Fix CJK emphasis delimiter detection in scanDelims

Treat CJK ideographs (codepoints > 0x2E7F) as punctuation in flanking
delimiter checks so that emphasis markers work correctly in mixed
CJK/ASCII contexts (e.g. 湾岸の**46%**を now renders bold correctly).

https://claude.ai/code/session_01RzF12DUuSS8R6Zw1NKwv9o

Author: claude

markdown_it/rules_inline/state_inline.py

@@ -148,6 +148,14 @@ def scanDelims(self, start: int, canSplitWord: bool) -> Scanned:
         isLastWhiteSpace = isWhiteSpace(ord(lastChar))
         isNextWhiteSpace = isWhiteSpace(ord(nextChar))
 
+        # Treat CJK ideographs as punctuation for flanking delimiter checks
+        # so that emphasis works correctly in mixed CJK/ASCII contexts.
+        # e.g. 湾岸の**46%**を should render <strong>46%</strong>
+        if not isNextPunctChar and ord(nextChar) > 0x2E7F:
+            isNextPunctChar = True
+        if not isLastPunctChar and ord(lastChar) > 0x2E7F:
+            isLastPunctChar = True
+
         left_flanking = not (
             isNextWhiteSpace
             or (isNextPunctChar and not (isLastWhiteSpace or isLastPunctChar))