/*
  Unix SMB/CIFS implementation.
  Character set conversion Extensions
  Copyright (C) Igor Vergeichik <iverg@mail.ru> 2001
  Copyright (C) Andrew Tridgell 2001
  Copyright (C) Simo Sorce 2001
  Copyright (C) Martin Pool 2003

  This program is free software; you can redistribute it and/or modify
  it under the terms of the GNU General Public License as published by
  the Free Software Foundation; either version 2 of the License, or
  (at your option) any later version.

  This program is distributed in the hope that it will be useful,
  but WITHOUT ANY WARRANTY; without even the implied warranty of
  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  GNU General Public License for more details.

  You should have received a copy of the GNU General Public License
  along with this program; if not, write to the Free Software
  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.

*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif /* HAVE_CONFIG_H */

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <string.h>
#include <ctype.h>
#include <errno.h>
#include <sys/stat.h>
#include <sys/param.h>
#ifdef HAVE_USABLE_ICONV
#include <iconv.h>
#endif
#if HAVE_LOCALE_H
#include <locale.h>
#endif
#if HAVE_LANGINFO_H
#include <langinfo.h>
#endif

#include <netatalk/endian.h>
#include <atalk/logger.h>
#include <atalk/unicode.h>
#include <atalk/util.h>
#include "byteorder.h"


/**
 * @file
 *
 * @brief Character-set conversion routines built on our iconv.
 *
 * @note Samba's internal character set (at least in the 3.0 series)
 * is always the same as the one for the Unix filesystem.  It is
 * <b>not</b> necessarily UTF-8 and may be different on machines that
 * need i18n filenames to be compatible with Unix software.  It does
 * have to be a superset of ASCII.  All multibyte sequences must start
 * with a byte with the high bit set.
 *
 * @sa lib/iconv.c
 */


#define MAX_CHARSETS 20

#define CHECK_FLAGS(a,b) (((a)!=NULL) ? (*(a) & (b)) : 0 )

static atalk_iconv_t conv_handles[MAX_CHARSETS][MAX_CHARSETS];
static char* charset_names[MAX_CHARSETS];
static struct charset_functions* charsets[MAX_CHARSETS];
static char hexdig[] = "0123456789abcdef";
#define hextoint( c )   ( isdigit( c ) ? c - '0' : c + 10 - 'a' )

static char* read_charsets_from_env(charset_t ch)
{
    char *name;

    switch (ch) {
    case CH_MAC:
        if (( name = getenv( "ATALK_MAC_CHARSET" )) != NULL )
            return name;
        else
            return "MAC_ROMAN";
        break;
    case CH_UNIX:
        if (( name = getenv( "ATALK_UNIX_CHARSET" )) != NULL )
            return name;
        else
            return "LOCALE";
        break;
    default:
        break;
    }
    return "ASCII";
}


/**
 * Return the name of a charset to give to iconv().
 **/
static const char *charset_name(charset_t ch)
{
    const char *ret = NULL;
    static int first = 1;
    static char macname[128];
    static char unixname[128];

    if (first) {
        memset(macname, 0, sizeof(macname));
        memset(unixname, 0, sizeof(unixname));
        first = 0;
    }

    if (ch == CH_UCS2) ret = "UCS-2";
    else if (ch == CH_UTF8) ret = "UTF8";
    else if (ch == CH_UTF8_MAC) ret = "UTF8-MAC";
    else if (ch == CH_UNIX) {
        if (unixname[0] == '\0') {
            ret = read_charsets_from_env(CH_UNIX);
            strlcpy(unixname, ret, sizeof(unixname));
        }
        else
            ret = unixname;
    }
    else if (ch == CH_MAC) {
        if (macname[0] == '\0') {
            ret = read_charsets_from_env(CH_MAC);
            strlcpy(macname, ret, sizeof(macname));
        }
        else
            ret = macname;
    }

    if (!ret)
        ret = charset_names[ch];

#if defined(HAVE_NL_LANGINFO) && defined(CODESET)
    if (ret && strcasecmp(ret, "LOCALE") == 0) {
        const char *ln = NULL;

