2 # Convert Packages.gz files into Sleepycat db files for efficient usage of
7 # Copyright (C) 2006 Jeroen van Wolffelaar <jeroen@wolffelaar.nl>
8 # This program is free software; you can redistribute it and/or modify
9 # it under the terms of the GNU General Public License as published by
10 # the Free Software Foundation; either version 2 of the License, or
11 # (at your option) any later version.
13 # This program is distributed in the hope that it will be useful,
14 # but WITHOUT ANY WARRANTY; without even the implied warranty of
15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 # GNU General Public License for more details.
18 # You should have received a copy of the GNU General Public License
19 # along with this program; if not, write to the Free Software
20 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
24 my $what = $ARGV[0] ? "non-free" : "*";
25 # max. distinct results for a given package postfix
26 my $MAX_PACKAGE_POSTFIXES = 100;
29 my %packages_small = ();
30 my %package_names = ();
31 my %package_postfixes = ();
32 my %sources_packages = ();
33 my %descriptions = ();
34 my @descriptions = ("we count lines one-based\000");
35 my %packages_descriptions = ();
36 my %descriptions_packages = ();
38 my @archives = ( 'us', 'non-US', 'security', 'volatile', 'backports' );
39 my @suites = ('oldstable', 'stable', 'testing', 'unstable', 'experimental');
43 for my $archive (@archives) {
44 for my $suite (@suites) {
46 print "Reading $archive/$suite...\n";
48 tie %packages_all_db, "DB_File", "packages_all_$suite.db.new",
49 O_RDWR|O_CREAT, 0666, $DB_BTREE
50 or die "Error creating DB: $!";
51 open PKG, "zcat /org/packages.debian.org/archive/$archive/$suite/$what/{,debian-installer/}binary-*/Packages.gz|";
58 while (/^(\S+):\s*(.*)\s*$/mg) {
59 my ($key, $value) = ($1, $2);
60 $value =~ s/\377/\n /g;
61 $key =~ tr [A-Z] [a-z];
65 next if exists($packages_all_db{"$data{'package'} $data{'architecture'} $data{'version'}"});
67 $package_names{$data{'package'}} = 1;
68 my $src = $data{'package'};
70 if ($data{'source'}) {
71 $src = $data{'source'};
73 if $src =~ s/\s+\(\s*=\s*(.*)\).*//; # strip version info
75 $data{'source'} = $src;
76 $data{'source-version'} = $src_version;
77 my $descr = $data{'description'};
79 if (exists($descriptions{$descr})) {
80 $did = $descriptions{$descr};
82 $did = 1 + $#descriptions;
83 $descriptions[$did] = $descr;
84 $descriptions{$descr} = $did;
86 $data{'description'} = $did;
87 $packages_descriptions{"$data{'package'} $data{'version'} $data{'architecture'}"} = $did;
88 $descriptions_packages{$did} .=
89 "$data{'package'} $data{'version'} $data{'architecture'}\000";
94 my $subsection = $data{section} || '-';
95 if ($data{section} && ($data{section} =~ m=/=o)) {
96 ($section, $subsection) = split m=/=o, $data{section}, 2;
98 $data{'section'} = $section;
99 $data{'subsection'} = $subsection;
100 $data{'priority'} ||= '-';
101 $packages_small{$data{'package'}} .= "$archive $suite $data{'architecture'} ".
102 "$section $subsection $data{'priority'} $data{'version'} $sdescr\000";
103 $sources_packages{$src} .=
104 "$archive $suite $data{'package'} $data{'version'} $data{'architecture'}\000";
105 $data{archive} = $archive;
106 while (my ($key, $value) = each (%data)) {
107 next if $key eq 'package' or $key eq 'version' or $key eq 'architecture'
108 or $key eq 'maintainer';
109 print STDERR "WARN: $key ($suite/$archive/$data{package}/$data{architecture}\n" unless defined $value;
110 $data .= "$key: $value\n";
112 $packages_all_db{"$data{'package'} $data{'architecture'} $data{'version'}"}
116 untie %packages_all_db;
120 print "Writing databases...\n";
121 my %packages_small_db;
122 tie %packages_small_db, "DB_File", "packages_small.db.new",
123 O_RDWR|O_CREAT, 0666, $DB_BTREE
124 or die "Error creating DB: $!";
125 while (my ($k, $v) = each(%packages_small)) {
127 $packages_small_db{$k} = $v;
129 untie %packages_small_db;
131 my %sources_packages_db;
132 tie %sources_packages_db, "DB_File", "sources_packages.db.new",
133 O_RDWR|O_CREAT, 0666, $DB_BTREE
134 or die "Error creating DB: $!";
135 while (my ($k, $v) = each(%sources_packages)) {
137 $sources_packages_db{$k} = $v;
139 untie %sources_packages_db;
141 my %packages_descriptions_db;
142 tie %packages_descriptions_db, "DB_File", "packages_descriptions.db.new",
143 O_RDWR|O_CREAT, 0666, $DB_BTREE
144 or die "Error creating DB: $!";
145 while (my ($k, $v) = each(%packages_descriptions)) {
146 $packages_descriptions_db{$k} = $v;
148 untie %packages_descriptions_db;
150 my %descriptions_packages_db;
151 tie %descriptions_packages_db, "DB_File", "descriptions_packages.db.new",
152 O_RDWR|O_CREAT, 0666, $DB_BTREE
153 or die "Error creating DB: $!";
154 while (my ($k, $v) = each(%descriptions_packages)) {
156 $descriptions_packages_db{$k} = $v;
158 untie %descriptions_packages_db;
161 tie %descriptions_db, "DB_File", "descriptions.db.new",
162 O_RDWR|O_CREAT, 0666, $DB_BTREE
163 or die "Error creating DB: $!";
164 open DESCR, "> descriptions.txt" or die "Error creating descriptions textfile";
165 for (my $i=1; $i<= $#descriptions; $i++) {
166 my $plain_description = $descriptions[$i];
167 $plain_description =~ s/\n .\n/ /og;
168 $plain_description =~ s/[\n \t]+/ /og;
169 print DESCR "$plain_description\n";
170 $descriptions_db{$i} = $descriptions[$i];
173 untie %descriptions_db;
175 # package names stuff:
176 for my $pkg (keys %package_names) {
177 for (my $i=0;$i<length($pkg)-1;$i++) {
178 my $before = substr($pkg, 0, $i);
179 my $after = substr($pkg, $i);
180 $before = "^" if $before eq ""; # otherwise split doesn't work properly
181 $package_postfixes{$after} .= "$before\0";
184 my %package_postfixes_db;
185 tie %package_postfixes_db, "DB_File", "package_postfixes.db.new",
186 O_RDWR|O_CREAT, 0666, $DB_BTREE
187 or die "Error creating DB: $!";
188 while (my ($k, $v) = each(%package_postfixes)) {
192 $nr = length($nr) + 1; # < number of hits
193 if ($nr > $MAX_PACKAGE_POSTFIXES) {
196 $package_postfixes_db{$k} = $v;
198 untie %package_postfixes_db;
200 rename("packages_small.db.new", "packages_small.db");
201 rename("sources_packages.db.new", "sources_packages.db");
202 for my $suite (@suites) {
203 rename("packages_all_$suite.db.new", "packages_all_$suite.db");
205 rename("packages_descriptions.db.new", "packages_descriptions.db");
206 rename("descriptions_packages.db.new", "descriptions_packages.db");
207 rename("descriptions.txt.new", "descriptions.txt");
208 rename("descriptions.db.new", "descriptions.db");
209 rename("package_postfixes.db.new", "package_postfixes.db");