Re: Insertion of character inside a TCP frame

Thanks for the help, at this point I applied Christophe solution and it worked fine. But I know it's not a long term solution because UCS-2 SMS are sent in UTF-8 so I'll encounter some problems when I'll work on this part.

I'll be sure to keep the list updated on my progress if I reach this point, it's a personal project (there are very good SMPP servers and the world does need another one, there are just none in CL).

On Tuesday, July 16th, 2024 at 4:49 PM, Jérôme Radix <[email protected]> wrote:

> Hi Steph,
>
> here is a little strace output (strace.out) showing that :
> - you effectively have 22 bytes
> - (length answer) returns 21 because of character C2 which is more than 127 and this mess FORMAT and UTF-8 (see Christophe answer)
>
> To help you see the characters in your answer string, in rbind, replace in sbcl :
> (format t "~A (~A)~%" answer (length answer))
> by
> (format t "~A (~A)~%" (sb-ext:string-to-octets answer) (length (sb-ext:string-to-octets answer)))
>
> it prints :
> #(0 0 0 21 194 128 0 0 1 0 0 0 0 0 0 0 1 83 77 80 80 0) (22)
>
> I would be glad to see your final solution,
>
> Regards,
> Jérôme.
>
> Le mar. 16 juil. 2024 à 10:00, Marco Antoniotti <[email protected]> a écrit :
>
>> Hi Stephane
>>
>> you may want to have a look at babel and maybe cl-pack for encoding and decoding. I am sure there are others around and some of them may suffer from some bit rot.
>>
>> All the best
>>
>> Marco
>>
>> On Tue, Jul 16, 2024 at 9:49 AM Christophe Rhodes <[email protected]> wrote:
>>
>>> "steph.tougard via Sbcl-help" <[email protected]> writes:
>>>
>>>> For those who don't know, SMPP is a binary protocol over TCP to send
>>>> SMS. Unlike SMPP or HTTP, each frame is sent or received between the
>>>
>>> SMPP is a binary protocol, but your Lisp code uses characters and
>>> strings. This means that there will be an encoding step on the Lisp
>>> side, to convert from characters to binary.
>>>
>>> In most lisps nowadays, the default encoding scheme is probably utf-8,
>>> which encodes the ASCII repertoire (the first 128 characters of Unicode)
>>> as single bytes, but uses multiple bytes for all other characters.
>>>
>>> I would guess that the string that you are attempting to send contains a
>>> character with char-code greater than 127, and that your stream is set
>>> up to convert characters using utf-8.
>>>
>>> You are using usocket, which does not allow users to specify external
>>> formats directly. You could try setting
>>> sb-ext:*default-external-format* to :latin-1 before starting your
>>> server, or you could send binary (rather than string) data to your
>>> socket, performing the character to binary conversion yourself rather
>>> than making it implicit.
>>>
>>> Christophe
>>>
>>> _______________________________________________
>>> Sbcl-help mailing list
>>> [email protected]
>>> https://lists.sourceforge.net/lists/listinfo/sbcl-help
>>
>> --
>>
>> Marco Antoniotti, Professor tel. +39 - 02 64 48 79 01
>> DISCo, University of Milan-Bicocca U14 2043 http://dcb.disco.unimib.it
>> Viale Sarca 336
>> I-20126 Milan (MI) ITALY
>> _______________________________________________
>> Sbcl-help mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/sbcl-help

_______________________________________________
Sbcl-help mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sbcl-help