#ifdef HAVE_SETLOCALE
        setlocale(LC_ALL, "");
#endif
        ln = nl_langinfo(CODESET);
        if (ln) {
            /* Check whether the charset name is supported
               by iconv */
            atalk_iconv_t handle = atalk_iconv_open(ln, "UCS-2");
            if (handle == (atalk_iconv_t) -1) {
                LOG(log_debug, logtype_default, "Locale charset '%s' unsupported, using ASCII instead", ln);
                ln = "ASCII";
            } else {
                atalk_iconv_close(handle);
            }
            if (ch==CH_UNIX)
                strlcpy(unixname, ln, sizeof(unixname));
        }
        ret = ln;
    }
#else /* system doesn't have LOCALE support */
    if (ch == CH_UNIX) ret = NULL;
#endif

    if (!ret || !*ret) ret = "ASCII";
    return ret;
}

static struct charset_functions* get_charset_functions (charset_t ch)
{
    if (charsets[ch] != NULL)
        return charsets[ch];

    charsets[ch] = find_charset_functions(charset_name(ch));

    return charsets[ch];
}


static void lazy_initialize_conv(void)
{
    static int initialized = 0;

    if (!initialized) {
        initialized = 1;
        init_iconv();
    }
}

charset_t add_charset(const char* name)
{
    static charset_t max_charset_t = NUM_CHARSETS-1;
    charset_t cur_charset_t = max_charset_t+1;
    unsigned int c1;

    lazy_initialize_conv();

    for (c1=0; c1<=max_charset_t;c1++) {
        if ( strcasecmp(name, charset_name(c1)) == 0)
            return (c1);
    }

    if ( cur_charset_t >= MAX_CHARSETS )  {
        LOG (log_debug, logtype_default, "Adding charset %s failed, too many charsets (max. %u allowed)",
             name, MAX_CHARSETS);
        return (charset_t) -1;
    }

    /* First try to setup the required conversions */

    conv_handles[cur_charset_t][CH_UCS2] = atalk_iconv_open( charset_name(CH_UCS2), name);
    if (conv_handles[cur_charset_t][CH_UCS2] == (atalk_iconv_t)-1) {
        LOG(log_error, logtype_default, "Required conversion from %s to %s not supported",
            name,  charset_name(CH_UCS2));
        conv_handles[cur_charset_t][CH_UCS2] = NULL;
        return (charset_t) -1;
    }

    conv_handles[CH_UCS2][cur_charset_t] = atalk_iconv_open( name, charset_name(CH_UCS2));
    if (conv_handles[CH_UCS2][cur_charset_t] == (atalk_iconv_t)-1) {
        LOG(log_error, logtype_default, "Required conversion from %s to %s not supported",
            charset_name(CH_UCS2), name);
        conv_handles[CH_UCS2][cur_charset_t] = NULL;
        return (charset_t) -1;
    }

    /* register the new charset_t name */
    charset_names[cur_charset_t] = strdup(name);

    charsets[cur_charset_t] = get_charset_functions (cur_charset_t);
    max_charset_t++;

#ifdef DEBUG
    LOG(log_debug9, logtype_default, "Added charset %s with handle %u", name, cur_charset_t);
#endif
    return (cur_charset_t);
}

/**
 * Initialize iconv conversion descriptors.
 *
 * This is called the first time it is needed, and also called again
 * every time the configuration is reloaded, because the charset or
 * codepage might have changed.
 **/
void init_iconv(void)
{
    int c1;

    for (c1=0;c1<NUM_CHARSETS;c1++) {
        const char *name = charset_name((charset_t)c1);

        conv_handles[c1][CH_UCS2] = atalk_iconv_open( charset_name(CH_UCS2), name);
        if (conv_handles[c1][CH_UCS2] == (atalk_iconv_t)-1) {
            LOG(log_error, logtype_default, "Required conversion from %s to %s not supported",
                name,  charset_name(CH_UCS2));
            conv_handles[c1][CH_UCS2] = NULL;
        }

        if (c1 != CH_UCS2) { /* avoid lost memory, make valgrind happy */
            conv_handles[CH_UCS2][c1] = atalk_iconv_open( name, charset_name(CH_UCS2));
            if (conv_handles[CH_UCS2][c1] == (atalk_iconv_t)-1) {
                LOG(log_error, logtype_default, "Required conversion from %s to %s not supported",
                    charset_name(CH_UCS2), name);
                conv_handles[CH_UCS2][c1] = NULL;
            }
        }

        charsets[c1] = get_charset_functions (c1);
    }
}

/**
 *
 **/
static size_t add_null(charset_t to, char *buf, size_t bytesleft, size_t len)
{
    /* Terminate the string */
    if (to == CH_UCS2 && bytesleft >= 2) {
        buf[len]   = 0;
        buf[len+1] = 0;

    }
    else if ( to != CH_UCS2 && bytesleft > 0 )
        buf[len]   = 0;
    else {
        errno = E2BIG;
        return (size_t)(-1);
    }

    return len;
}


/**
 * Convert string from one encoding to another, making error checking etc
 *
 * @param src pointer to source string (multibyte or singlebyte)
 * @param srclen length of the source string in bytes
 * @param dest pointer to destination string (multibyte or singlebyte)
 * @param destlen maximal length allowed for string
 * @returns the number of bytes occupied in the destination
 **/
static size_t convert_string_internal(charset_t from, charset_t to,
                                      void const *src, size_t srclen,
                                      void *dest, size_t destlen)
{
    size_t i_len, o_len;
    size_t retval;
    const char* inbuf = (const char*)src;
    char* outbuf = (char*)dest;
    char* o_save = outbuf;
    atalk_iconv_t descriptor;

    /* Fixed based on Samba 3.0.6 */
    if (srclen == (size_t)-1) {
        if (from == CH_UCS2) {
            srclen = (strlen_w((const ucs2_t *)src)) * 2;
        } else {
            srclen = strlen((const char *)src);
        }
    }


    lazy_initialize_conv();

    descriptor = conv_handles[from][to];

    if (descriptor == (atalk_iconv_t)-1 || descriptor == (atalk_iconv_t)0) {
        return (size_t) -1;
    }

    i_len=srclen;
    o_len=destlen;
    retval = atalk_iconv(descriptor,  &inbuf, &i_len, &outbuf, &o_len);
    if(retval==(size_t)-1) {
        const char *reason="unknown error";
        switch(errno) {
        case EINVAL:
            reason="Incomplete multibyte sequence";
            break;
        case E2BIG:
            reason="No more room";
            break;
        case EILSEQ:
            reason="Illegal multibyte sequence";
            break;
        }
        LOG(log_debug, logtype_default,"Conversion error: %s",reason);
        return (size_t)-1;
    }

    /* Terminate the string */
    return add_null( to, o_save, o_len, destlen -o_len);
}


size_t convert_string(charset_t from, charset_t to,
                      void const *src, size_t srclen,
                      void *dest, size_t destlen)
{
    size_t i_len, o_len;
    ucs2_t *u;
    ucs2_t buffer[MAXPATHLEN];
    ucs2_t buffer2[MAXPATHLEN];

    /* convert from_set to UCS2 */
    if ((size_t)-1 == ( o_len = convert_string_internal( from, CH_UCS2, src, srclen,
                                                           (char*) buffer, sizeof(buffer))) ) {
        LOG(log_error, logtype_default, "Conversion failed ( %s to CH_UCS2 )", charset_name(from));
        return (size_t) -1;
    }

    /* Do pre/decomposition */
    i_len = sizeof(buffer2);
    u = buffer2;
    if (charsets[to] && (charsets[to]->flags & CHARSET_DECOMPOSED) ) {
        if ( (size_t)-1 == (i_len = decompose_w(buffer, o_len, u, &i_len)) )
            return (size_t)-1;
    }
    else if (!charsets[from] || (charsets[from]->flags & CHARSET_DECOMPOSED)) {
        if ( (size_t)-1 == (i_len = precompose_w(buffer, o_len, u, &i_len)) )
            return (size_t)-1;
    }
    else {
        u = buffer;
        i_len = o_len;
    }
    /* Convert UCS2 to to_set */
    if ((size_t)(-1) == ( o_len = convert_string_internal( CH_UCS2, to, (char*) u, i_len, dest, destlen)) ) {
        LOG(log_error, logtype_default, "Conversion failed (CH_UCS2 to %s):%s", charset_name(to), strerror(errno));
        return (size_t) -1;
    }

    return o_len;
}



/**
 * Convert between character sets, allocating a new buffer for the result.
 *
 * @param srclen length of source buffer.
 * @param dest always set at least to NULL
 * @note -1 is not accepted for srclen.
 *
 * @returns Size in bytes of the converted string; or -1 in case of error.
 **/

static size_t convert_string_allocate_internal(charset_t from, charset_t to,
                                               void const *src, size_t srclen, char **dest)
{
    size_t i_len, o_len, destlen;
    size_t retval;
    const char *inbuf = (const char *)src;
    char *outbuf = NULL, *ob = NULL;
    atalk_iconv_t descriptor;

