2 files changed, 24 insertions, 0 deletions
diff --git a/src/UniConversion.cxx b/src/UniConversion.cxx
index 4da9e102a..d0028d65e 100644
--- a/src/UniConversion.cxx
+++ b/src/UniConversion.cxx
@@ -8,6 +8,7 @@
 #include <stdlib.h>
 
 #include <stdexcept>
+#include <string>
 
 #include "UniConversion.h"
 
@@ -304,6 +305,28 @@ int UTF8DrawBytes(const unsigned char *us, int len) {
 	return (utf8StatusNext & UTF8MaskInvalid) ? 1 : (utf8StatusNext & UTF8MaskWidth);
 }
 
+// Replace invalid bytes in UTF-8 with the replacement character
+std::string FixInvalidUTF8(const std::string &text) {
+	std::string result;
+	const unsigned char *us = reinterpret_cast<const unsigned char *>(text.c_str());
+	size_t remaining = text.size();
+	while (remaining > 0) {
+		const int utf8Status = UTF8Classify(us, static_cast<int>(remaining));
+		if (utf8Status & UTF8MaskInvalid) {
+			// Replacement character 0xFFFD = UTF8:"efbfbd".
+			result.append("\xef\xbf\xbd");
+			us++;
+			remaining--;
+		} else {
+			const int len = utf8Status&UTF8MaskWidth;
+			result.append(reinterpret_cast<const char *>(us), len);
+			us += len;
+			remaining -= len;
+		}
+	}
+	return result;
+}
+
 #ifdef SCI_NAMESPACE
 }
 #endif
diff --git a/src/UniConversion.h b/src/UniConversion.h
index aeb13f0c2..c5867a664 100644
--- a/src/UniConversion.h
+++ b/src/UniConversion.h
@@ -23,6 +23,7 @@ size_t UTF16Length(const char *s, size_t len);
 size_t UTF16FromUTF8(const char *s, size_t len, wchar_t *tbuf, size_t tlen);
 unsigned int UTF32FromUTF8(const char *s, unsigned int len, unsigned int *tbuf, unsigned int tlen);
 unsigned int UTF16FromUTF32Character(unsigned int val, wchar_t *tbuf);
+std::string FixInvalidUTF8(const std::string &text);
 
 extern int UTF8BytesOfLead[256];
 void UTF8BytesOfLeadInitialise();