[PATCH] uqmi: corrected too short received SMS

[Daniel Golle]

Hi Henrik,

thank you for submitting this patch. I've also noticed that problem
long ago but it wasn't important for me at the time, so I didn't go
into fixing it.

On Sat, Mar 12, 2022 at 12:33:54AM +0100, Henrik Ginstmark wrote:
> When characters with ascii values bigger than 0x7f are used, the
> length of the received text
> message is too short.
> 
> Test message sent: 123äÄ123
> Before correction:
> root at OpenWrt:/tmp# uqmi -d /dev/cdc-wdm0 --get-message 20
> Raw text: 31 32 33 7b 5b 31 32 33
> {
>         "smsc": "+46724400001",
>         "sender": "+46xxxxxxxxx",
>         "timestamp": "2022-03-11 18:48:10",
>         "text": "123äÄ1"
> }
> 
> after correction:
> root at OpenWrt:/tmp# uqmi -d /dev/cdc-wdm0 --get-message 20
> Raw text: 31 32 33 7b 5b 31 32 33
> {
>         "smsc": "+46724400001",
>         "sender": "+46xxxxxxxxx",
>         "timestamp": "2022-03-11 18:48:10",
>         "text": "123äÄ123"
> }
> 
> Signed-off-by: Henrik Ginstmark <henrik at ginstmark.se>
> ---
>  uqmi/src/commands-wms.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/uqmi/src/commands-wms.c b/uqmi/src/commands-wms.c
> index 700d79f..a58fd6a 100644
> --- a/uqmi/src/commands-wms.c
> +++ b/uqmi/src/commands-wms.c
> @@ -222,8 +222,8 @@ static int decode_udh(const unsigned char *data)
>  static void decode_7bit_field(char *name, const unsigned char *data,
> int data_len, int bit_offset)

I've fixed the line-wrappign problems here ...

>  {
>         char *dest = blobmsg_alloc_string_buffer(&status, name, 3 *
> data_len + 2);

... and here ...
> -       pdu_decode_7bit_str(dest, data, CEILDIV(data_len * 7, 8), bit_offset);
> -       dest[data_len] = 0;
> +       int out_len = pdu_decode_7bit_str(dest, data, CEILDIV(data_len
> * 7, 8), bit_offset);

... and here, so the patch would apply.

> +       dest[out_len] = 0;
>         blobmsg_add_string_buffer(&status);
>  }
> 
> --
> 2.34.1

Please use 'git format-send-email' or 'git format-patch' next time to
avoid the MUA messing around with line-wrapping.
For this submission it's ok, I've already fixed it manually.


Cheers


Daniel