UTF8 decoding algorithm in VC

In order to save network bandwidth, many web applications use UTF8 encoding methods, such as MSN, etc., UTF8 encoding principle is actually very simple, here is a VC instance below Windows.

#define unkown_char '? # Define isbyteofonebytechar (b) (((b) & 0x80) == 0x00) #define isheadoftwobyteschar (b) (((b) & 0xe0) == 0xc0) #define isheadoftHreebyteschar (b) (( (b) & 0xF0) == 0xE0) #define isComingByteOfMultBytesChar (b) (((b) & 0xC0) == 0x80) #define isHeadOfUTF8Char (b) / isByteOfOneByteChar (b) || / isByteOfOneByteChar (b) || / isHeadOfThreeBytesChar (b) Data Data :: Utf8Decode (const data & utf8hex) {int L = uTF8HEX.LENGTH (); unsigned short * unicode = new unsigned short [l * 2]; int outi = 0; for (int i = 0; i L) {Unicode [outi ] = unkown_char; break;} byte b2 = (byte) UTF8HEX.GETCHAR (i 1); if (iScomingByteofmultByteschar (b2)) {unsigned short tmp = b1 & 0x1f ; TMP = (TMP << 6) (B2 & 0x3F); Unicode [outi ] = TMP; i = 2;} else IF (isheadofutf8char (b2)) {UNI Code [outi ] = unkown_char; i = 1;} else {unicode [outi ] = unkown_char; i = 2;}} else if (isheadoftofthreebyteschar (b1)) {if (i 2> L) {Unicode [OUTI ] = Unkown_char; Break;} else if (i 3> L) {byte b2 = (byte) UTF8HEX.GETCHAR (I 1); if (ISBYTEOFONEBYTECHAR (B2)) {unicode [outi ] = unkown_char; i = 1; Continue;} else {unicode [outi ] = unkown_char; break;

}} BYTE b2 = (BYTE) utf8hex.getChar (i 1); BYTE b3 = (BYTE) utf8hex.getChar (i 2); if (isComingByteOfMultBytesChar (b2) && isComingByteOfMultBytesChar (b3)) {unsigned short tmp = b1 & 0x0f; TMP = TMP << 12; unsigned short tmp2 = b2 & 0x3f; TMP2 = TMP2 << 6; Unicode [OUTI ] = TMP TMP2 (B3 & 0x3F); i = 3;} else f (isheadofutf8char (b2)) {unicode [outi ] = unkown_char; i = 1;} else if (isheadofutf8char (b3)) {unicode [outi ] = unkown_char; i = 2;} else {unicode [outi ] = unkown_char; i = 3;}} else {unicode [outi ] = unkown_char; i = 1;}} unicode [outi] = 0; Data Ret (outi * 4, true); # ifdef win32 int count = widechartomultibyte (0, 0 , Unicode, Outi, Ret.mbuf, Outi * 4, NULL, NULL); if (count == 0) CPLOG (log_err, "get utf byte, error number =% lu", getLastError ()); # else # Endif IF (count> 0) {RET.MBUF [count] = '/ 0'; ret.mlength = count;} delete [] unicode;

Return Ret;}