    *dest = NULL;

    if (src == NULL || srclen == (size_t)-1)
        return (size_t)-1;

    lazy_initialize_conv();

    descriptor = conv_handles[from][to];

    if (descriptor == (atalk_iconv_t)-1 || descriptor == (atalk_iconv_t)0) {
        /* conversion not supported, return -1*/
        LOG(log_debug, logtype_default, "convert_string_allocate: conversion not supported!");
        return -1;
    }

    destlen = MAX(srclen, 512);
convert:
    destlen = destlen * 2;
    outbuf = (char *)realloc(ob, destlen);
    if (!outbuf) {
        LOG(log_debug, logtype_default,"convert_string_allocate: realloc failed!");
        SAFE_FREE(ob);
        return (size_t)-1;
    } else {
        ob = outbuf;
    }
    inbuf = src;   /* this restarts the whole conversion if buffer needed to be increased */
    i_len = srclen;
    o_len = destlen;
    retval = atalk_iconv(descriptor,
                         &inbuf, &i_len,
                         &outbuf, &o_len);
    if(retval == (size_t)-1)        {
        const char *reason="unknown error";
        switch(errno) {
        case EINVAL:
            reason="Incomplete multibyte sequence";
            break;
        case E2BIG:
            goto convert;
        case EILSEQ:
            reason="Illegal multibyte sequence";
            break;
        }
        LOG(log_debug, logtype_default,"Conversion error: %s(%s)",reason,inbuf);
        SAFE_FREE(ob);
        return (size_t)-1;
    }


    destlen = destlen - o_len;

    /* Terminate the string */
    if (to == CH_UCS2 && o_len >= 2) {
        ob[destlen] = 0;
        ob[destlen+1] = 0;
        *dest = (char *)realloc(ob,destlen+2);
    }
    else if ( to != CH_UCS2 && o_len > 0 ) {
        ob[destlen] = 0;
        *dest = (char *)realloc(ob,destlen+1);
    }
    else {
        goto convert; /* realloc */
    }

    if (destlen && !*dest) {
        LOG(log_debug, logtype_default, "convert_string_allocate: out of memory!");
        SAFE_FREE(ob);
        return (size_t)-1;
    }

    return destlen;
}


size_t convert_string_allocate(charset_t from, charset_t to,
                               void const *src, size_t srclen,
                               char ** dest)
{
    size_t i_len, o_len;
    ucs2_t *u;
    ucs2_t buffer[MAXPATHLEN];
    ucs2_t buffer2[MAXPATHLEN];

    *dest = NULL;

    /* convert from_set to UCS2 */
    if ((size_t)(-1) == ( o_len = convert_string_internal( from, CH_UCS2, src, srclen,
                                                           buffer, sizeof(buffer))) ) {
        LOG(log_error, logtype_default, "Conversion failed ( %s to CH_UCS2 )", charset_name(from));
        return (size_t) -1;
    }

    /* Do pre/decomposition */
    i_len = sizeof(buffer2);
    u = buffer2;
    if (charsets[to] && (charsets[to]->flags & CHARSET_DECOMPOSED) ) {
        if ( (size_t)-1 == (i_len = decompose_w(buffer, o_len, u, &i_len)) )
            return (size_t)-1;
    }
    else if ( !charsets[from] || (charsets[from]->flags & CHARSET_DECOMPOSED) ) {
        if ( (size_t)-1 == (i_len = precompose_w(buffer, o_len, u, &i_len)) )
            return (size_t)-1;
    }
    else {
        u = buffer;
        i_len = o_len;
    }

    /* Convert UCS2 to to_set */
    if ((size_t)-1 == ( o_len = convert_string_allocate_internal( CH_UCS2, to, (char*)u, i_len, dest)) )
        LOG(log_error, logtype_default, "Conversion failed (CH_UCS2 to %s):%s", charset_name(to), strerror(errno));

    return o_len;

