revised extended autnum patch, was: Re: Bit of tweaking to IRRToolSet web site
spz at serpens.de
Tue Feb 24 08:42:50 UTC 2009
Thus wrote Nick Hilliard (nick at inex.ie):
> On 15/02/2009 10:52, S.P.Zeidler wrote:
> I'm running into trouble with this patch. First, rpsl_asnum.hh has been
> omitted again. I've used the one attached to ticket #9, added "extern
*sigh* yes, I can't svn add so svn diff will ignore it. I'm too used to
actually have access to the repository I'm working with. :-7
> bool opt_asdot;" to the top, put in a check for opt_asdot in the "if"
> statement, and changed rpsl/symbols.hh to include this header. This
> makes the code compile. However, it's still not recognising asn32s
>> pancake:/home/nick/tmp/IRRToolSet-4.8.5/src/peval> ./peval -asplain AS3.3
>> peval: AS3.3
>> ***Error: badly formed filter.
I assume you don't use gmake, and thus rpsl.y.cc etc don't get
regenerated for you. Diffing for these is a right bastard since the
difference in generator version drops 3/4 of the files into the diff,
even without a functional difference, that's why they are not included.
bison -y -d -p `basename rpsl.y .y` rpsl.y
mv y.tab.c `basename rpsl.y`.cc
mv y.tab.h `basename rpsl.y`.hh
flex -P`basename rpsl.l .l` rpsl.l
mv lex.`basename rpsl.l .l`.c `basename rpsl.l`.cc
But I noticed I'll need to go over it again anyway, since I seem to have
been 'asplaining' a bit too enthusiastically:
113% peval AS3.3
114% peval -asplain AS3.3
115% peval -no-as -asplain AS3.3
116% whois3 -h whois.radb.net AS196611
% No entries found for the selected source(s).
This is going to pose an interesting problem if/when the registries start
to mix asdot and asplain, I guess irrtoolset ought to try and match both.
Fixing this will take at least until after next weekend, maybe longer.
spz at serpens.de (S.P.Zeidler)
More information about the irrtoolset