#ifdef USE_UNICODE

#include "wc.h"
#include "ucs.h"
#include "utf8.h"
#include "wtf.h"

wc_uint8 WC_UTF8_MAP[ 0x100 ] = {
   8, 8, 8, 8, 8, 8, 8, 8,  8, 8, 8, 8, 8, 8, 8, 8,
   8, 8, 8, 8, 8, 8, 8, 8,  8, 8, 8, 8, 8, 8, 8, 8,
   1, 1, 1, 1, 1, 1, 1, 1,  1, 1, 1, 1, 1, 1, 1, 1,
   1, 1, 1, 1, 1, 1, 1, 1,  1, 1, 1, 1, 1, 1, 1, 1,
   1, 1, 1, 1, 1, 1, 1, 1,  1, 1, 1, 1, 1, 1, 1, 1,
   1, 1, 1, 1, 1, 1, 1, 1,  1, 1, 1, 1, 1, 1, 1, 1,
   1, 1, 1, 1, 1, 1, 1, 1,  1, 1, 1, 1, 1, 1, 1, 1,
   1, 1, 1, 1, 1, 1, 1, 1,  1, 1, 1, 1, 1, 1, 1, 8,

   0, 0, 0, 0, 0, 0, 0, 0,  0, 0, 0, 0, 0, 0, 0, 0,
   0, 0, 0, 0, 0, 0, 0, 0,  0, 0, 0, 0, 0, 0, 0, 0,
   0, 0, 0, 0, 0, 0, 0, 0,  0, 0, 0, 0, 0, 0, 0, 0,
   0, 0, 0, 0, 0, 0, 0, 0,  0, 0, 0, 0, 0, 0, 0, 0,
   2, 2, 2, 2, 2, 2, 2, 2,  2, 2, 2, 2, 2, 2, 2, 2,
   2, 2, 2, 2, 2, 2, 2, 2,  2, 2, 2, 2, 2, 2, 2, 2,
   3, 3, 3, 3, 3, 3, 3, 3,  3, 3, 3, 3, 3, 3, 3, 3,
   4, 4, 4, 4, 4, 4, 4, 4,  5, 5, 5, 5, 6, 6, 7, 7,
};

static wc_uchar utf8_buf[7];

size_t
wc_ucs_to_utf8(wc_uint32 ucs, wc_uchar *utf8)
{
    if (ucs < WC_C_UTF8_L2) {
	utf8[0] =   ucs;
	utf8[1] = 0;
	return 1;
    } else if (ucs < WC_C_UTF8_L3) {
	utf8[0] =  (ucs >> 6)          | 0xc0;
	utf8[1] =  (ucs        & 0x3f) | 0x80;
	utf8[2] = 0;
	return 2;
    } else if (ucs < WC_C_UTF8_L4) {
	utf8[0] =  (ucs >> 12)         | 0xe0;
	utf8[1] = ((ucs >> 6)  & 0x3f) | 0x80;
	utf8[2] =  (ucs        & 0x3f) | 0x80;
	utf8[3] = 0;
	return 3;
    } else if (ucs < WC_C_UTF8_L5) {
	utf8[0] =  (ucs >> 18)         | 0xf0;
	utf8[1] = ((ucs >> 12) & 0x3f) | 0x80;
	utf8[2] = ((ucs >> 6)  & 0x3f) | 0x80;
	utf8[3] =  (ucs        & 0x3f) | 0x80;
	utf8[4] = 0;
	return 4;
    } else if (ucs < WC_C_UTF8_L6) {
	utf8[0] =  (ucs >> 24)         | 0xf8;
	utf8[1] = ((ucs >> 18) & 0x3f) | 0x80;
	utf8[2] = ((ucs >> 12) & 0x3f) | 0x80;
	utf8[3] = ((ucs >> 6)  & 0x3f) | 0x80;
	utf8[4] =  (ucs        & 0x3f) | 0x80;
	utf8[5] = 0;
	return 5;
    } else if (ucs <= WC_C_UCS4_END) {
	utf8[0] =  (ucs >> 30)         | 0xfc;
	utf8[1] = ((ucs >> 24) & 0x3f) | 0x80;
	utf8[2] = ((ucs >> 18) & 0x3f) | 0x80;
	utf8[3] = ((ucs >> 12) & 0x3f) | 0x80;
	utf8[4] = ((ucs >> 6)  & 0x3f) | 0x80;
	utf8[5] =  (ucs        & 0x3f) | 0x80;
	utf8[6] = 0;
	return 6;
    } else {
	utf8[0] = 0;
	return 0;
    }
}

wc_uint32
wc_utf8_to_ucs(wc_uchar *utf8)
{
    wc_uint32 ucs;

