On Thu, 01 Jul 2010 18:08:47 -0500
Denis Kenzior <[email protected]> wrote:

> Hi Kristen,
> 
> >>>> Then this code is incorrect, as it handles len 3 attrs only at the end
> >>>> of the array.  Also please note that SIM Toolkit Text Attributes are
> >>>> always coded on 4 bytes (see 102.223 Section 8.72 for details).  You
> >>>> might want to assume 4 byte alignment or invent a data structure for
> >>>> these attributes.
> >>>
> >>> I should have said that attrs_len does not always have to be a multiple
> >>> of 4 -- it is allowed to only have 3 byte attrs as the last attribute,
> >>> so the code is correct.  However, since not all attributes are
> >>> required to be of len 4, we can't just assume attrs_len is a multiple
> >>> of 4.
> >>>
> >>
> >> How do you figure this?  From 23.040 Section 9.2.3.24.10.1.1       Text 
> >> Formatting
> >>
> >> "Octet 4
> >>
> >> This Octet may be omitted by setting the IED length accordingly."
> > 
> > This isn't the clearest language, so my interpretation of this is
> > you may omit one byte if you set the length properly.  Clearly
> > the spec tell you that the byte is optional, so I think it would
> > be wrong to assume that all are 4 bytes.  Also, since obviously
> > we can't support mixed 3 and 4 byte fields since we'd have no
> > way to know if the next byte was the start of the next attribute,
> > it's got to be at the end of the array.
> 
> Actually it is not.  Each Text attribute is put into an SMS header, with
> the header having an IEI (Information Element Identifier), Length and
> Data.  See SMS_IEI_TEXT_FORMAT in smsutil.h
> 
> Hence, when an EMS PDU is parsed, you know whether this is a 3 byte or a
> 4 byte element.
> 
> Regards,
> -Denis

The way this function is - we have no knowledge of the pdu.
Are you wanting to change the arguments to this function to include
an sms header?  or a pdu? Or are you saying the caller is responsible 
for padding out the text attribute into a 4 byte attribute?
_______________________________________________
ofono mailing list
[email protected]
http://lists.ofono.org/listinfo/ofono

Reply via email to