}

size_t charset_strupper(charset_t ch, const char *src, size_t srclen, char *dest, size_t destlen)
{
    size_t size;
    char *buffer;

    size = convert_string_allocate_internal(ch, CH_UCS2, src, srclen,
                                            (char**) &buffer);
    if (size == (size_t)-1) {
        SAFE_FREE(buffer);
        return size;
    }
    if (!strupper_w((ucs2_t *)buffer) && (dest == src)) {
        free(buffer);
        return srclen;
    }

    size = convert_string_internal(CH_UCS2, ch, buffer, size, dest, destlen);
    free(buffer);
    return size;
}

size_t charset_strlower(charset_t ch, const char *src, size_t srclen, char *dest, size_t destlen)
{
    size_t size;
    char *buffer;

    size = convert_string_allocate_internal(ch, CH_UCS2, src, srclen,
                                            (char **) &buffer);
    if (size == (size_t)-1) {
        SAFE_FREE(buffer);
        return size;
    }
    if (!strlower_w((ucs2_t *)buffer) && (dest == src)) {
        free(buffer);
        return srclen;
    }

    size = convert_string_internal(CH_UCS2, ch, buffer, size, dest, destlen);
    free(buffer);
    return size;
}


size_t unix_strupper(const char *src, size_t srclen, char *dest, size_t destlen)
{
    return charset_strupper( CH_UNIX, src, srclen, dest, destlen);
}

size_t unix_strlower(const char *src, size_t srclen, char *dest, size_t destlen)
{
    return charset_strlower( CH_UNIX, src, srclen, dest, destlen);
}

size_t utf8_strupper(const char *src, size_t srclen, char *dest, size_t destlen)
{
    return charset_strupper( CH_UTF8, src, srclen, dest, destlen);
}

size_t utf8_strlower(const char *src, size_t srclen, char *dest, size_t destlen)
{
    return charset_strlower( CH_UTF8, src, srclen, dest, destlen);
}

/**
 * Copy a string from a charset_t char* src to a UCS2 destination, allocating a buffer
 *
 * @param dest always set at least to NULL
 *
 * @returns The number of bytes occupied by the string in the destination
 *         or -1 in case of error.
 **/

size_t charset_to_ucs2_allocate(charset_t ch, ucs2_t **dest, const char *src)
{
    size_t src_len = strlen(src);

    *dest = NULL;
    return convert_string_allocate(ch, CH_UCS2, src, src_len, (char**) dest);
}

/** -----------------------------------
 * Copy a string from a charset_t char* src to a UTF-8 destination, allocating a buffer
 *
 * @param dest always set at least to NULL
 *
 * @returns The number of bytes occupied by the string in the destination
 **/

size_t charset_to_utf8_allocate(charset_t ch, char **dest, const char *src)
{
    size_t src_len = strlen(src);

    *dest = NULL;
    return convert_string_allocate(ch, CH_UTF8, src, src_len, dest);
}

/** -----------------------------------
 * Copy a string from a UCS2 src to a unix char * destination, allocating a buffer
 *
 * @param dest always set at least to NULL
 *
 * @returns The number of bytes occupied by the string in the destination
 **/

size_t ucs2_to_charset(charset_t ch, const ucs2_t *src, char *dest, size_t destlen)
{
    size_t src_len = (strlen_w(src)) * sizeof(ucs2_t);
    return convert_string(CH_UCS2, ch, src, src_len, dest, destlen);
}

/* --------------------------------- */
size_t ucs2_to_charset_allocate(charset_t ch, char **dest, const ucs2_t *src)
{
    size_t src_len = (strlen_w(src)) * sizeof(ucs2_t);
    *dest = NULL;
    return convert_string_allocate(CH_UCS2, ch, src, src_len, dest);
}

/** ---------------------------------
 * Copy a string from a UTF-8 src to a unix char * destination, allocating a buffer
 *
 * @param dest always set at least to NULL
 *
 * @returns The number of bytes occupied by the string in the destination
 **/

size_t utf8_to_charset_allocate(charset_t ch, char **dest, const char *src)
{
    size_t src_len = strlen(src);
    *dest = NULL;
    return convert_string_allocate(CH_UTF8, ch, src, src_len, dest);
}

size_t charset_precompose ( charset_t ch, char * src, size_t inlen, char * dst, size_t outlen)
{
    char *buffer;
    ucs2_t u[MAXPATHLEN];
    size_t len;
    size_t ilen;

    if ((size_t)(-1) == (len = convert_string_allocate_internal(ch, CH_UCS2, src, inlen, &buffer)) )
        return len;

    ilen=sizeof(u);

    if ( (size_t)-1 == (ilen = precompose_w((ucs2_t *)buffer, len, u, &ilen)) ) {
        free (buffer);
        return (size_t)(-1);
    }

    if ((size_t)(-1) == (len = convert_string_internal( CH_UCS2, ch, (char*)u, ilen, dst, outlen)) ) {
        free (buffer);
        return (size_t)(-1);
    }

