Merge main: adopt icu_properties approach, drop unic-ucd-category fallback

Agent-Logs-Url: https://github.com/RustPython/RustPython/sessions/43f9f7dc-f5af-48b8-b93b-6363bcda334c

Co-authored-by: youknowone <69878+youknowone@users.noreply.github.com>

Cargo.lock

@@ -3739,6 +3739,7 @@ dependencies = [
  "num_enum",
  "optional",
  "rustpython-wtf8",
+ "unic-ucd-category",
 ]
 
 [[package]]

crates/sre_engine/src/string.rs

@@ -428,7 +428,6 @@ pub(crate) const fn is_uni_linebreak(ch: u32) -> bool {
 }
 #[inline]
 pub(crate) fn is_uni_alnum(ch: u32) -> bool {
-    // TODO: check with cpython
     char::try_from(ch).is_ok_and(|c| {
         GeneralCategoryGroup::Letter
             .union(GeneralCategoryGroup::Number)

extra_tests/snippets/builtin_str.py

@@ -891,3 +891,13 @@ assert id(b) != id(b * 0)
 assert id(b) != id(b * 1)
 assert id(b) != id(1 * b)
 assert id(b) != id(b * 2)
+
+
+# Regression tests for isalpha/isalnum Unicode General Category correctness.
+# These characters are in letter categories (Ll/Lo) and should return True,
+# but were missed in older Unicode tables used by unic-ucd-category.
+# See: https://github.com/RustPython/RustPython/pull/7520#issuecomment-4148322294
+for _cp in [1376, 1416, 1519, 2160, 2161, 2162, 2163, 2164, 2165, 2166]:
+    _c = chr(_cp)
+    assert _c.isalpha(), f"U+{_cp:04X} should be isalpha"
+    assert _c.isalnum(), f"U+{_cp:04X} should be isalnum"

extra_tests/snippets/builtin_str_unicode.py

@@ -11,6 +11,7 @@ c = ᚴ * 3
 
 assert c == "👋👋👋"
 
+import re
 import unicodedata
 
 assert unicodedata.category("a") == "Ll"
@@ -38,3 +39,10 @@ assert b"xn--pythn-mua.org.".decode("idna") == "pyth\xf6n.org."
 # TODO: add east_asian_width and mirrored
 # assert unicodedata.ucd_3_2_0.east_asian_width('\u231a') == 'N'
 # assert not unicodedata.ucd_3_2_0.mirrored("\u0f3a")
+
+# U+0345 COMBINING GREEK YPOGEGRAMMENI (category Mn) should not be alphanumeric.
+# CPython's isalpha/isalnum use Unicode letter categories (Lu/Ll/Lt/Lm/Lo),
+# not the broader Unicode Alphabetic derived property.
+assert not "\u0345".isalpha(), "isalpha should not match Mn category characters"
+assert not "\u0345".isalnum(), "isalnum should not match Mn category characters"
+assert not re.match(r"\w", "\u0345"), r"\w should not match U+0345 (category Mn)"