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
27 my $what = $ARGV[0] ? "non-free" : "*";
28 # max. distinct results for a given package postfix
29 my $MAX_PACKAGE_POSTFIXES = 100;
34 use Packages::Config qw( $TOPDIR $DBDIR @ARCHIVES @SUITES );
35 &Packages::Config::init( './' );
36 my %packages_small = ();
37 my %virtual_packages = ();
38 my %package_names = ();
39 my %package_postfixes = ();
40 my %sources_packages = ();
41 my %descriptions = ();
42 my @descriptions = ("we count lines one-based\000");
43 my %packages_descriptions = ();
44 my %descriptions_packages = ();
52 for my $archive (@ARCHIVES) {
53 for my $suite (@SUITES) {
55 print "Reading $archive/$suite...\n";
57 tie %packages_all_db, "DB_File", "$DBDIR/packages_all_$suite.db.new",
58 O_RDWR|O_CREAT, 0666, $DB_BTREE
59 or die "Error creating DB: $!";
60 open PKG, "zcat $TOPDIR/archive/$archive/$suite/$what/{,debian-installer/}binary-*/Packages.gz|";
67 while (/^(\S+):\s*(.*)\s*$/mg) {
68 my ($key, $value) = ($1, $2);
69 $value =~ s/\377/\n /g;
70 $key =~ tr [A-Z] [a-z];
74 next if exists($packages_all_db{"$data{'package'} $data{'architecture'} $data{'version'}"});
76 if ($data{'provides'}) {
77 foreach (split /\s*,\s*/, $data{'provides'}) {
78 $virtual_packages{$_}{$suite}{$data{'package'}}++;
81 $package_names{$data{'package'}} = 1;
82 my $src = $data{'package'};
84 if ($data{'source'}) {
85 $src = $data{'source'};
86 $src =~ s/\s+.*//; # strip version info
88 $data{'source'} = $src;
89 my $descr = $data{'description'};
91 if (exists($descriptions{$descr})) {
92 $did = $descriptions{$descr};
94 $did = 1 + $#descriptions;
95 $descriptions[$did] = $descr;
96 $descriptions{$descr} = $did;
98 $data{'description'} = $did;
99 $packages_descriptions{"$data{'package'} $data{'version'} $data{'architecture'}"} = $did;
100 $descriptions_packages{$did} .=
101 "$data{'package'} $data{'version'} $data{'architecture'}\000";
104 $sdescr =~ s/\n.*//s;
105 my $section = 'main';
106 my $subsection = $data{section} || '-';
107 if ($data{section} && ($data{section} =~ m=/=o)) {
108 ($section, $subsection) = split m=/=o, $data{section}, 2;
109 ($subsection, $section) = split m=/=o, $data{section}, 2
110 if $section eq 'non-US';
112 $data{'section'} = $section;
113 $data{'subsection'} = $subsection;
114 $data{'priority'} ||= '-';
115 $sections{$suite}{$archive}{$section}++;
116 $subsections{$suite}{$archive}{$subsection}++;
117 $priorities{$suite}{$archive}{$data{priority}}++;
118 my $pkgitem = "$archive $suite $data{'architecture'} ".
119 "$section $subsection $data{'priority'} $data{'version'} $sdescr\0";
120 my $previtem = $packages_small{$data{'package'}}{$suite}{$data{'architecture'}}
122 $packages_small{$data{'package'}}{$suite}{$data{'architecture'}} = $pkgitem
123 if version_cmp($data{'version'}, (split /\s/o, $previtem)[6]) >= 0;
124 $previtem = $packages_small{$data{'package'}}{$suite}{'any'}
126 $packages_small{$data{'package'}}{$suite}{'any'} = $pkgitem
127 if version_cmp($data{'version'}, (split /\s/o, $previtem)[6]) >= 0;
128 $sources_packages{$src} .=
129 "$archive $suite $data{'package'} $data{'version'} $data{'architecture'}\000";
130 $data{archive} = $archive;
131 while (my ($key, $value) = each (%data)) {
132 next if $key eq 'package' or $key eq 'version' or $key eq 'architecture'
133 or $key eq 'maintainer';
134 print STDERR "WARN: $key ($suite/$archive/$data{package}/$data{architecture}\n" unless defined $value;
135 $data .= "$key\00$value\00";
138 $packages_all_db{"$data{'package'} $data{'architecture'} $data{'version'}"}
142 untie %packages_all_db;
146 print "Writing databases...\n";
147 my %packages_small_db;
148 tie %packages_small_db, "DB_File", "$DBDIR/packages_small.db.new",
149 O_RDWR|O_CREAT, 0666, $DB_BTREE
