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
28 my $what = $ARGV[0] ? "non-free" : "*";
29 # max. distinct results for a given package postfix
30 my $MAX_PACKAGE_POSTFIXES = 100;
35 use Packages::Config qw( $TOPDIR $DBDIR @ARCHIVES @SUITES );
36 &Packages::Config::init( './' );
37 my %packages_small = ();
38 my %virtual_packages = ();
39 my %package_names = ();
40 my %package_postfixes = ();
41 my %sources_packages = ();
42 my %descriptions = ();
43 my @descriptions = ("we count lines one-based\000");
44 my %packages_descriptions = ();
45 my %descriptions_packages = ();
53 for my $suite (@SUITES) {
54 my %package_names_suite = ();
56 tie %packages_all_db, "DB_File", "$DBDIR/packages_all_$suite.db.new",
57 O_RDWR|O_CREAT, 0666, $DB_BTREE
58 or die "Error creating DB: $!";
60 for my $archive (@ARCHIVES) {
61 print "Reading $archive/$suite...\n";
62 open PKG, "zcat $TOPDIR/archive/$archive/$suite/$what/{,debian-installer/}binary-*/Packages.gz|";
69 while (/^(\S+):\s*(.*)\s*$/mg) {
70 my ($key, $value) = ($1, $2);
71 $value =~ s/\377/\n /g;
72 $key =~ tr [A-Z] [a-z];
76 next if exists($packages_all_db{"$data{'package'} $data{'architecture'} $data{'version'}"});
77 # Skip arch:all for amd64&kfreebsd, too often broken
78 next if ($archive eq 'amd64' or $archive eq 'kfreebsd')
79 and $data{architecture} eq 'all';
81 if ($data{'provides'}) {
82 foreach (split /\s*,\s*/, $data{'provides'}) {
83 $virtual_packages{$_}{$suite}{$data{'package'}}++;
84 $packages_small{$_} ||= {};
87 $package_names{$data{'package'}} = 1;
88 $package_names_suite{$data{'package'}} = 1;
89 my $src = $data{'package'};
91 if ($data{'source'}) {
92 $src = $data{'source'};
93 $src =~ s/\s+.*//; # strip version info
95 $data{'source'} = $src;
96 my $descr = $data{'description'};
98 if (exists($descriptions{$descr})) {
99 $did = $descriptions{$descr};
101 $did = 1 + $#descriptions;
102 $descriptions[$did] = $descr;
103 $descriptions{$descr} = $did;
105 $data{'description'} = $did;
106 $packages_descriptions{"$data{'package'} $data{'version'} $data{'architecture'}"} = $did;
107 $descriptions_packages{$did} .=
108 "$data{'package'} $data{'version'} $data{'architecture'}\000";
111 $sdescr =~ s/\n.*//s;
112 my $section = 'main';
113 my $subsection = $data{section} || '-';
114 if ($data{section} && ($data{section} =~ m=/=o)) {
115 ($section, $subsection) = split m=/=o, $data{section}, 2;
116 ($subsection, $section) = split m=/=o, $data{section}, 2
117 if $section eq 'non-US';
119 $data{'section'} = $section;
120 $data{'subsection'} = $subsection;
121 $data{'priority'} ||= '-';
122 $sections{$suite}{$section}++;
123 $subsections{$suite}{$subsection}++;
124 $priorities{$suite}{$data{priority}}++;
125 my $pkgitem = "$archive $suite $data{'architecture'} ".
126 "$section $subsection $data{'priority'} $data{'version'} $sdescr\0";
127 my $previtem = ($packages_small{$data{'package'}}{$suite}{$data{'architecture'}}
129 $packages_small{$data{'package'}}{$suite}{$data{'architecture'}} = $pkgitem
130 if version_cmp($data{'version'}, (split /\s/o, $previtem)[6]) > 0;
131 $previtem = ($packages_small{$data{'package'}}{$suite}{'any'}
133 $packages_small{$data{'package'}}{$suite}{'any'} = $pkgitem
134 if version_cmp($data{'version'}, (split /\s/o, $previtem)[6]) > 0;
135 $sources_packages{$src} .=
136 "$archive $suite $data{'package'} $data{'version'} $data{'architecture'}\000";
137 $data{archive} = $archive;
138 while (my ($key, $value) = each (%data)) {
139 next if $key eq 'package' or $key eq 'version' or $key eq 'architecture'
140 or $key eq 'maintainer';
141 print STDERR "WARN: $key ($suite/$archive/$data{package}/$data{architecture}\n" unless defined $value;
142 $data .= "$key\00$value\00";
145 $packages_all_db{"$data{'package'} $data{'architecture'} $data{'version'}"}
150 open NAMES, '>', "$DBDIR/package_names_$suite.txt.new"
151 or die "Error creating package names list: $!";
152 foreach (sort keys %package_names_suite) {
157 untie %packages_all_db;
160 print "Writing databases...\n";