    switch (WC_UTF8_MAP[utf8[0]]) {
    case 1:
	ucs =  (wc_uint32) utf8[0];
	if (ucs >= WC_C_UTF8_L2)
	    break;
	return ucs;
    case 2:
	ucs = ((wc_uint32)(utf8[0] & 0x1f) << 6)
	    |  (wc_uint32)(utf8[1] & 0x3f);
	if (ucs < WC_C_UTF8_L2)
	    break;
	return ucs;
    case 3:
	ucs = ((wc_uint32)(utf8[0] & 0x0f) << 12)
	    | ((wc_uint32)(utf8[1] & 0x3f) << 6)
	    |  (wc_uint32)(utf8[2] & 0x3f);
	if (ucs < WC_C_UTF8_L3)
	    break;
	return ucs;
    case 4:
	ucs = ((wc_uint32)(utf8[0] & 0x07) << 18)
	    | ((wc_uint32)(utf8[1] & 0x3f) << 12)
	    | ((wc_uint32)(utf8[2] & 0x3f) << 6)
	    |  (wc_uint32)(utf8[3] & 0x3f);
	if (ucs < WC_C_UTF8_L4)
	    break;
	return ucs;
    case 5:
	ucs = ((wc_uint32)(utf8[0] & 0x03) << 24)
	    | ((wc_uint32)(utf8[1] & 0x3f) << 18)
	    | ((wc_uint32)(utf8[2] & 0x3f) << 12)
	    | ((wc_uint32)(utf8[3] & 0x3f) << 6)
	    |  (wc_uint32)(utf8[4] & 0x3f);
	if (ucs < WC_C_UTF8_L5)
	    break;
	return ucs;
    case 6:
	ucs = ((wc_uint32)(utf8[0] & 0x01) << 30)
	    | ((wc_uint32)(utf8[1] & 0x3f) << 24)
	    | ((wc_uint32)(utf8[2] & 0x3f) << 18)
	    | ((wc_uint32)(utf8[3] & 0x3f) << 12)
	    | ((wc_uint32)(utf8[4] & 0x3f) << 6)
	    |  (wc_uint32)(utf8[5] & 0x3f);
	if (ucs < WC_C_UTF8_L6)
	    break;
	return ucs;
    default:
	break;
    }
    return WC_C_UCS4_ERROR;
}

Str
wc_conv_from_utf8(Str is, wc_ces ces)
{
    Str os;
    wc_uchar *sp = (wc_uchar *)is->ptr;
    wc_uchar *ep = sp + is->length;
    wc_uchar *p;
    wc_uchar *q = NULL;
    int state = WC_UTF8_NOSTATE;
    size_t next = 0;
    wc_uint32 ucs;
    wc_status st;

    for (p = sp; p < ep && *p < 0x80; p++)
	;
    if (p == ep)
	return is;
    os = Strnew_size(is->length * 4 / 3);
    if (p > sp)
	Strcat_charp_n(os, is->ptr, (int)(p - sp));

    st.tag = NULL;
    st.ntag = 0;
    for (; p < ep; p++) {
	switch (state) {
	case WC_UTF8_NOSTATE:
	    next = WC_UTF8_MAP[*p];
	    switch (next) {
	    case 1:
		wtf_push_ucs(os, (wc_uint32)*p, &st);
		break;
	    case 8:
		Strcat_char(os, (char)*p);
		break;
	    case 0:
	    case 7:
		wtf_push_unknown(os, p, 1);
		break;
	    default:
		q = p;
		next--;
		state = WC_UTF8_NEXT;
		break;
	    }
	    break;
	case WC_UTF8_NEXT:
	    if (WC_UTF8_MAP[*p]) {
		wtf_push_unknown(os, q, p - q + 1);
		state = WC_UTF8_NOSTATE;
		break;
	    }
	    if (--next)
		break;
	    state = WC_UTF8_NOSTATE;
	    ucs = wc_utf8_to_ucs(q);
	    if (ucs == WC_C_UCS4_ERROR ||
		(ucs >= WC_C_UCS2_SURROGATE && ucs <= WC_C_UCS2_SURROGATE_END))
		wtf_push_unknown(os, q, p - q + 1);
	    else if (ucs != WC_C_UCS2_BOM)
		wtf_push_ucs(os, ucs, &st);
	    break;
	}
    }
    switch (state) {
    case WC_UTF8_NEXT:
	wtf_push_unknown(os, q, p - q);
	break;
    }
    return os;
}

static int
wc_push_tag_to_utf8(Str os, int ntag)
{
    char *p;

