Added option --input-encoding <ascii | utf8> that allows to use UTF-8 literals in...

[re2c] / src / encoding / utf8 / utf8.h

diff --git a/src/encoding/utf8/utf8.h b/src/encoding/utf8/utf8.h
index efb0a309b75195a55bde7b5006868efadca6b6e1..1f593e2e41cc999f790e1f62890a6bef6d53be6b 100644 (file)
--- a/src/encoding/utf8/utf8.h
+++ b/src/encoding/utf8/utf8.h
@@ -29,6 +29,7 @@ public:
     static const uint32_t PREFIX_2BYTE;
     static const uint32_t PREFIX_3BYTE;
     static const uint32_t PREFIX_4BYTE;
+    static const uint32_t PREFIX_5BYTE;
 
     static const uint32_t SHIFT;
     static const uint32_t MASK;
@@ -36,6 +37,9 @@ public:
     // UTF-8 bytestring for given Unicode rune
     static uint32_t rune_to_bytes(uint32_t * s, rune r);
 
+    // read Unicode rune for the given (pre-validated) UTF-8 bytestring
+    static uint32_t decode_unsafe(const char *str);
+
     // length of UTF-8 bytestring for given Unicode rune
     static uint32_t rune_length(rune r);