Fix UTF-8 validation checks

The bitwise comparisons are only for 8-bit characters, not 32-bit
characters.

The checks were mixing up binary and hex.  :-)

diff --git a/lily/text-interface.cc b/lily/text-interface.cc
index c7d425bcb5909f6afe205efa2db46ae97007582e..8f2f7edc062d2f2a8fb31f2720c38fea0d13d5a8 100644 (file)
--- a/lily/text-interface.cc
+++ b/lily/text-interface.cc
@@ -31,15 +31,15 @@ replace_whitespace (string *str)
       char cur = (*str)[i];
       
       // U+10000 - U+10FFFF
-      if ((cur & 0x11110000) == 0x11110000)
+      if ((cur & 0xF0) == 0xF0)
        char_len = 4;
       // U+0800 - U+FFFF
-      else if ((cur & 0x11100000) == 0x11100000)
+      else if ((cur & 0xE0) == 0xE0)
        char_len = 3;
       // U+0080 - U+07FF
-      else if ((cur & 0x11000000) == 0x11000000)
+      else if ((cur & 0xC0) == 0xC0)
        char_len = 2;
-      else if (cur & 0x10000000)
+      else if (cur & 0x80)
        programming_error ("invalid utf-8 string");
       else
        // avoid the locale-dependent isspace