[Popcon-commits] cvs commit to popularity-contest by pere

[popcon-commits at lists.alioth.debian.org]

Update of /cvsroot/popcon/popularity-contest
In directory haydn:/tmp/cvs-serv9262

Added Files:
	clean-genpkglist clean-filter 
Log Message:
Add scripts to clean out non-debian packages from the entries, to
make it possible to publish anonymous raw data.  Based on code from
Alain Schroeder.


--- NEW FILE: clean-genpkglist ---
#!/bin/sh
#
# Author: Alain Schroeder
# Date:   2005-07-09
# Modified by Petter Reinholdtsen 2005-07-15

tempname=`tempfile`
temp2=`tempfile`
resultfile=packages

archs="alpha arm hppa i386 ia64 m68k mips mipsel powerpc s390 sparc"
mirror="http://ftp.fi.debian.org/debian"

for i in $archs; do
        wget "$mirror/dists/Debian3.1r0/main/binary-${i}/Packages.gz" -O "$tempname"
        zcat $tempname | grep "^Package: " | sed -e 's/^Package: //' >> $temp2
done

sort -u $temp2 > $resultfile

rm $tempname $temp2;

--- NEW FILE: clean-filter ---
#!/usr/bin/perl
#
# Author: Alain Schroeder
# Date:   2005-07-09
# Modified by Petter Reinholdtsen 2005-07-15


use strict;

my %table;
my $foundentry = 0;
my $ignorecount = 0;
my $debug = 0;

open (PACKAGES, '< packages') or
    die ("packages not found. please run get-packages.sh first\n");

while (<PACKAGES>) {
    chomp;
    $table{$_} = "1";
}
close PACKAGES;

while (my $line = <STDIN>) {
    
    if ($line =~ m/^POPULARITY-CONTEST-0/i) {
	$foundentry = 1;
	$ignorecount = 0;
    }
    if ($line =~ m/^END-POPULARITY-CONTEST-0/i) {
	$foundentry = 0;
	print "# Ignored $ignorecount entries\n";
    } if ($line =~ m/^\d* \d* (\S*).*/i && !exists $table{$1}) {
	print STDERR "Ignoring package $1\n" if $debug;
	$ignorecount++;
	next;
    }
    print $line;
}