Function utf8StrWidth

linenoise.c:441–478

Calculate the display width of a UTF-8 string of 'len' bytes. * This is used for cursor positioning in the terminal. * Handles grapheme clusters: characters joined by ZWJ contribute 0 width * after the first character in the sequence. * ANSI CSI escape sequences (e.g. color codes in the prompt) are treated * as zero-width. */

         (cp >= 0xFFE0 && cp <= 0xFFE6) ||    /* Fullwidth Signs */
         (cp >= 0x1F1E6 && cp <= 0x1F1FF) ||  /* Regional Indicators (flags) */
         (cp >= 0x1F300 && cp <= 0x1F64F) ||  /* Misc Symbols and Emoticons */
         (cp >= 0x1F680 && cp <= 0x1F6FF) ||  /* Transport and Map Symbols */
         (cp >= 0x1F900 && cp <= 0x1F9FF) ||  /* Supplemental Symbols */
         (cp >= 0x1FA00 && cp <= 0x1FAFF) ||  /* Chess, Extended-A */
         (cp >= 0x20000 && cp <= 0x2FFFF)))   /* CJK Extension B and beyond */
        return 2;

    return 1; /* Default: single width */
}

int linenoiseCharacterWidth(uint32_t cp) {
    return utf8CharWidth(cp);
}

size_t linenoiseNextGrapheme(const char *s, size_t len, int *width) {
    uint32_t cp;
    size_t used = linenoiseUtf8Decode(s, len, &cp);
    if (!used) { if (width) *width = 0; return 0; }
    int w = utf8CharWidth(cp), regional = isRegionalIndicator(cp), joined = 0;
    while (used < len) {
        size_t n = linenoiseUtf8Decode(s + used, len - used, &cp);
        if (!n) break;
        if (joined) {
            if (utf8CharWidth(cp) > w) w = utf8CharWidth(cp);
            joined = 0;
        } else if (isZWJ(cp)) {
            joined = 1;
        } else if (isGraphemeExtend(cp)) {
            if (cp == 0xfe0f && w) w = 2;
        } else if (regional && isRegionalIndicator(cp)) {
            regional = 0;
        } else break;
        used += n;
    }
    if (width) *width = w;
    return used;
}

/* If s[] points at an ANSI CSI escape sequence (e.g. a color change like
 * ESC [ 1 ; 32 m), return its length in bytes. Otherwise return 0.
 *