    free(buffer);
    return (len);
}

size_t charset_decompose ( charset_t ch, char * src, size_t inlen, char * dst, size_t outlen)
{
    char *buffer;
    ucs2_t u[MAXPATHLEN];
    size_t len;
    size_t ilen;

    if ((size_t)(-1) == (len = convert_string_allocate_internal(ch, CH_UCS2, src, inlen, &buffer)) )
        return len;

    ilen=sizeof(u);

    if ( (size_t)-1 == (ilen = decompose_w((ucs2_t *)buffer, len, u, &ilen)) ) {
        free (buffer);
        return (size_t)(-1);
    }

    if ((size_t)(-1) == (len = convert_string_internal( CH_UCS2, ch, (char*)u, ilen, dst, outlen)) ) {
        free (buffer);
        return (size_t)(-1);
    }

    free(buffer);
    return (len);
}

size_t utf8_precompose ( char * src, size_t inlen, char * dst, size_t outlen)
{
    return charset_precompose ( CH_UTF8, src, inlen, dst, outlen);
}

size_t utf8_decompose ( char * src, size_t inlen, char * dst, size_t outlen)
{
    return charset_decompose ( CH_UTF8, src, inlen, dst, outlen);
}

#if 0
static char  debugbuf[ MAXPATHLEN +1 ];
char * debug_out ( char * seq, size_t len)
{
    size_t i = 0;
    unsigned char *p;
    char *q;

    p = (unsigned char*) seq;
    q = debugbuf;

    for ( i = 0; i<=(len-1); i++)
    {
        sprintf(q, "%2.2x.", *p);
        q += 3;
        p++;
    }
    *q=0;
    q = debugbuf;
    return q;
}
#endif

/*
 * Convert from MB to UCS2 charset
 * Flags:
 *      CONV_UNESCAPEHEX:    ':XX' will be converted to an UCS2 character
 *      CONV_IGNORE:         return the first convertable characters.
 *      CONV_FORCE:  force convertion
 * FIXME:
 *      This will *not* work if the destination charset is not multibyte, i.e. UCS2->UCS2 will fail
 *      The (un)escape scheme is not compatible to the old cap style escape. This is bad, we need it
 *      for e.g. HFS cdroms.
 */

static size_t pull_charset_flags (charset_t from_set, charset_t cap_set, const char *src, size_t srclen, char* dest, size_t destlen, u_int16_t *flags)
{
    const u_int16_t option = (flags ? *flags : 0);
    size_t i_len, o_len;
    size_t j = 0;
    const char* inbuf = (const char*)src;
    char* outbuf = dest;
    atalk_iconv_t descriptor;
    atalk_iconv_t descriptor_cap;

    if (srclen == (size_t)-1)
        srclen = strlen(src) + 1;

    descriptor = conv_handles[from_set][CH_UCS2];
    descriptor_cap = conv_handles[cap_set][CH_UCS2];

    if (descriptor == (atalk_iconv_t)-1 || descriptor == (atalk_iconv_t)0) {
        errno = EINVAL;
        return (size_t)-1;
    }

    i_len=srclen;
    o_len=destlen;

    while (i_len > 0) {
        if ((option & CONV_UNESCAPEHEX)) {
            for (j = 0; j < i_len; ++j) {
                if (inbuf[j] == ':') break;
            }
            j = i_len - j;
            i_len -= j;
        }

        if (i_len > 0 &&
            atalk_iconv(descriptor, &inbuf, &i_len, &outbuf, &o_len) == (size_t)-1) {
            if (errno == EILSEQ || errno == EINVAL) {
                errno = EILSEQ;
                if ((option & CONV_IGNORE)) {
                    *flags |= CONV_REQMANGLE;
                    return destlen - o_len;
                }
                if ((option & CONV__EILSEQ)) {
                    if (o_len < 2) {
                        errno = E2BIG;
                        goto end;
                    }
                    *((ucs2_t *)outbuf) = (ucs2_t) IGNORE_CHAR; /**inbuf */
                    inbuf++;
                    i_len--;
                    outbuf += 2;
                    o_len -= 2;
                    /* FIXME reset stat ? */
                    continue;
                }
            }
            goto end;
        }

        if (j) {
            /* we're at the start on an hex encoded ucs2 char */
            char h[MAXPATHLEN];
            size_t hlen = 0;

