/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright(c) 1998 Sun Microsystems, Inc.
 * All rights reserved.
 */
#include <stdio.h>
#include <errno.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/isa_defs.h>
#include <unicode_gb2312.h>
#include "common_defs.h"

#define MSB             0x80
#define NON_ID_CHAR     '?'

typedef struct _icv_state {
        short   _ustate;
        char    _cbuf[3];
        boolean little_endian;
        boolean bom_written;
} _iconv_st;

enum    _USTATE { U0, U1, U2, U3, U4, U5, U6 };

int unicode_to_gb(char, char, char *, int, int *);

/*
 * Open; called from iconv_open()
 */
void *
_icv_open()
{
        _iconv_st *st;

        if ((st = (_iconv_st *)malloc(sizeof(_iconv_st))) == NULL) {
                errno = ENOMEM;
                return ((void *) -1);
        }

        st->_ustate = U0;
        st->little_endian = false;
        st->bom_written = false;
#if defined(UCS_2LE)
        st->little_endian = true;
        st->bom_written = true;
#endif
        return ((void *)st);
}


/*
 * Close; called from iconv_close()
 */
void
_icv_close(_iconv_st *st)
{
        if (st == NULL)
                errno = EBADF;
        else
                free(st);
}


/*
 * Actual conversion; called from iconv()
 */
size_t
_icv_iconv(_iconv_st *st, char **inbuf, size_t*inbytesleft,
                        char **outbuf, size_t*outbytesleft)
{
        char    c1, c2;
        int     n;
        int     uconv_num = 0;

        if (st == NULL) {
                errno = EBADF;
                return ((size_t)-1);
        }

        if (inbuf == NULL || *inbuf == NULL) { /* Reset request. */
                st->_ustate = U0;
                return ((size_t)0);
        }

        errno = 0;

        while (*inbytesleft > 0 && *outbytesleft > 0) {

            uchar_t first_byte;

            switch (st->_ustate) {
            case U0:
                /*
                 * Code converion for UCS-2LE to support Samba
                 */
                if (st->little_endian) {
                    st->_ustate = U1;
                    st->_cbuf[0] = **inbuf;
                }
                else if ((**inbuf & MSB) == 0) {        /* ASCII */
                    **outbuf = **inbuf;
                    (*outbuf)++; (*outbytesleft)--;
                } else if ((**inbuf & 0xe0) == 0xc0) { /* 0xc2..0xdf */

                    /* invalid sequence if the first byte is either 0xc0 or 0xc1 */
                    if ( number_of_bytes_in_utf8_char[((uchar_t)**inbuf)] == ICV_TYPE_ILLEGAL_CHAR )
                        errno = EILSEQ;
                    else {
                        st->_ustate = U1;
                        st->_cbuf[0] = **inbuf;
                    }
                } else if ((**inbuf & 0xf0) == 0xe0) { /* 0xe0..0xef */
                    st->_ustate = U2;
                    st->_cbuf[0] = **inbuf;
                } else {
                    /* four bytes of UTF-8 sequence */
                    if ( number_of_bytes_in_utf8_char[((uchar_t)**inbuf)] == ICV_TYPE_ILLEGAL_CHAR )
                        errno = EILSEQ;
                    else
                     {
                        st->_ustate = U4;
                        st->_cbuf[0] = **inbuf;
                     }
                }
                break;
            case U1:
                if ((**inbuf & 0xc0) == MSB || st->little_endian) {     /* Two-byte UTF */
                    int uconv_num_internal = 0;

                    /*
                     * Code conversion for UCS-2LE to support Samba
                     */
                    if (st->little_endian) {
                        c1 = **inbuf;
                        c2 = st->_cbuf[0];

                        /*
                         * It's ASCII
                         */
                        if (c1 == 0 && (c2 & MSB) == 0) {
                          *(*outbuf)++ = c2;
                          (*outbytesleft) --;
                          st->_ustate = U0;
                          break;
                        }
                    } else {
                        c1 = (st->_cbuf[0]&0x1c)>>2;
                        c2 = ((st->_cbuf[0]&0x03)<<6) | ((**inbuf)&0x3f);
                    }
                    n = unicode_to_gb(c1, c2, *outbuf, *outbytesleft, &uconv_num_internal);
                    if (n > 0) {
                        (*outbuf) += n, (*outbytesleft) -= n;

                        uconv_num += uconv_num_internal;

                        st->_ustate = U0;
                    } else if (n == 0) {
                        errno = E2BIG;
                    } else { /* n == -1 if unicode is either FFFE or 0xFFFF */
                        errno = EILSEQ;
                    }
                } else {
                    errno = EILSEQ;
                }
                break;
            case U2:

                first_byte = st->_cbuf[0];