    if (ntag) {
	p = wc_ucs_get_tag(ntag);
	if (p == NULL)
	    ntag = 0;
    }
    if (ntag) {
	wc_ucs_to_utf8(WC_C_LANGUAGE_TAG, utf8_buf);
	Strcat_charp(os, (char *)utf8_buf);
	for (; *p; p++) {
	    wc_ucs_to_utf8(WC_C_LANGUAGE_TAG0 | *p, utf8_buf);
	    Strcat_charp(os, (char *)utf8_buf);
	}
    } else {
	wc_ucs_to_utf8(WC_C_CANCEL_TAG, utf8_buf);
	Strcat_charp(os, (char *)utf8_buf);
    }
    return ntag;
}

void
wc_push_to_utf8(Str os, wc_wchar_t cc, wc_status *st)
{
  while (1) {
    switch (WC_CCS_SET(cc.ccs)) {
    case WC_CCS_US_ASCII:
	if (st->ntag)
	    st->ntag = wc_push_tag_to_utf8(os, 0);
	Strcat_char(os, (char)(cc.code & 0x7f));
	return;
    case WC_CCS_UCS2:
    case WC_CCS_UCS4:
	if (st->ntag)
	    st->ntag = wc_push_tag_to_utf8(os, 0);
	wc_ucs_to_utf8(cc.code, utf8_buf);
	Strcat_charp(os, (char *)utf8_buf);
	return;
    case WC_CCS_UCS_TAG:
	if (WcOption.use_language_tag && wc_ucs_tag_to_tag(cc.code) != st->ntag)
	    st->ntag = wc_push_tag_to_utf8(os, wc_ucs_tag_to_tag(cc.code));
	wc_ucs_to_utf8(wc_ucs_tag_to_ucs(cc.code), utf8_buf);
	Strcat_charp(os, (char *)utf8_buf);
	return;
    case WC_CCS_ISO_8859_1:
	if (st->ntag)
	    st->ntag = wc_push_tag_to_utf8(os, 0);
	wc_ucs_to_utf8((cc.code | 0x80), utf8_buf);
	Strcat_charp(os, (char *)utf8_buf);
	return;
    case WC_CCS_UNKNOWN_W:
	if (!WcOption.no_replace) {
	    if (st->ntag)
	        st->ntag = wc_push_tag_to_utf8(os, 0);
	    Strcat_charp(os, WC_REPLACE_W);
	}
	return;
    case WC_CCS_UNKNOWN:
	if (!WcOption.no_replace) {
	    if (st->ntag)
	        st->ntag = wc_push_tag_to_utf8(os, 0);
	    Strcat_charp(os, WC_REPLACE);
	}
	return;
    default:
	if (WcOption.ucs_conv &&
		(cc.code = wc_any_to_ucs(cc)) != WC_C_UCS4_ERROR)
	    cc.ccs = WC_CCS_UCS2;
	else
	    cc.ccs = WC_CCS_IS_WIDE(cc.ccs) ? WC_CCS_UNKNOWN_W : WC_CCS_UNKNOWN;
	continue;
    }
  }
}

void
wc_push_to_utf8_end(Str os, wc_status *st)
{
    if (st->ntag)
	st->ntag = wc_push_tag_to_utf8(os, 0);
    return;
}

Str
wc_char_conv_from_utf8(wc_uchar c, wc_status *st)
{
    static Str os;
    static wc_uchar buf[6];
    static size_t nbuf, next;
    wc_uint32 ucs;

    if (st->state == -1) {
	st->state = WC_UTF8_NOSTATE;
	os = Strnew_size(8);
	st->tag = NULL;
	st->ntag = 0;
	nbuf = 0;
    }

    switch (st->state) {
    case WC_UTF8_NOSTATE:
	switch (next = WC_UTF8_MAP[c]) {
	case 1:
	    wtf_push_ucs(os, (wc_uint32)c, st);
	    break;
	case 8:
	    Strcat_char(os, (char)c);
	    break;
	case 0:
	case 7:
	    break;
	default:
	    buf[nbuf++] = c;
	    next--;
	    st->state = WC_UTF8_NEXT;
	    return NULL;
	}
	break;
    case WC_UTF8_NEXT:
	if (WC_UTF8_MAP[c])
	    break;
	buf[nbuf++] = c;
	if (--next)
	    return NULL;
	ucs = wc_utf8_to_ucs(buf);
	if (ucs == WC_C_UCS4_ERROR ||
	    (ucs >= WC_C_UCS2_SURROGATE && ucs <= WC_C_UCS2_SURROGATE_END))
	    break;
	if (ucs != WC_C_UCS2_BOM)
	    wtf_push_ucs(os, ucs, st);
	break;
    }
    st->state = -1;
    return os;
}

#endif