From 3e505885258c27393e39df1ba2215af49065b382 Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Tue, 25 Mar 2025 16:52:01 +0530
Subject: [PATCH] Add a test for PUA recog

---
 gen/wcwidth.py           |  2 +-
 kitty/char-props.h       | 11 +++++++++--
 kitty/data-types.c       | 14 ++++++++++++++
 kitty_tests/datatypes.py |  2 ++
 4 files changed, 26 insertions(+), 3 deletions(-)

diff --git a/gen/wcwidth.py b/gen/wcwidth.py
index 9a4d584a5..7deb463c1 100755
--- a/gen/wcwidth.py
+++ b/gen/wcwidth.py
@@ -717,7 +717,7 @@ def aw(s: Iterable[int], width: int) -> None:
         raw = f.read()
         nraw = re.sub(r'\d+/\*=width_shift\*/', f'{width_shift}/*=width_shift*/', raw)
         nraw = re.sub(r'// CharPropsDeclaration.+?// EndCharPropsDeclaration', CharProps.c_declaration(), nraw, flags=re.DOTALL)
-        nraw = re.sub(r'// UCBDeclaration.+?// EndUCBDeclaration', buf.getvalue(), nraw, flags=re.DOTALL)
+        nraw = re.sub(r'// UCBDeclaration.+?// EndUCBDeclaration', buf.getvalue().rstrip(), nraw, flags=re.DOTALL)
         if nraw != raw:
             f.seek(0)
             f.truncate()
diff --git a/kitty/char-props.h b/kitty/char-props.h
index 494fa635a..095d7668d 100644
--- a/kitty/char-props.h
+++ b/kitty/char-props.h
@@ -46,8 +46,6 @@ typedef enum UnicodeCategory {
 // EndUCBDeclaration
 
 
-
-
 // CharPropsDeclaration
 // Uses 23 bits
 typedef union CharProps {
@@ -101,3 +99,12 @@ void grapheme_segmentation_reset(GraphemeSegmentationState *s);
 bool grapheme_segmentation_step(GraphemeSegmentationState *s, CharProps ch);
 static inline int wcwidth_std(CharProps ch) { return (int)ch.shifted_width - 4/*=width_shift*/; }
 static inline bool is_private_use(CharProps ch) { return ch.category == UC_Co; }
+static inline const char* char_category(CharProps cp) {
+#define a(x) case UC_##x: return #x
+    switch((UnicodeCategory)cp.category) {
+        a(Cn); a(Cc); a(Zs); a(Po); a(Sc); a(Ps); a(Pe); a(Sm); a(Pd); a(Nd); a(Lu); a(Sk); a(Pc); a(Ll); a(So); a(Lo); a(Pi); a(Cf);
+        a(No); a(Pf); a(Lt); a(Lm); a(Mn); a(Me); a(Mc); a(Nl); a(Zl); a(Zp); a(Cs); a(Co);
+    }
+    return "Cn";
+#undef a
+}
diff --git a/kitty/data-types.c b/kitty/data-types.c
index 931d300d1..8b792bd11 100644
--- a/kitty/data-types.c
+++ b/kitty/data-types.c
@@ -632,6 +632,19 @@ py_run_atexit_cleanup_functions(PyObject *self UNUSED, PyObject *args UNUSED) {
     Py_RETURN_NONE;
 }
 
+static PyObject*
+py_char_props_for(PyObject *self UNUSED, PyObject *ch) {
+    if (!PyUnicode_Check(ch) || PyUnicode_GET_LENGTH(ch) != 1) { PyErr_SetString(PyExc_TypeError, "must suply a single character"); return NULL; }
+    char_type c = PyUnicode_READ_CHAR(ch, 0);
+    CharProps cp = char_props_for(c);
+#define B(x) #x, cp.x ? Py_True : Py_False
+    return Py_BuildValue("{ si sO sB sB ss }",
+        "width", wcwidth_std(cp), B(is_extended_pictographic), "grapheme_break", cp.grapheme_break,
+        "indic_conjunct_break", cp.indic_conjunct_break, "category", char_category(cp)
+    );
+#undef B
+}
+
 static PyMethodDef module_methods[] = {
     METHODB(replace_c0_codes_except_nl_space_tab, METH_O),
     {"wcwidth", (PyCFunction)wcwidth_wrap, METH_O, ""},
@@ -647,6 +660,7 @@ static PyMethodDef module_methods[] = {
     {"base64_encode_into", (PyCFunction)base64_encode_into, METH_VARARGS, ""},
     {"base64_decode", (PyCFunction)(void (*) (void))(pybase64_decode), METH_O, ""},
     {"base64_decode_into", (PyCFunction)base64_decode_into, METH_VARARGS, ""},
+    {"char_props_for", py_char_props_for, METH_O, ""},
     {"split_into_graphemes", (PyCFunction)split_into_graphemes, METH_O, ""},
     {"thread_write", (PyCFunction)cm_thread_write, METH_VARARGS, ""},
     {"redirect_std_streams", (PyCFunction)redirect_std_streams, METH_VARARGS, ""},
diff --git a/kitty_tests/datatypes.py b/kitty_tests/datatypes.py
index ab5033318..88700f3cc 100644
--- a/kitty_tests/datatypes.py
+++ b/kitty_tests/datatypes.py
@@ -11,6 +11,7 @@
     Color,
     HistoryBuf,
     LineBuf,
+    char_props_for,
     expand_ansi_c_escapes,
     parse_input_from_terminal,
     replace_c0_codes_except_nl_space_tab,
@@ -640,3 +641,4 @@ def test_split_into_graphemes(self):
             expected = test['data']
             actual = split_into_graphemes(''.join(expected))
             self.ae(expected, actual, f'Test #{i} failed: {test["comment"]}')
+        self.assertEqual(char_props_for('\ue000')['category'], 'Co')