            i_len = j, j = 0;
            while (i_len >= 3 && inbuf[0] == ':' &&
                   isxdigit(inbuf[1]) && isxdigit(inbuf[2])) {
                h[hlen++] = (hextoint(inbuf[1]) << 4) | hextoint(inbuf[2]);
                inbuf += 3;
                i_len -= 3;
            }
            if (hlen) {
                const char *h_buf = h;
                if (atalk_iconv(descriptor_cap, &h_buf, &hlen, &outbuf, &o_len) == (size_t)-1) {
                    i_len += hlen * 3;
                    inbuf -= hlen * 3;
                    if (errno == EILSEQ && (option & CONV_IGNORE)) {
                        *flags |= CONV_REQMANGLE;
                        return destlen - o_len;
                    }
                    goto end;
                }
            } else {
                /* We have an invalid :xx sequence */
                errno = EILSEQ;
                if ((option & CONV_IGNORE)) {
                    *flags |= CONV_REQMANGLE;
                    return destlen - o_len;
                }
                goto end;
            }
        }
    }
end:
    return (i_len + j == 0 || (option & CONV_FORCE)) ? destlen - o_len : (size_t)-1;
}

/*
 * Convert from UCS2 to MB charset
 * Flags:
 *      CONV_ESCAPEDOTS: escape leading dots
 *      CONV_ESCAPEHEX:  unconvertable characters and '/' will be escaped to :XX
 *      CONV_IGNORE:     return the first convertable characters.
 *      CONV__EILSEQ:    unconvertable characters will be replaced with '_'
 *      CONV_FORCE:  force convertion
 * FIXME:
 *      CONV_IGNORE and CONV_ESCAPEHEX can't work together. Should we check this ?
 *      This will *not* work if the destination charset is not multibyte, i.e. UCS2->UCS2 will fail
 *      The escape scheme is not compatible to the old cap style escape. This is bad, we need it
 *      for e.g. HFS cdroms.
 */


static size_t push_charset_flags (charset_t to_set, charset_t cap_set, char* src, size_t srclen, char* dest, size_t destlen, u_int16_t *flags)
{
    const u_int16_t option = (flags ? *flags : 0);
    size_t i_len, o_len, i;
    size_t j = 0;
    const char* inbuf = (const char*)src;
    char* outbuf = (char*)dest;
    atalk_iconv_t descriptor;
    atalk_iconv_t descriptor_cap;
    char escch;                 /* 150210: uninitialized OK, depends on j */

    descriptor = conv_handles[CH_UCS2][to_set];
    descriptor_cap = conv_handles[CH_UCS2][cap_set];

    if (descriptor == (atalk_iconv_t)-1 || descriptor == (atalk_iconv_t)0) {
        errno = EINVAL;
        return (size_t) -1;
    }

    i_len=srclen;
    o_len=destlen;

    if ((option & CONV_ESCAPEDOTS) &&
        i_len >= 2 && SVAL(inbuf, 0) == 0x002e) { /* 0x002e = . */
        if (o_len < 3) {
            errno = E2BIG;
            goto end;
        }
        *outbuf++ = ':';
        *outbuf++ = '2';
        *outbuf++ = 'e';
        o_len -= 3;
        inbuf += 2;
        i_len -= 2;
        *flags |= CONV_REQESCAPE;
    }

    while (i_len >= 2) {
        if ((option & CONV_ESCAPEHEX)) {
            for (i = 0; i < i_len; i += 2) {
                ucs2_t c = SVAL(inbuf, i);
                switch (c) {
                case 0x003a: /* 0x003a = ':' */
                    if ( ! (option & CONV_ALLOW_COLON)) {
                        errno = EILSEQ;
                        goto end;
                    }
                    escch = c;
                    j = i_len - i;
                    i_len = i;
                    break;
                case 0x002f: /* 0x002f = '/' */
                    escch = c;
                    j = i_len - i;
                    i_len = i;
                    break;
                }
            }
        }
        while (i_len > 0 &&
               atalk_iconv(descriptor, &inbuf, &i_len, &outbuf, &o_len) == (size_t)-1) {
            if (errno == EILSEQ) {
                if ((option & CONV_IGNORE)) {
                    *flags |= CONV_REQMANGLE;
                    return destlen - o_len;
                }
                if ((option & CONV_ESCAPEHEX)) {
                    const size_t bufsiz = o_len / 3 + 1;
                    char *buf = malloc(bufsiz);
                    size_t buflen;

