utf-8 masks depend on endianness

Rather than reversing 32 bit numbers, when checking if the analysed code
units represent a valid utf-8 encoding, we change the constants base on
system's endianness.

Author: grisumbras

include/boost/json/detail/utf8.hpp

@@ -16,6 +16,12 @@
 #include <cstring>
 #include <cstdint>
 
+#ifdef BOOST_JSON_BIG_ENDIAN
+# define BOOST_JSON_MK_NUM4(b4, b3, b2, b1) 0x ## b1 ## b2 ## b3 ## b4
+#else
+# define BOOST_JSON_MK_NUM4(b4, b3, b2, b1) 0x ## b4 ## b3 ## b2 ## b1
+#endif
+
 namespace boost {
 namespace json {
 namespace detail {
@@ -74,46 +80,55 @@ inline
 bool
 is_valid_utf8(const char* p, uint16_t first)
 {
-    uint32_t v;
+    std::uint32_t v = 0;
     switch(first >> 8)
     {
     default:
         return false;
 
     // 2 bytes, second byte [80, BF]
     case 1:
-        v = load_little_endian<2>(p);
-        return (v & 0xC000) == 0x8000;
+        std::memcpy(&v, p, 2);
+        return ( v & BOOST_JSON_MK_NUM4(00,00,C0,00) )
+            == BOOST_JSON_MK_NUM4(00,00,80,00);
 
     // 3 bytes, second byte [A0, BF]
     case 2:
-        v = load_little_endian<3>(p);
-        return (v & 0xC0E000) == 0x80A000;
+        std::memcpy(&v, p, 3);
+        return ( v & BOOST_JSON_MK_NUM4(00,C0,E0,00) )
+            == BOOST_JSON_MK_NUM4(00,80,A0,00);
 
     // 3 bytes, second byte [80, BF]
     case 3:
-        v = load_little_endian<3>(p);
-        return (v & 0xC0C000) == 0x808000;
+        std::memcpy(&v, p, 3);
+        return ( v & BOOST_JSON_MK_NUM4(00,C0,C0,00) )
+            == BOOST_JSON_MK_NUM4(00,80,80,00);
 
     // 3 bytes, second byte [80, 9F]
     case 4:
-        v = load_little_endian<3>(p);
-        return (v & 0xC0E000) == 0x808000;
+        std::memcpy(&v, p, 3);
+        return ( v & BOOST_JSON_MK_NUM4(00,C0,E0,00) )
+            == BOOST_JSON_MK_NUM4(00,80,80,00);
 
     // 4 bytes, second byte [90, BF]
     case 5:
-        v = load_little_endian<4>(p);
-        return (v & 0xC0C0FF00) + 0x7F7F7000 <= 0x2F00;
+        std::memcpy(&v, p, 4);
+        return (
+            ( v & BOOST_JSON_MK_NUM4(C0, C0, C0, 00) )
+                == BOOST_JSON_MK_NUM4(80, 80, 80, 00) )
+            &&   ( v & BOOST_JSON_MK_NUM4(00, 00, 30, 00) );
 
     // 4 bytes, second byte [80, BF]
     case 6:
-        v = load_little_endian<4>(p);
-        return (v & 0xC0C0C000) == 0x80808000;
+        std::memcpy(&v, p, 4);
+        return ( v & BOOST_JSON_MK_NUM4(C0,C0,C0,00) )
+            == BOOST_JSON_MK_NUM4(80,80,80,00);
 
     // 4 bytes, second byte [80, 8F]
     case 7:
-        v = load_little_endian<4>(p);
-        return (v & 0xC0C0F000) == 0x80808000;
+        std::memcpy(&v, p, 4);
+        return ( v & BOOST_JSON_MK_NUM4(C0,C0,F0,00) )
+            == BOOST_JSON_MK_NUM4(80,80,80,00);
     }
 }