gh-129117: Expose _PyUnicode_IsXidContinue/Start in unicodedata

Doc/library/unicodedata.rst

@@ -144,6 +144,36 @@ following functions:
  1
 
 
+.. function:: isxidstart(chr, /)
+
+ Return ``True`` if *chr* is a valid identifier start per the
+ `Unicode Standard Annex #31 <https://www.unicode.org/reports/tr31/>`_,
+ that is, it has the ``XID_Start`` property. Return ``False`` otherwise.
+ For example::
+
+ >>> unicodedata.isxidstart('S')
+ True
+ >>> unicodedata.isxidstart('0')
+ False
+
+ .. versionadded:: next
+
+
+.. function:: isxidcontinue(chr, /)
+
+ Return ``True`` if *chr* is a valid identifier character per the
+ `Unicode Standard Annex #31 <https://www.unicode.org/reports/tr31/>`_,
+ that is, it has the ``XID_Continue`` property. Return ``False`` otherwise.
+ For example::
+
+ >>> unicodedata.isxidcontinue('S')
+ True
+ >>> unicodedata.isxidcontinue(' ')
+ False
+
+ .. versionadded:: next
+
+
 .. function:: decomposition(chr)
 
  Returns the character decomposition mapping assigned to the character

Doc/whatsnew/3.15.rst

@@ -794,6 +794,11 @@ unicodedata
 
 * The Unicode database has been updated to Unicode 17.0.0.
 
+* Add :func:`unicodedata.isxidstart` and :func:`unicodedata.isxidcontinue`
+ functions to check whether a character can start or continue a
+ `Unicode Standard Annex #31 <https://www.unicode.org/reports/tr31/>`_ identifier.
+ (Contributed by Stan Ulbrych in :gh:`129117`.)
+
 
 wave
 ----

Include/internal/pycore_unicodectype.h

@@ -0,0 +1,25 @@
+#ifndef Py_INTERNAL_UNICODECTYPE_H
+#define Py_INTERNAL_UNICODECTYPE_H
+#ifdef __cplusplus
+extern "C"{
+#endif
+
+#ifndef Py_BUILD_CORE
+# error "this header requires Py_BUILD_CORE define"
+#endif
+
+extern int _PyUnicode_ToLowerFull(Py_UCS4 ch, Py_UCS4 *res);
+extern int _PyUnicode_ToTitleFull(Py_UCS4 ch, Py_UCS4 *res);
+extern int _PyUnicode_ToUpperFull(Py_UCS4 ch, Py_UCS4 *res);
+extern int _PyUnicode_ToFoldedFull(Py_UCS4 ch, Py_UCS4 *res);
+extern int _PyUnicode_IsCaseIgnorable(Py_UCS4 ch);
+extern int _PyUnicode_IsCased(Py_UCS4 ch);
+
+// Export for 'unicodedata' shared extension.
+PyAPI_FUNC(int) _PyUnicode_IsXidStart(Py_UCS4 ch);
+PyAPI_FUNC(int) _PyUnicode_IsXidContinue(Py_UCS4 ch);
+
+#ifdef __cplusplus
+}
+#endif
+#endif /* !Py_INTERNAL_UNICODECTYPE_H */

Include/internal/pycore_unicodeobject.h

@@ -74,18 +74,6 @@ _PyUnicodeWriter_WriteCharInline(_PyUnicodeWriter *writer, Py_UCS4 ch)
  return 0;
 }
 
-
-/* --- Characters Type APIs ----------------------------------------------- */
-
-extern int _PyUnicode_IsXidStart(Py_UCS4 ch);
-extern int _PyUnicode_IsXidContinue(Py_UCS4 ch);
-extern int _PyUnicode_ToLowerFull(Py_UCS4 ch, Py_UCS4 *res);
-extern int _PyUnicode_ToTitleFull(Py_UCS4 ch, Py_UCS4 *res);
-extern int _PyUnicode_ToUpperFull(Py_UCS4 ch, Py_UCS4 *res);
-extern int _PyUnicode_ToFoldedFull(Py_UCS4 ch, Py_UCS4 *res);
-extern int _PyUnicode_IsCaseIgnorable(Py_UCS4 ch);
-extern int _PyUnicode_IsCased(Py_UCS4 ch);
-
 /* --- Unicode API -------------------------------------------------------- */
 