                /* if the first byte is 0xed, it is illegal sequence if the second
                 * one is between 0xa0 and 0xbf because surrogate section is ill-formed
                 */
                if (((uchar_t)**inbuf) < valid_min_2nd_byte[first_byte] ||
                     ((uchar_t)**inbuf) > valid_max_2nd_byte[first_byte] )
                    errno = EILSEQ;
                else
                  {
                    st->_ustate = U3;
                    st->_cbuf[1] = **inbuf;
                  }
                break;
            case U3:
                if ((**inbuf & 0xc0) == MSB) {  /* Three-byte UTF */
                    int uconv_num_internal = 0;

                    c1 = ((st->_cbuf[0]&0x0f)<<4) | ((st->_cbuf[1]&0x3c)>>2);
                    c2 = ((st->_cbuf[1]&0x03)<<6) | ((**inbuf)&0x3f);
                    n = unicode_to_gb(c1, c2, *outbuf, *outbytesleft, &uconv_num_internal);
                    if (n > 0) {
                        (*outbuf) += n, (*outbytesleft) -= n;

                        uconv_num += uconv_num_internal;

                        st->_ustate = U0;
                    } else if ( n == 0 ) {
                        errno = E2BIG;
                    } else { /* n == -1 if unicode is either 0xFFFE or 0xFFFF */
                        errno = EILSEQ;
                    }
                } else {
                    errno = EILSEQ;
                }
                break;
             case U4:

                first_byte = st->_cbuf[0];

                /* if the first byte is 0xf0, it is illegal sequence if
                 * the second one is between 0x80 and 0x8f
                 * for Four-Byte UTF: U+10000..U+10FFFF
                 */
                if (((uchar_t)**inbuf) < valid_min_2nd_byte[first_byte] ||
                     ((uchar_t)**inbuf) > valid_max_2nd_byte[first_byte] )
                    errno = EILSEQ;
                else
                    {
                      st->_ustate = U5;
                      st->_cbuf[1] = **inbuf;
                    }
                break;
             case U5:
                if ((**inbuf & 0xc0) == MSB) /* 0x80..0xbf */
                  {
                     st->_ustate = U6;
                     st->_cbuf[2] = **inbuf;
                  }
                else
                  errno = EILSEQ;
                break;
             case U6:
                if ((**inbuf & 0xc0) == MSB) /* 0x80..0xbf */
                  {
                    /* all gb2312 characters are in Unicode Plane 0
                     * so replace these other 16 planes with 0x3f3f
                     */
                    if ( *outbytesleft < 2 )
                       errno = E2BIG;
                    else
                       {
                          **outbuf = NON_ID_CHAR;
                          *(*outbuf+1) = NON_ID_CHAR;
                          (*outbytesleft) -= 2;

                          uconv_num++;

                          st->_ustate = U0;
                       }
                  }
                else
                  errno = EILSEQ;
                break;
            }

            if (errno) break;

            (*inbuf)++; (*inbytesleft)--;
        }

        if (*inbytesleft == 0 && st->_ustate != U0)
                errno = EINVAL;

        if (*inbytesleft > 0 && *outbytesleft == 0)
                errno = E2BIG;

        if (errno) {
           int num_reversed_bytes = 0;

           switch (st->_ustate)
             {
              case U1:
                num_reversed_bytes = 1;
                break;
              case U2:
                num_reversed_bytes = 1;
                break;
              case U3:
                num_reversed_bytes = 2;
                break;
              case U4:
                num_reversed_bytes = 1;
                break;
              case U5:
                num_reversed_bytes = 2;
                break;
              case U6:
                num_reversed_bytes = 3;
                break;
             }

           /*
            * if error, *inbuf points to the byte following the last byte
            * successfully used in conversion.
            */
           *inbuf -= num_reversed_bytes;
           *inbytesleft += num_reversed_bytes;
           st->_ustate = U0;

           return ((size_t)-1);
        }

        return uconv_num;
}

/* return values: 0 - no enough space to hold the GB2312 code
 *               -1 - illegal sequence
 *               >0 - buffer length
 */
int unicode_to_gb(char in_byte1, char in_byte2, char *buf, int buflen, int *uconv_num)
{
        int     gb, unicode;
        int     i, l, h;

        if (buflen < 2)
                return 0;
        unicode = ((in_byte1 & 0xff) << 8) + (in_byte2 & 0xff);
        /* 0xfffe and 0xffff should not be allowed */
        if ( unicode == 0xFFFE || unicode == 0xFFFF ) return -1;

        for (l = 0, h = UNICODEMAX; l < h; ) {
                if (unicode_gb_tab[l].key == unicode) {
                        i = l;
                        break;
                }
                if (unicode_gb_tab[h].key == unicode) {
                        i = h;
                        break;
                }
                i = (l + h) / 2;
                if (unicode_gb_tab[i].key == unicode)
                        break;
                if (unicode_gb_tab[i].key < unicode)
                        l = i + 1;
                else    h = i - 1;
        }
        if (unicode == unicode_gb_tab[i].key) {
                gb = unicode_gb_tab[i].value;
                *buf = ((gb & 0xff00) >> 8) | MSB;
                *(buf+1) = (gb & 0xff) | MSB;
        } else {
                *buf = NON_ID_CHAR;
                *(buf+1) = NON_ID_CHAR;

                /* non-identical conversion */
                *uconv_num = 1;
        }

        return 2;
}