,
                  Py_ssize_t field_len, int *quoted,
                  int copy_phase)
 {
@@ -1179,7 +1179,7 @@ join_check_rec_size(WriterObj *self, Py_ssize_t rec_len)
 static int
 join_append(WriterObj *self, PyObject *field, int quoted)
 {
-    unsigned int field_kind = -1;
+    int field_kind = -1;
     const void *field_data = NULL;
     Py_ssize_t field_len = 0;
     Py_ssize_t rec_len;
@@ -1211,7 +1211,7 @@ static int
 join_append_lineterminator(WriterObj *self)
 {
     Py_ssize_t terminator_len, i;
-    unsigned int term_kind;
+    int term_kind;
     const void *term_data;
 
     terminator_len = PyUnicode_GET_LENGTH(self->dialect->lineterminator);
diff --git a/Modules/_datetimemodule.c b/Modules/_datetimemodule.c
@@ -5284,7 +5284,7 @@ _sanitize_isoformat_str(PyObject *dtstr)
     //
     // The result of this, if not NULL, returns a new reference
     const void* const unicode_data = PyUnicode_DATA(dtstr);
-    const unsigned int kind = PyUnicode_KIND(dtstr);
+    const int kind = PyUnicode_KIND(dtstr);
 
     // Depending on the format of the string, the separator can only ever be
     // in positions 7, 8 or 10. We'll check each of these for a surrogate and
diff --git a/Modules/_decimal/_decimal.c b/Modules/_decimal/_decimal.c
@@ -1918,7 +1918,7 @@ dec_dealloc(PyObject *dec)
 /******************************************************************************/
 
 Py_LOCAL_INLINE(int)
-is_space(enum PyUnicode_Kind kind, const void *data, Py_ssize_t pos)
+is_space(int kind, const void *data, Py_ssize_t pos)
 {
     Py_UCS4 ch = PyUnicode_READ(kind, data, pos);
     return Py_UNICODE_ISSPACE(ch);
@@ -1935,7 +1935,7 @@ is_space(enum PyUnicode_Kind kind, const void *data, Py_ssize_t pos)
 static char *
 numeric_as_ascii(PyObject *u, int strip_ws, int ignore_underscores)
 {
-    enum PyUnicode_Kind kind;
+    int kind;
     const void *data;
     Py_UCS4 ch;
     char *res, *cp;
diff --git a/Modules/_elementtree.c b/Modules/_elementtree.c
@@ -1121,7 +1121,7 @@ checkpath(PyObject* tag)
     if (PyUnicode_Check(tag)) {
         const Py_ssize_t len = PyUnicode_GET_LENGTH(tag);
         const void *data = PyUnicode_DATA(tag);
-        unsigned int kind = PyUnicode_KIND(tag);
+        int kind = PyUnicode_KIND(tag);
         if (len >= 3 && PyUnicode_READ(kind, data, 0) == '{' && (
                 PyUnicode_READ(kind, data, 1) == '}' || (
                 PyUnicode_READ(kind, data, 1) == '*' &&
diff --git a/Modules/_operator.c b/Modules/_operator.c
@@ -1230,9 +1230,6 @@ attrgetter_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
     /* prepare attr while checking args */
     for (idx = 0; idx < nattrs; ++idx) {
         PyObject *item = PyTuple_GET_ITEM(args, idx);
-        Py_ssize_t item_len;
-        const void *data;
-        unsigned int kind;
         int dot_count;
 
         if (!PyUnicode_Check(item)) {
@@ -1245,9 +1242,9 @@ attrgetter_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
             Py_DECREF(attr);
             return NULL;
         }
-        item_len = PyUnicode_GET_LENGTH(item);
-        kind = PyUnicode_KIND(item);
-        data = PyUnicode_DATA(item);
+        Py_ssize_t item_len = PyUnicode_GET_LENGTH(item);
+        int kind = PyUnicode_KIND(item);
+        const void *data = PyUnicode_DATA(item);
 
         /* check whether the string is dotted */
         dot_count = 0;
diff --git a/Modules/_pickle.c b/Modules/_pickle.c
@@ -2579,7 +2579,7 @@ raw_unicode_escape(PyObject *obj)
     char *p;
     Py_ssize_t i, size;
     const void *data;
-    unsigned int kind;
+    int kind;
     _PyBytesWriter writer;
 
     if (PyUnicode_READY(obj))
diff --git a/Modules/pyexpat.c b/Modules/pyexpat.c
@@ -1088,7 +1088,7 @@ PyUnknownEncodingHandler(void *encodingHandlerData,
     PyObject* u;
     int i;
     const void *data;
-    unsigned int kind;
+    int kind;
 
     if (PyErr_Occurred())
         return XML_STATUS_ERROR;
diff --git a/Objects/bytesobject.c b/Objects/bytesobject.c
@@ -2396,7 +2396,7 @@ _PyBytes_FromHex(PyObject *string, int use_bytearray)
 
     if (!PyUnicode_IS_ASCII(string)) {
         const void *data = PyUnicode_DATA(string);
-        unsigned int kind = PyUnicode_KIND(string);
+        int kind = PyUnicode_KIND(string);
         Py_ssize_t i;
 
         /* search for the first non-ASCII character */
diff --git a/Objects/longobject.c b/Objects/longobject.c
@@ -1714,7 +1714,7 @@ long_to_decimal_string_internal(PyObject *aa,
     digit *pout, *pin, rem, tenpow;
     int negative;
     int d;
-    enum PyUnicode_Kind kind;
+    int kind;
 
     a = (PyLongObject *)aa;
     if (a == NULL || !PyLong_Check(a)) {
@@ -1904,7 +1904,7 @@ long_format_binary(PyObject *aa, int base, int alternate,
     PyObject *v = NULL;
     Py_ssize_t sz;
     Py_ssize_t size_a;
-    enum PyUnicode_Kind kind;
+    int kind;
     int negative;
     int bits;
 
diff --git a/Objects/stringlib/localeutil.h b/Objects/stringlib/localeutil.h
@@ -75,7 +75,7 @@ InsertThousandsGrouping_fill(_PyUnicodeWriter *writer, Py_ssize_t *buffer_pos,
 
     if (n_zeros) {
         *buffer_pos -= n_zeros;
-        enum PyUnicode_Kind kind = PyUnicode_KIND(writer->buffer);
+        int kind = PyUnicode_KIND(writer->buffer);
         void *data = PyUnicode_DATA(writer->buffer);
         unicode_fill(kind, data, '0', *buffer_pos, n_zeros);
     }
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
diff --git a/Python/formatter_unicode.c b/Python/formatter_unicode.c
