Chris
This is great information, it should help. Yes the 703 number is the phone
number. The reason why I can't just create an IPD file using the simulator
is because when a user erases a message, it does not erase from the
blackberry. I'll send you a new image demonstrating what I'm talking about
tomorrow. Basically if the message is under a few words and has been
deleted, you can still see in clear txt looking at the dmp file. The
problem comes in where you have an sms message or even an email that has
been compressed [because of its length] and deleted by the user. Creating
a new IPD file wont recover the deleted messages even if you can see that
they're still in the dmp file.
I figured out some of the indicators of an SMS message, however I haven't
yet figured out what bytes flag a message as deleted. If I can figure this
out, I can simply change the flag from deleted to active and have the
simulator do the decompression of the message.
On Thu, Feb 25, 2010 at 4:55 PM, Chris Frey <cdf...@foursquare.net> wrote:
> On Wed, Feb 24, 2010 at 10:50:01AM -0500, Jamaal Speights wrote:
> > I was also speaking of getting help with the Barry code :) I will email
> > you a zip file containing an example compressed SMS message and an
> > uncompressed message. The zip file is the 8330-fs.dmp file from the
> > simulator. Messages under 6 words are uncompressed. Over that they are
> > compressed.
>
> Hi Jamaal,
>
> I looked at the dump file you sent, and I'm assuming that the 703 number
> is a phone number?
>
> The format of the raw message is different than the USB wire format.
> A snippet from your data:
>
> 01d48680 29 80 77 48 32 00 01 07 05 20 15 48 09 00 84 d0 |).wH2....
> .H....|
> 01d48690 54 65 73 74 20 74 65 73 74 00 00 00 03 00 15 48 |Test
> test......H|
>
> It almost looks like there is a length field (09 00) then a code (84 d0)
> and then the string ("Test test").
>
> But these codes are different from what Barry expects. In a sample SMS
> message:
>
> 00000060 XX XX XX XX XX XX XX XX 00 1a 00 04 4a 75 73 74
> |XXXXXXXX....Just|
> 00000070 20 73 6f 6d 65 74 68 69 6e 67 20 66 6f 72 20 64 | something for
> d|
> 00000080 65 73 73 65 72 74 0f 00 07 00 04 00 00 00 00 00
> |essert..........|
>
> The size is there (1a 00) and the field code (04) and then the string
> ("Just something for dessert").
>
> The Barry code expects these codes, and won't work with data that uses
> different codes.
>
> Now, you might be able to just change the codes, but this will be some
> reverse engineering work and testing work on your own. See the file
> src/r_sms.cc:
>
> // SMS Field Codes
> #define SMSFC_METADATA 0x01
> #define SMSFC_ADDRESS 0x02
> #define SMSFC_BODY 0x04
>
> The code expects single byte codes, which you may be able to get away with,
> if you ignore the d0 in your data. (The d0 seems static in the data I
> saw).
>
> There was a different code used for a compressed field (86 d0) vs. an
> uncompressed field (84 d0). I'm not sure what kind of compression
> is used. None of the USB fields were compressed, that I've seen.
>
>
> > The data in the image file i am sending is different from the IPD file.
> > During the backup process, the data on the blackberry is uncompressed and
> > place into the IDP file as a string.
>
> Are you able to generate an IPD file from your dump image? This might
> be an easier route if possible.
>
> - Chris
>
>
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Barry-devel mailing list
> Barry-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/barry-devel
>
------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Barry-devel mailing list
Barry-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/barry-devel