161 my %packages_small_db;
162 tie %packages_small_db, "DB_File", "$DBDIR/packages_small.db.new",
163 O_RDWR|O_CREAT, 0666, $DB_BTREE
164 or die "Error creating DB: $!";
165 while (my ($pkg, $v) = each(%packages_small)) {
166 my ($res1, $res2, $res3) = ("", "", "");
167 while (my ($suite, $v2) = each %$v) {
168 $res2 .= $v2->{'any'};
169 while (my ($arch, $v3) = each %$v2) {
170 next if $arch eq 'any' or $v3 eq $v2->{'any'};
175 if (exists $virtual_packages{$pkg}) {
176 while (my ($suite, $v2) = each %{$virtual_packages{$pkg}}) {
177 $res1 .= "$suite\01".(join ' ', keys %$v2)."\01";
184 my $res = "$res1$res2$res3";
186 $packages_small_db{$pkg} = $res;
188 untie %packages_small_db;
190 my %sources_packages_db;
191 tie %sources_packages_db, "DB_File", "$DBDIR/sources_packages.db.new",
192 O_RDWR|O_CREAT, 0666, $DB_BTREE
193 or die "Error creating DB: $!";
194 while (my ($k, $v) = each(%sources_packages)) {
196 $sources_packages_db{$k} = $v;
198 untie %sources_packages_db;
200 my %packages_descriptions_db;
201 tie %packages_descriptions_db, "DB_File", "$DBDIR/packages_descriptions.db.new",
202 O_RDWR|O_CREAT, 0666, $DB_BTREE
203 or die "Error creating DB: $!";
204 while (my ($k, $v) = each(%packages_descriptions)) {
205 $packages_descriptions_db{$k} = $v;
207 untie %packages_descriptions_db;
209 my %descriptions_packages_db;
210 tie %descriptions_packages_db, "DB_File", "$DBDIR/descriptions_packages.db.new",
211 O_RDWR|O_CREAT, 0666, $DB_BTREE
212 or die "Error creating DB: $!";
213 while (my ($k, $v) = each(%descriptions_packages)) {
215 $descriptions_packages_db{$k} = $v;
217 untie %descriptions_packages_db;
220 tie %descriptions_db, "DB_File", "$DBDIR/descriptions.db.new",
221 O_RDWR|O_CREAT, 0666, $DB_BTREE
222 or die "Error creating DB: $!";
223 open DESCR, ">", "$DBDIR/descriptions.txt" or die "Error creating descriptions textfile";
224 for (my $i=1; $i<= $#descriptions; $i++) {
225 my $plain_description = $descriptions[$i];
226 # WARNING: This needs to correspond with what happens in
227 # Packages/Search.pm:do_fulltext_search
228 $plain_description =~ tr [A-Z] [a-z];
229 # ensure one space on both ends
230 $plain_description = " $plain_description ";
231 $plain_description =~ s/[(),.-]+//og;
232 $plain_description =~ s#[^a-z0-9_/+]+# #og;
233 print DESCR "$plain_description\n";
234 $descriptions_db{$i} = $descriptions[$i];
237 untie %descriptions_db;
239 # package names stuff:
240 for my $pkg (keys %package_names) {
241 for (my $i=0;$i<length($pkg)-1;$i++) {
242 my $before = substr($pkg, 0, $i);
243 my $after = substr($pkg, $i);
244 $before = "^" if $before eq ""; # otherwise split doesn't work properly
245 $package_postfixes{$after} .= "$before\0";
248 my %package_postfixes_db;
249 tie %package_postfixes_db, "DB_File", "$DBDIR/package_postfixes.db.new",
250 O_RDWR|O_CREAT, 0666, $DB_BTREE
251 or die "Error creating DB: $!";
252 while (my ($k, $v) = each(%package_postfixes)) {
256 $nr = length($nr) + 1; # < number of hits
257 if ($nr > $MAX_PACKAGE_POSTFIXES) {
260 $package_postfixes_db{$k} = $v;
262 untie %package_postfixes_db;
264 store \%sections, "$DBDIR/sections.info";
265 store \%subsections, "$DBDIR/subsections.info";
266 store \%priorities, "$DBDIR/priorities.info";
268 rename("$DBDIR/packages_small.db.new", "$DBDIR/packages_small.db");
269 rename("$DBDIR/sources_packages.db.new", "$DBDIR/sources_packages.db");
270 for my $suite (@SUITES) {
271 rename("$DBDIR/packages_all_$suite.db.new",
272 "$DBDIR/packages_all_$suite.db");
273 rename("$DBDIR/package_names_$suite.txt.new",
274 "$DBDIR/package_names_$suite.txt");
276 rename("$DBDIR/packages_descriptions.db.new",
277 "$DBDIR/packages_descriptions.db");
278 rename("$DBDIR/descriptions_packages.db.new",
279 "$DBDIR/descriptions_packages.db");
280 rename("$DBDIR/descriptions.txt.new", "$DBDIR/descriptions.txt");
281 rename("$DBDIR/descriptions.db.new", "$DBDIR/descriptions.db");
282 rename("$DBDIR/package_postfixes.db.new", "$DBDIR/package_postfixes.db");