From 97e89c408a08414779433cc1a649ef6c5a2a3e86 Mon Sep 17 00:00:00 2001
From: Jean-Francois Dockes <jf@dockes.org>
Date: Wed, 25 Mar 2020 16:57:42 +0100
Subject: [PATCH] korean splitter: only break korean stretch on non-korean
 alphabetic (e.g. not numbers or punctuation)

---
 src/common/textsplitko.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/common/textsplitko.cpp b/src/common/textsplitko.cpp
index f3257b6d..d0c1bdc0 100644
--- a/src/common/textsplitko.cpp
+++ b/src/common/textsplitko.cpp
@@ -124,9 +124,9 @@ bool TextSplit::ko_to_words(Utf8Iter *itp, unsigned int *cp)
     string::size_type orgbytepos = it.getBpos();
     for (; !it.eof(); it++) {
         c = *it;
-        if (!isHANGUL(c) && !(isspace(c) || ispunct(c))) {
+        if (!isHANGUL(c) && isalpha(c)) {
             // Done with Korean stretch, process and go back to main routine
-            //std::cerr << "Broke on char " << int(c) << endl;
+            std::cerr << "Broke on char " << (std::string)it << endl;
             break;
         } else {
             it.appendchartostring(inputdata);