RE: Character set

Hi All,

In a similar problem to Davids (below), we are doing a test to an SMPP 
SMSC and they are sending back message text encoded as 7 bit packed GSM 
alphabet.

Even though the SMPP driver calls:
charset_gsm_to_latin1(msg->sms.msgdata); 

this does not convert 7 bit packed GSM data to latin 1, it only 
converts 8 bit packed GSM data. Should this function unpack 7 bit data or
is it presumed that will always only get 8 bit GSM encoded characters?

Comments welcome,

Michael.

> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]]On Behalf Of
> David Tully
> Sent: 09 July 2003 14:50
> To: [email protected]
> Subject: Character set
> 
> 
> Hi
> 
> I'm struggling to get this connection fully working - it's 
> the one with the
> non-standard message_id that I was mailing about last week.
> 
> I'm told it's GSM-7..
> 
> 2003-07-09 12:46:39 [6] DEBUG:   data_coding: 0 = 0x00000000
> 2003-07-09 12:46:39 [6] DEBUG:   sm_default_msg_id: 0 = 0x00000000
> 2003-07-09 12:46:39 [6] DEBUG:   sm_length: 4 = 0x00000004
> 2003-07-09 12:46:39 [6] DEBUG:   short_message:
> 2003-07-09 12:46:39 [6] DEBUG:    Octet string at 0x80d1f78:
> 2003-07-09 12:46:39 [6] DEBUG:      len:  4
> 2003-07-09 12:46:39 [6] DEBUG:      size: 5
> 2003-07-09 12:46:39 [6] DEBUG:      immutable: 0
> 2003-07-09 12:46:39 [6] DEBUG:      data: d4 f2 9c 0e         
>       ....
> 2003-07-09 12:46:39 [6] DEBUG:    Octet string dump ends.
> 2003-07-09 12:46:39 [6] DEBUG: SMPP PDU dump ends.
> 
> 
> I had a look at the kannel source and I saw it converts from 
> gsm-7 to latin
> by default if no alt_charset is set.
> I'm just getting rubbish with this though..
> 
> The message is 'Test' and is 7-bit packed:
> 
> D4 1 1010100 = T
> F2 11 110010 = e
> 9C 100 11100 = s
> 0E 0000 1110 = t
> 
> 
> How do I convert it?
> 
> Regards,
> 
> David.
> 
> 
>