                    if (!buf)
                        goto end;
                    i = i_len;
                    for (buflen = 1; buflen <= bufsiz; ++buflen) {
                        char *b = buf;
                        size_t o = buflen;
                        if (atalk_iconv(descriptor_cap, &inbuf, &i, &b, &o) != (size_t)-1) {
                            buflen -= o;
                            break;
                        } else if (errno != E2BIG) {
                            SAFE_FREE(buf);
                            goto end;
                        } else if (o < buflen) {
                            buflen -= o;
                            break;
                        }
                    }
                    if (o_len < buflen * 3) {
                        SAFE_FREE(buf);
                        errno = E2BIG;
                        goto end;
                    }
                    o_len -= buflen * 3;
                    i_len = i;
                    for (i = 0; i < buflen; ++i) {
                        *outbuf++ = ':';
                        *outbuf++ = hexdig[(buf[i] >> 4) & 0x0f];
                        *outbuf++ = hexdig[buf[i] & 0x0f];
                    }
                    SAFE_FREE(buf);
                    *flags |= CONV_REQESCAPE;
                    continue;
                }
            }
            goto end;
        }

        if (j) {
            i_len = j, j = 0;
            if (o_len < 3) {
                errno = E2BIG;
                goto end;
            }
            switch (escch) {
            case '/':
                *outbuf++ = ':';
                *outbuf++ = '2';
                *outbuf++ = 'f';
                break;
            case ':':
                *outbuf++ = ':';
                *outbuf++ = '3';
                *outbuf++ = 'a';
                break;
            default:
                /*
                 *  THIS SHOULD NEVER BE REACHED !!!
                 *  As a safety net I put in a ' ' here
                 */
                *outbuf++ = ':';
                *outbuf++ = '2';
                *outbuf++ = '0';
                break;
            }
            o_len -= 3;
            inbuf += 2;
            i_len -= 2;
        }
    }
    if (i_len > 0) errno = EINVAL;
end:
    return (i_len + j == 0 || (option & CONV_FORCE)) ? destlen - o_len : (size_t)-1;
}

/*
 * FIXME the size is a mess we really need a malloc/free logic
 *`dest size must be dest_len +2
 */
size_t convert_charset ( charset_t from_set, charset_t to_set, charset_t cap_charset, const char *src, size_t src_len, char *dest, size_t dest_len, u_int16_t *flags)
{
    size_t i_len, o_len;
    ucs2_t *u;
    ucs2_t buffer[MAXPATHLEN +2];
    ucs2_t buffer2[MAXPATHLEN +2];

    lazy_initialize_conv();

    /* convert from_set to UCS2 */
    if ((size_t)(-1) == ( o_len = pull_charset_flags( from_set, cap_charset, src, src_len,
                                                      (char *) buffer, sizeof(buffer) -2, flags)) ) {
        LOG(log_error, logtype_default, "Conversion failed ( %s to CH_UCS2 )", charset_name(from_set));
        return (size_t) -1;
    }

    if ( o_len == 0)
        return o_len;

    /* Do pre/decomposition */
    i_len = sizeof(buffer2) -2;
    u = buffer2;
    if (CHECK_FLAGS(flags, CONV_DECOMPOSE) || (charsets[to_set] && (charsets[to_set]->flags & CHARSET_DECOMPOSED)) ) {
        if ( (size_t)-1 == (i_len = decompose_w(buffer, o_len, u, &i_len)) )
            return (size_t)(-1);
    }
    else if (CHECK_FLAGS(flags, CONV_PRECOMPOSE) || !charsets[from_set] || (charsets[from_set]->flags & CHARSET_DECOMPOSED)) {
        if ( (size_t)-1 == (i_len = precompose_w(buffer, o_len, u, &i_len)) )
            return (size_t)(-1);
    }
    else {
        u = buffer;
        i_len = o_len;
    }
    /* null terminate */
    u[i_len] = 0;
    u[i_len +1] = 0;

    /* Do case conversions */
    if (CHECK_FLAGS(flags, CONV_TOUPPER)) {
        strupper_w(u);
    }
    else if (CHECK_FLAGS(flags, CONV_TOLOWER)) {
        strlower_w(u);
    }

    /* Convert UCS2 to to_set */
    if ((size_t)(-1) == ( o_len = push_charset_flags( to_set, cap_charset, (char *)u, i_len, dest, dest_len, flags )) ) {
        LOG(log_error, logtype_default,
            "Conversion failed (CH_UCS2 to %s):%s", charset_name(to_set), strerror(errno));
        return (size_t) -1;
    }
    /* null terminate */
    dest[o_len] = 0;
    dest[o_len +1] = 0;

    return o_len;
}