question about encoding sequance

[yueyue papa]

My changeds gnokii in windows iconv solution. But Is there anyone
could answer my previous encoding question.

win_iconv:　(GTK+ depends this )
　　　http://www.gtk.org/download-windows.html

I do not use dll, but using the source code integrated in gnokii.


1> copy the iconv.h to any include file in gnokii
2> add ICONV define in iconv.h
   #define ICONV_CONST const
   The ICONV_CONST is used in gnokii

3> add the win_iconv.c
   The souce code needs to add in gnokii library, because the libary
needs this feature

4> change config.h
   #define HAVE_ICONV 1   /*minye.li*/

Try to rebuild the project, see whether there is compiler error. Solve
it if thre is missing part.

5> Change code gsm-encoding.c
static char application_encoding[64] = "";
==>
static char application_encoding[64] = "gb2312";

The data will change flowcontrol in

GNOKII_API const char *gn_char_get_encoding()

   In function: static int char_wctomb(char *dst, wchar_t src, MBSTATE *mbs)

cd = iconv_open(gn_char_get_encoding(), "WCHAR_T");
==>
cd = iconv_open(gn_char_get_encoding(), "UTF16");

   In function: unsigned int char_unicode_decode(unsigned char* dest,
const unsigned char* src, int len)

   wchar_t wc = src[i * 2] << 8 | src[(i * 2) + 1];
==>
   wchar_t wc = src[i * 2] | (src[(i * 2) + 1] << 8 );

Rebuild and run again.

The chinese SMS could be displayed.

There are other parts not clear
1> How unicode is encoded in SMS
   little indian, or big indian.

2> WCHAR_T has not bee complete changed in code.
   cd = iconv_open(gn_char_get_encoding(), "WCHAR_T");

3> how to encode/decode unicode byte sequance in SMS

4> phone book name display has not added UNICODE sequance.

---------- Forwarded message ----------
From: yueyue papa <address@hidden>
Date: Jun 3, 2008 4:57 PM
Subject: question about encoding sequance
To: address@hidden


Hi All,


Is there spec related how the unicode is encode in SMS, using big
indian or litle indian, or the indian flag in the SMS ?  After I
changed the wc encode sequance, the iconv works fine. Is this the
encode problem or my problem for how to call iconv. I am in windows
platform and use win_iconv as the code page convert tool.

/* null terminates the string */
unsigned int char_unicode_decode(unsigned char* dest, const unsigned
char* src, int len)
{
       int i, length = 0, pos = 0;
       MBSTATE mbs;

       MBSTATE_DEC_CLEAR(mbs);
       for (i = 0; i < len / 2; i++) {
#if 0
               wchar_t wc = src[i * 2] << 8 | src[(i * 2) + 1];
//original code
#else
               wchar_t wc = src[i * 2]  | src[(i * 2) + 1] << 8;  //I
changed code
#endif
               length = char_uni_alphabet_decode(wc, dest, &mbs);
               dest += length;
               pos += length;
       }
       *dest = 0;
       return pos;
}


Lee