fix(melo-test): strip standalone Hangul jamo before synthesis

MeloTTS-Korean's symbol vocab only includes precomposed Hangul
syllables. Bare Hangul Compatibility Jamo characters (ㅋㅋㅋ, ㅠㅠ, ㅎㅎ)
and Hangul Jamo block code points blow up in
`cleaned_text_to_sequence` with `KeyError: 'ㅋ'` (or similar).

Pre-clean the request text in `/tts` to drop those code point ranges
before handing the string to the model, collapsing the leftover
whitespace. If sanitization yields an empty string, return 400 with
an explanation instead of a 500 stack trace.

Also document the limitation in the README troubleshooting section.

melo-test/server.py

@@ -8,6 +8,7 @@ GET  /             -> index.html (테스트 UI)
 """
 import os
 import io
+import re
 import time
 import tempfile
 import logging
@@ -62,16 +63,38 @@ class TTSReq(BaseModel):
     speed: float = 1.0
 
 
+# MeloTTS-Korean 의 symbol vocab 은 완성형 음절(가, 나, ...)만 포함한다.
+# 'ㅋㅋㅋ', 'ㅠㅠ', 'ㅎㅎ' 같이 한글 자모(Hangul Compatibility Jamo /
+# Hangul Jamo) 가 단독으로 들어오면 cleaned_text_to_sequence 에서
+# KeyError 로 죽는다. 합성 전에 미리 제거한다.
+_STANDALONE_JAMO_RE = re.compile(r"[\u1100-\u11FF\u3130-\u318F\uA960-\uA97F\uD7B0-\uD7FF]+")
+
+
+def _sanitize_for_melo(text: str) -> str:
+    # 단독 자모를 공백으로 치환한 뒤 연속 공백을 하나로 정리.
+    cleaned = _STANDALONE_JAMO_RE.sub(" ", text)
+    return re.sub(r"\s+", " ", cleaned).strip()
+
+
 @app.post("/tts")
 def tts(req: TTSReq):
     model = state.get("model")
     if model is None:
         raise HTTPException(503, "model not ready")
-    text = (req.text or "").strip()
-    if not text:
+    raw = (req.text or "").strip()
+    if not raw:
         raise HTTPException(400, "empty text")
-    if len(text) > 500:
+    if len(raw) > 500:
         raise HTTPException(400, "too long (<=500)")
+    text = _sanitize_for_melo(raw)
+    if not text:
+        raise HTTPException(
+            400,
+            "text contains only standalone Hangul jamo (e.g. ㅋㅋㅋ); "
+            "MeloTTS-Korean cannot synthesize these.",
+        )
+    if text != raw:
+        log.info("sanitized standalone jamo: %r -> %r", raw[:60], text[:60])
 
     speed = max(0.5, min(2.0, float(req.speed or 1.0)))
     fd, out_path = tempfile.mkstemp(suffix=".wav")