150 or die "Error creating DB: $!";
151 while (my ($pkg, $v) = each(%packages_small)) {
152 my ($res1, $res2, $res3) = ("", "", "");
153 while (my ($suite, $v2) = each %$v) {
154 $res2 .= $v2->{'any'};
155 while (my ($arch, $v3) = each %$v2) {
156 next if $arch eq 'any' or $v3 eq $v2->{'any'};
161 if (exists $virtual_packages{$pkg}) {
162 while (my ($suite, $v2) = each %{$virtual_packages{$pkg}}) {
163 $res1 .= "$suite\01".(join ' ', keys %$v2)."\01";
170 my $res = "$res1$res2$res3";
172 $packages_small_db{$pkg} = $res;
174 untie %packages_small_db;
176 my %sources_packages_db;
177 tie %sources_packages_db, "DB_File", "$DBDIR/sources_packages.db.new",
178 O_RDWR|O_CREAT, 0666, $DB_BTREE
179 or die "Error creating DB: $!";
180 while (my ($k, $v) = each(%sources_packages)) {
182 $sources_packages_db{$k} = $v;
184 untie %sources_packages_db;
186 my %packages_descriptions_db;
187 tie %packages_descriptions_db, "DB_File", "$DBDIR/packages_descriptions.db.new",
188 O_RDWR|O_CREAT, 0666, $DB_BTREE
189 or die "Error creating DB: $!";
190 while (my ($k, $v) = each(%packages_descriptions)) {
191 $packages_descriptions_db{$k} = $v;
193 untie %packages_descriptions_db;
195 my %descriptions_packages_db;
196 tie %descriptions_packages_db, "DB_File", "$DBDIR/descriptions_packages.db.new",
197 O_RDWR|O_CREAT, 0666, $DB_BTREE
198 or die "Error creating DB: $!";
199 while (my ($k, $v) = each(%descriptions_packages)) {
201 $descriptions_packages_db{$k} = $v;
203 untie %descriptions_packages_db;
206 tie %descriptions_db, "DB_File", "$DBDIR/descriptions.db.new",
207 O_RDWR|O_CREAT, 0666, $DB_BTREE
208 or die "Error creating DB: $!";
209 open DESCR, ">", "$DBDIR/descriptions.txt" or die "Error creating descriptions textfile";
210 for (my $i=1; $i<= $#descriptions; $i++) {
211 my $plain_description = $descriptions[$i];
212 # WARNING: This needs to correspond with what happens in
213 # Packages/Search.pm:do_fulltext_search
214 $plain_description =~ tr [A-Z] [a-z];
215 # ensure one space on both ends
216 $plain_description = " $plain_description ";
217 $plain_description =~ s/[(),.-]+//og;
218 $plain_description =~ s#[^a-z0-9_/+]+# #og;
219 print DESCR "$plain_description\n";
220 $descriptions_db{$i} = $descriptions[$i];
223 untie %descriptions_db;
225 # package names stuff:
226 for my $pkg (keys %package_names) {
227 for (my $i=0;$i<length($pkg)-1;$i++) {
228 my $before = substr($pkg, 0, $i);
229 my $after = substr($pkg, $i);
230 $before = "^" if $before eq ""; # otherwise split doesn't work properly
231 $package_postfixes{$after} .= "$before\0";
234 my %package_postfixes_db;
235 tie %package_postfixes_db, "DB_File", "$DBDIR/package_postfixes.db.new",
236 O_RDWR|O_CREAT, 0666, $DB_BTREE
237 or die "Error creating DB: $!";
238 while (my ($k, $v) = each(%package_postfixes)) {
242 $nr = length($nr) + 1; # < number of hits
243 if ($nr > $MAX_PACKAGE_POSTFIXES) {
246 $package_postfixes_db{$k} = $v;
248 untie %package_postfixes_db;
250 store \%sections, "$DBDIR/sections.info";
251 store \%subsections, "$DBDIR/subsections.info";
252 store \%priorities, "$DBDIR/priorities.info";
254 rename("$DBDIR/packages_small.db.new", "$DBDIR/packages_small.db");
255 rename("$DBDIR/sources_packages.db.new", "$DBDIR/sources_packages.db");
256 for my $suite (@SUITES) {
257 rename("$DBDIR/packages_all_$suite.db.new",
258 "$DBDIR/packages_all_$suite.db");
260 rename("$DBDIR/packages_descriptions.db.new",
261 "$DBDIR/packages_descriptions.db");
262 rename("$DBDIR/descriptions_packages.db.new",
263 "$DBDIR/descriptions_packages.db");
264 rename("$DBDIR/descriptions.txt.new", "$DBDIR/descriptions.txt");
265 rename("$DBDIR/descriptions.db.new", "$DBDIR/descriptions.db");
266 rename("$DBDIR/package_postfixes.db.new", "$DBDIR/package_postfixes.db");