 // Export for '_json' shared extension

Lib/test/test_unicodedata.py

@@ -276,6 +276,33 @@ def test_east_asian_width_9_0_changes(self):
  self.assertEqual(self.db.ucd_3_2_0.east_asian_width('\u231a'), 'N')
  self.assertEqual(self.db.east_asian_width('\u231a'), 'W')
 
+ def test_isxidstart(self):
+ self.assertTrue(self.db.isxidstart('S'))
+ self.assertTrue(self.db.isxidstart('\u0AD0')) # GUJARATI OM
+ self.assertTrue(self.db.isxidstart('\u0EC6')) # LAO KO LA
+ self.assertTrue(self.db.isxidstart('\u17DC')) # KHMER SIGN AVAKRAHASANYA
+ self.assertTrue(self.db.isxidstart('\uA015')) # YI SYLLABLE WU
+ self.assertTrue(self.db.isxidstart('\uFE7B')) # ARABIC KASRA MEDIAL FORM
+
+ self.assertFalse(self.db.isxidstart(' '))
+ self.assertFalse(self.db.isxidstart('0'))
+ self.assertRaises(TypeError, self.db.isxidstart)
+ self.assertRaises(TypeError, self.db.isxidstart, 'xx')
+
+ def test_isxidcontinue(self):
+ self.assertTrue(self.db.isxidcontinue('S'))
+ self.assertTrue(self.db.isxidcontinue('_'))
+ self.assertTrue(self.db.isxidcontinue('0'))
+ self.assertTrue(self.db.isxidcontinue('\u00BA')) # MASCULINE ORDINAL INDICATOR
+ self.assertTrue(self.db.isxidcontinue('\u0640')) # ARABIC TATWEEL
+ self.assertTrue(self.db.isxidcontinue('\u0710')) # SYRIAC LETTER ALAPH
+ self.assertTrue(self.db.isxidcontinue('\u0B3E')) # ORIYA VOWEL SIGN AA
+ self.assertTrue(self.db.isxidcontinue('\u17D7')) # KHMER SIGN LEK TOO
+
+ self.assertFalse(self.db.isxidcontinue(' '))
+ self.assertRaises(TypeError, self.db.isxidcontinue)
+ self.assertRaises(TypeError, self.db.isxidcontinue, 'xx')
+
 class UnicodeMiscTest(UnicodeDatabaseTest):
 
  @cpython_only

Makefile.pre.in

@@ -1433,6 +1433,7 @@ PYTHON_HEADERS= \
  $(srcdir)/Include/internal/pycore_typeobject.h \
  $(srcdir)/Include/internal/pycore_typevarobject.h \
  $(srcdir)/Include/internal/pycore_ucnhash.h \
+ $(srcdir)/Include/internal/pycore_unicodectype.h \
  $(srcdir)/Include/internal/pycore_unicodeobject.h \
  $(srcdir)/Include/internal/pycore_unicodeobject_generated.h \
  $(srcdir)/Include/internal/pycore_unionobject.h \

Misc/NEWS.d/next/Library/2025-10-17-20-42-38.gh-issue-129117.X9jr4p.rst

@@ -0,0 +1,3 @@
+:mod:`unicodedata`: Add :func:`~unicodedata.isxidstart` and
+:func:`~unicodedata.isxidcontinue` functions to check whether a character can
+start or continue a `Unicode Standard Annex #31 <https://www.unicode.org/reports/tr31/>`_ identifier.

Modules/unicodedata.c

@@ -19,6 +19,7 @@
 #include "Python.h"
 #include "pycore_object.h" // _PyObject_VisitType()
 #include "pycore_ucnhash.h" // _PyUnicode_Name_CAPI
+#include "pycore_unicodectype.h" // _PyUnicode_IsXidStart()
 
 #include <stdbool.h>
 #include <stddef.h> // offsetof()
@@ -1525,6 +1526,58 @@ unicodedata_UCD_name_impl(PyObject *self, int chr, PyObject *default_value)
  return PyUnicode_FromString(name);
 }
 
