UTF-8 to UTF-16

The C function below properly converts UTF-8 characters to UTF-16, for use in Windows functions, etc. The example also shows how to retrieve individual characters from a UTF-16 string. If you don't want to include windows.h, use this header:

typedef unsigned short WCHAR;
typedef signed long LONG;
typedef unsigned char BYTE;
typedef unsigned short WORD;
typedef unsigned long DWORD;
typedef int BOOL;
#ifndef NULL
#define NULL ((void*)0)
#endif
#ifndef TRUE
#define TRUE 1
#endif
#ifndef FALSE
#define FALSE 0
#endif

The code follows. I release it to the public domain.

typedef struct {
 BYTE byte1f,byte1t,byte2f,byte2t,
      valueMask,trail;
} UTF8TBLROW;


static const UTF8TBLROW utf8tbl[] = {
 {0x00,0x7F,0   ,0   ,0x7F,0},
 {0xC2,0xDF,0x80,0xBF,0x1F,1},
 {0xE0,0xE0,0xA0,0xBF,0x0F,2},
 {0xE1,0xEC,0x80,0xBF,0x0F,2},
 {0xED,0xED,0x80,0x9F,0x0F,2},
 {0xEE,0xEF,0x80,0xBF,0x0F,2},
 {0xF0,0xF0,0x90,0xBF,0x07,3},
 {0xF1,0xF3,0x80,0xBF,0x07,3},
 {0xF4,0xF4,0x80,0x8F,0x07,3},
 {0   ,0   ,0   ,0   ,0   ,0}
};


WCHAR *Utf8ToUtf16(char *string){
 if(!string)return NULL;
 DWORD bytes=(strlen(string)*2+1)*sizeof(WCHAR);
 WCHAR *result=malloc(bytes);
 WCHAR *presult=result;
 while(1){
  BYTE b=*string++;
  if(b==0){//null terminator
   *(presult++)=0;
   break;
  } else if(b<0x80){//ASCII character
   *(presult++)=b;
  } else {
   DWORD ret=0;
   UTF8TBLROW *pRow=&utf8tbl[0];
   BOOL bad=FALSE;
   //b is the byte retrieved at beginning of loop
   while(pRow-mym>valueMask&&((b<pRow-mym>byte1f)||(b>pRow->byte1t))){
    pRow++;
   }
   ret=(DWORD)(b&pRow->valueMask);
   if(pRow->trail==0){
    //check for 0 trailing bytes already made above
    bad=TRUE;
   } else if(pRow->trail==1){
    b=*string++; // second byte
    if(b<pRow-mym>byte2f||b>pRow->byte2t)bad=TRUE;
    ret=(ret<<6)|(b&0x3F);
   } else if(pRow->trail==2){
    b=*string++; // second byte
    if(b<pRow-mym>byte2f||b>pRow->byte2t)bad=TRUE;
    ret=(ret<<6)|(b&0x3F);
    b=*string++; // third byte
    if(b<0x80||bmym>0xBF)bad=TRUE;
    ret=(ret<<6)|(b&0x3F);
   } else if(pRow->trail==3){
    b=*string++; // second byte
    if(b<pRow-mym>byte2f||b>pRow->byte2t)bad=TRUE;
    ret=(ret<<6)|(b&0x3F);
    b=*string++; // third byte
    if(b<0x80||bmym>0xBF)bad=TRUE;
    ret=(ret<<6)|(b&0x3F);
    b=*string++; // fourth byte
    if(b<0x80||bmym>0xBF)bad=TRUE;
    ret=(ret<<6)|(b&0x3F);
   }
   if(bad){
    free(result);
    return NULL;
   } else if((ret&0xFFFFF800)==0xD800){
    free(result);
    return NULL;
   } else if((ret&0xFFFF0000)==0){ // BMP character
    *(presult++)=ret;
   } else { // SMP character
    ret-=0x10000;
    *(presult++)=((ret>>10)&0x03FF)|0xD800;
    *(presult++)=((ret    )&0x03FF)|0xDC00;
   }
  }
 }
 return result;
}