utf8_encode.c

/*
 * ascii whitespace:
 * U+0009
 * U+000A
 * U+000C
 * U+000D
 * U+0020
*/
 /*
 * code point, four-to-six hex chars
 */
static int is_ascii_code_point(int val)
{
        /* ascii code point U+0000 to U+007F, inclusive
         * 0 to 127
         */
        return val >= 0x0000 && val <= 0x007F;
}

/*
 * Encode an integer to a utf8 unicode byte sequence
 * https://encoding.spec.whatwg.org/#code-point
 * https://encoding.spec.whatwg.org/#utf-8-encoder
 *
 * 0x80 == 128
 * x = 0x80
 * 0000 0000 1000 0000
 * x << 1
 *      left shift by one (multiply it by two)
 * 0000 0001 0000 0000
 *
 * 8
 * x = 8
 * 0000 0000 0000 1000
 * x << 5
 * 0000 0001 0000 0000
 * x = 256
 *        [0,255]
 * byte = [0x00, 0xFF]
 */
static int utf8encode(char* buf, int codepoint)
{
        int count;
        int offset;
        int i;
        int temp;

        if(is_ascii_code_point(codepoint)) {
                printf("ascii code point %d\n", codepoint);
                buf[0] = (char)codepoint;
                return 1;
        }

        if(codepoint >= 0x0080 && codepoint <= 0x07FF) {
                count = 1;
                offset = 0xC0;
        } else if(codepoint >= 0x0800 && codepoint <= 0xFFFF) {
                count = 2;
                offset = 0xE0;
        } else if(codepoint >= 0x10000 && codepoint <= 0x10FFFF) {
                count = 3;
                offset = 0xF0;
        }

        i = 0;
        buf[i] = (codepoint >> (6 * count)) + offset;
        i++;

        while(count > 0) {
                temp = codepoint >> (6 * (count - 1));
                buf[i++] = 0x80 | (temp & 0x3f);
                count--;
        }

        return i;
}