+/*[clinic input]
+unicodedata.UCD.isxidstart
+
+ self: self
+ chr: int(accept={str})
+ /
+
+Return True if the character has the XID_Start property, else False.
+
+[clinic start generated code]*/
+
+static PyObject *
+unicodedata_UCD_isxidstart_impl(PyObject *self, int chr)
+/*[clinic end generated code: output=944005823c72c3ef input=9353f88d709c21fb]*/
+{
+ if (UCD_Check(self)){
+ const change_record *old = get_old_record(self, chr);
+ if (old->category_changed == 0){
+ /* unassigned */
+ Py_RETURN_FALSE;
+ }
+ }
+
+ return PyBool_FromLong(_PyUnicode_IsXidStart(chr));
+}
+
+/*[clinic input]
+unicodedata.UCD.isxidcontinue
+
+ self: self
+ chr: int(accept={str})
+ /
+
+Return True if the character has the XID_Continue property, else False.
+
+[clinic start generated code]*/
+
+static PyObject *
+unicodedata_UCD_isxidcontinue_impl(PyObject *self, int chr)
+/*[clinic end generated code: output=9438dcbff5ca3e41 input=bbb8dd3ac0d2d709]*/
+{
+ if (UCD_Check(self)){
+ const change_record *old = get_old_record(self, chr);
+ if (old->category_changed == 0){
+ /* unassigned */
+ Py_RETURN_FALSE;
+ }
+ }
+
+ return PyBool_FromLong(_PyUnicode_IsXidContinue(chr));
+}
+
 /*[clinic input]
 unicodedata.UCD.lookup
 
@@ -1590,6 +1643,8 @@ static PyMethodDef unicodedata_functions[] ={
  UNICODEDATA_UCD_EAST_ASIAN_WIDTH_METHODDEF
  UNICODEDATA_UCD_DECOMPOSITION_METHODDEF
  UNICODEDATA_UCD_NAME_METHODDEF
+ UNICODEDATA_UCD_ISXIDSTART_METHODDEF
+ UNICODEDATA_UCD_ISXIDCONTINUE_METHODDEF
  UNICODEDATA_UCD_LOOKUP_METHODDEF
  UNICODEDATA_UCD_IS_NORMALIZED_METHODDEF
  UNICODEDATA_UCD_NORMALIZE_METHODDEF

Objects/unicodectype.c

@@ -9,6 +9,7 @@
 */
 
 #include "Python.h"
+#include "pycore_unicodectype.h" // export _PyUnicode_IsXidStart(), _PyUnicode_IsXidContinue()
 
 #define ALPHA_MASK 0x01
 #define DECIMAL_MASK 0x02

Objects/unicodeobject.c

@@ -57,6 +57,7 @@ OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 #include "pycore_pylifecycle.h" // _Py_SetFileSystemEncoding()
 #include "pycore_pystate.h" // _PyInterpreterState_GET()
 #include "pycore_ucnhash.h" // _PyUnicode_Name_CAPI
+#include "pycore_unicodectype.h" // _PyUnicode_IsXidStart
 #include "pycore_unicodeobject.h" // struct _Py_unicode_state
 #include "pycore_unicodeobject_generated.h" // _PyUnicode_InitStaticStrings()
 

PCbuild/pythoncore.vcxproj

@@ -328,6 +328,7 @@
  <ClInclude Include="..\Include\internal\pycore_typevarobject.h" />
  <ClInclude Include="..\Include\internal\pycore_ucnhash.h" />
  <ClInclude Include="..\Include\internal\pycore_unionobject.h" />
+ <ClInclude Include="..\Include\internal\pycore_unicodectype.h" />
  <ClInclude Include="..\Include\internal\pycore_unicodeobject.h" />
  <ClInclude Include="..\Include\internal\pycore_unicodeobject_generated.h" />
  <ClInclude Include="..\Include\internal\pycore_uniqueid.h" />

PCbuild/pythoncore.vcxproj.filters

@@ -528,6 +528,9 @@
  <ClInclude Include="..\Include\cpython\initconfig.h">
  <Filter>Include\cpython</Filter>
  </ClInclude>
+ <ClInclude Include="..\Include\internal\pycore_unicodectype.h">
+ <Filter>Include\internal</Filter>
+ </ClInclude>
  <ClInclude Include="..\Include\internal\pycore_unicodeobject.h">
  <Filter>Include\internal</Filter>
  </ClInclude>