[aprssig] Protocol parsing questions

[Robert Bruninga]

There are two documents written to help you parse APRS:
www.aprs.org/txt/APRS-parser-default.txt
www.aprs.org/txt/APRS-parser.txt 

The first one makes sure that you at least estimate a vicinity
plot on a new callsign no matter what his packet, so that he
exists within your system with the best guess position until you
subsequently get a valid posit.

The second document above shows you the process for working your
way through the parser.  If a packet passes all the way through
the parser and does not match any formats, then put it in the
BAD-PACKET list.  But if it is the only packet you have received
from a station, then put it in the STATUS list as the best info
you have from that station.

Bob, WB4APR

> -----Original Message-----
> I'm getting back into APRS after a few years (oh, about 15, I 
> think).  I'm in the process of writing a ground-up Java 
> parser, as my eventual goal is to run a native application on 
> the Android G1 that uses it as a position reporting device.
> 
> I've run across a few packets that I can't determine the 
> contents of based on the 1.0.1 and 1.1 specifications.  For 
> example, the packet below:
> 
> 
> CCDES-1>APRSWX,WIDE1-1,WIDE2-2,qAR,W3AHL-4:@200150z3957.34N/07
> 535.07W_108/000g000t029r000p000P000h48b10170L000gt41.93vw0.243
bv12.78fr000.C100
> 
> 
> We have a 'g000' that says what the gust is.  Then we have a 
> 'gt41...'  I have no idea what that is.  There's also the 
> question of when to stop parsing, and treat the remainder as 
> a comment.  In the 1.0.1 PDF file, a number of the example 
> weather packets are shown with 'wRSW'.  'w' should be the 
> software version, but it's not define in the list that 
> follows.  Is there a known sigil for deciding the remainder 
> of the packet are comments?
> 
> I also see a number of improperly formatted packets hitting 
> aprs-is.  Like this one:
> 
> 
> "WN0X-2>AP1WWX,KA0GFC-5,WIDE1,KB0WZI-3*,WIDE2-1,qAR,AC0JK-5:@2
00812z3837.76N/09220.63W_014/00
> 
> 
> Note the incomplete speed field.  What's the appropriate 
> course of action?  Discard it?
> 
> Then there are these packets:
> 
> 
> N1HQ>APS227,TCPIP*,qAC,THIRD:N/
> N1HQ>APS227,TCPIP*,qAC,THIRD:08
> N1HQ>APS227,TCPIP*,qAC,THIRD:54
> N1HQ>APS227,TCPIP*,qAC,THIRD:0.
> 
> 
> I'm guessing these might be a QSO or something, but I don't 
> understand how to programmatically identify them.  
> 
> Can someone point me to some documentation i might be 
> overlooking?  Or provide some insight from experience?
> 
> Thanks,
> --jc (K4JCW)
> 
>