[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: checks/fields rewrite



Frank Lichtenheld <djpig@debian.org> writes:
> On Sun, Mar 28, 2004 at 09:42:44PM +0200, Marc 'HE' Brockschmidt wrote:
>> Frank Lichtenheld <djpig@debian.org> writes:
>> > On Sun, Mar 28, 2004 at 11:00:46AM +0200, Marc 'HE' Brockschmidt wrote:
>> >> Frank Lichtenheld <djpig@debian.org> writes:
>> >> > @@ -202,6 +212,7 @@
>> >> >  
>> >> >  			my (@seen_libstdcs, @seen_tcls, @seen_tclxs, @seen_tks, @seen_tkxs, @seen_libpngs);
>> >> >  
>> >> > +#FIXME: this check is fully broken, or better: where is the check?
>> >> Don't understand what you mean. The @seen_* arrays are used *after* the
>> >> loop.
>> > I actually meant the line below the comment:
>> >  print "E: $pkg $type: alternates-not-allowed $field\n"
>> >      if (! grep { $_ eq $field } qw(depends pre-depends recommends suggests));
>> 
>> Oh, i had that one fixed in the second version I've sent to the list.
> Ahh, seen it.
>
> Checked the new version.
> Some remaining issues:
>
>  1) While populating the @seen_* arrays you only honor the first
>  alternative, e.g.:
> push @seen_libstdcs, $alternatives[0]->[0] if defined $known_libstdcs{$alternatives[0]->[0]};
>
>  Why is that?

This i copied from the old check, but i think it's because you can only
substitute the packages checked for with other version of these
packages. A group of alternatives starting with one of these packages
has to be something like tcl8.2 | tcl8.3, which is no problem. We should
only report cases where the dependency looks like tcl8.2, [...] tcl8.3
or so.

>  2) In the dependency checking some tests do not make sense for all
>  relations. Partly you have honored that but it is missing for at least:
>
>   - needlessly-depends-on-awk (only check for (pre-)dep , rec, sug)
>     (btw. you seem to miss a ! here

Arrr, fixed.

>   - depends-on-libdb1-compat (dito)
>   - doc-package-depends-on-main-package (dito)
>   - depends-on-essential-package-without-using-version (dito)

Fixed.

>  3) The binary depends check misses the check
>
>   - package-has-a-duplicate-relation

I dumped this as i didn't understand the code. After looking through the
list of current problems in the archive reported by this check, i think
one can replace this with a far simpler solution.

>  4) The source depends check misses the checks
>
>   - depends-on-essential-package-without-using-version
>   - bad-relation

Added.

Marc
-- 
$_=')(hBCdzVnS})3..0}_$;//::niam/s~=)]3[))_$(rellac(=_$({pam(esrever })e$.)4/3*
)e$(htgnel+23(rhc,"u"(kcapnu ,""nioj ;|_- |/+9-0z-aZ-A|rt~=e$;_$=e${pam tnirp{y
V2ajFGabus} yV2ajFGa&{gwmclBHIbus}gwmclBHI&{yVGa09mbbus}yVGa09mb&{hBCdzVnSbus';
s/\n//g;s/bus/\nbus/g;eval scalar reverse   # <mailto:marc@marcbrockschmidt.de>




Reply to: