3 # search_packages.pl -- CGI interface to the Packages files on packages.debian.org
5 # Copyright (C) 1998 James Treacy
6 # Copyright (C) 2000, 2001 Josip Rodin
7 # Copyright (C) 2001 Adam Heath
8 # Copyright (C) 2004 Martin Schulze
9 # Copyright (C) 2004-2006 Frank Lichtenheld
11 # use is allowed under the terms of the GNU Public License (GPL)
12 # see http://www.fsf.org/copyleft/gpl.html for a copy of the license
15 use CGI qw( -oldstyle_urls );
16 use CGI::Carp qw( fatalsToBrowser );
26 use Packages::Search qw( :all );
27 use Packages::HTML ();
29 my $thisscript = "search_packages.pl";
30 my $HOME = "http://www.debian.org";
32 my $SEARCHPAGE = "http://packages.debian.org/";
33 my @SUITES = qw( oldstable stable testing unstable experimental );
35 my @SECTIONS = qw( main contrib non-free );
36 my @ARCHIVES = qw( us security installer );
37 my @ARCHITECTURES = qw( alpha amd64 arm hppa hurd-i386 i386 ia64
38 kfreebsd-i386 mips mipsel powerpc s390 sparc );
39 my %SUITES = map { $_ => 1 } @SUITES;
40 my %SECTIONS = map { $_ => 1 } @SECTIONS;
41 my %ARCHIVES = map { $_ => 1 } @ARCHIVES;
42 my %ARCHITECTURES = map { $_ => 1 } @ARCHITECTURES;
44 $ENV{PATH} = "/bin:/usr/bin";
46 # Read in all the variables set by the form
49 my $pet0 = new Benchmark;
50 # use this to disable debugging in production mode completly
51 my $debug_allowed = 1;
52 my $debug = $debug_allowed && $input->param("debug");
53 $Search::Param::debug = 1 if $debug > 1;
55 # If you want, just print out a list of all of the variables and exit.
56 print $input->header if $debug;
60 if (my $path = $input->param('path')) {
61 my @components = map { lc $_ } split /\//, $path;
63 foreach (@components) {
65 $input->param('suite', $_);
66 } elsif ($SECTIONS{$_}) {
67 $input->param('section', $_);
68 } elsif ($ARCHIVES{$_}) {
69 $input->param('archive', $_);
70 }elsif ($ARCHITECTURES{$_}) {
71 $input->param('arch', $_);
76 my %params_def = ( keywords => { default => undef, match => '^\s*([-+\@\w\/.:]+)\s*$' },
77 suite => { default => 'stable', match => '^(\w+)$',
78 alias => 'version', array => ',',
79 replace => { all => \@SUITES } },
80 case => { default => 'insensitive', match => '^(\w+)$' },
81 official => { default => 0, match => '^(\w+)$' },
82 use_cache => { default => 1, match => '^(\w+)$' },
83 subword => { default => 0, match => '^(\w+)$' },
84 exact => { default => undef, match => '^(\w+)$' },
85 searchon => { default => 'all', match => '^(\w+)$' },
86 section => { default => 'all', match => '^([\w-]+)$',
87 alias => 'release', array => ',',
88 replace => { all => \@SECTIONS } },
89 arch => { default => 'any', match => '^(\w+)$',
90 array => ',', replace =>
91 { any => \@ARCHITECTURES } },
92 archive => { default => 'all', match => '^(\w+)$',
93 array => ',', replace =>
94 { all => \@ARCHIVES } },
95 format => { default => 'html', match => '^(\w+)$' },
97 my %params = Packages::Search::parse_params( $input, \%params_def );
99 my $format = $params{values}{format}{final};
100 #XXX: Don't use alternative output formats yet
103 if ($format eq 'html') {
104 print $input->header;
105 } elsif ($format eq 'xml') {
106 # print $input->header( -type=>'application/rdf+xml' );
107 print $input->header( -type=>'text/plain' );
110 if ($params{errors}{keywords}) {
111 print "Error: keyword not valid or missing" if $format eq 'html';
114 my $keyword = $params{values}{keywords}{final};
115 my @suites = @{$params{values}{suite}{final}};
116 my $official = $params{values}{official}{final};
117 my $use_cache = $params{values}{use_cache}{final};
118 my $case = $params{values}{case}{final};
119 my $case_bool = ( $case !~ /insensitive/ );
120 my $subword = $params{values}{subword}{final};
121 my $exact = $params{values}{exact}{final};
122 $exact = !$subword unless defined $exact;
123 my $searchon = $params{values}{searchon}{final};
124 my @sections = @{$params{values}{section}{final}};
125 my @archs = @{$params{values}{arch}{final}};
126 my $page = $params{values}{page}{final};
127 my $results_per_page = $params{values}{number}{final};
129 # for URL construction
130 my $suites_param = join ',', @{$params{values}{suite}{no_replace}};
131 my $sections_param = join ',', @{$params{values}{section}{no_replace}};
132 my $archs_param = join ',', @{$params{values}{arch}{no_replace}};
135 my $keyword_enc = encode_entities $keyword;
136 my $searchon_enc = encode_entities $searchon;
137 my $suites_enc = encode_entities join ', ', @{$params{values}{suite}{no_replace}};
138 my $sections_enc = encode_entities join ', ', @{$params{values}{section}{no_replace}};
139 my $archs_enc = encode_entities join ', ', @{$params{values}{arch}{no_replace}};
140 my $pet1 = new Benchmark;
141 my $petd = timediff($pet1, $pet0);
142 print "DEBUG: Parameter evaluation took ".timestr($petd)."<br>" if $debug;
144 if ($format eq 'html') {
145 print Packages::HTML::header( title => 'Package Search Results' ,
147 title_tag => 'Debian Package Search Results',
148 print_title_above => 1,
149 print_search_field => 'packages',
150 search_field_values => {
151 keywords => $keyword_enc,
152 searchon => $searchon,
154 suite => $suites_enc,
155 section => $sections_enc,
163 # read the configuration
165 if (!open (C, "../config.sh")) {
166 print "\nInternal Error: Cannot open configuration file.\n\n"
167 if $format eq 'html';
171 $topdir = $1 if (/^\s*topdir="?(.*)"?\s*$/);
175 my $DBDIR = $topdir . "/files/db";
176 my $search_on_sources = 0;
178 my $st0 = new Benchmark;
181 if ($searchon eq 'sourcenames') {
182 $search_on_sources = 1;
185 my %suites = map { $_ => 1 } @suites;
186 my %sections = map { $_ => 1 } @sections;
187 my %archs = map { $_ => 1 } @archs;
189 print "DEBUG: suites=@suites, sections=@sections, archs=@archs<br>" if $debug > 2;
192 my ($hash, $key, $results) = @_;
193 my $result = $hash->{$key};
194 foreach (split /\000/, $result) {
195 my @data = split ( /\s/, $_, 7 );
196 print "DEBUG: Considering entry ".join( ':', @data)."<br>" if $debug > 2;
197 if ($suites{$data[0]} && ($archs{$data[1]} || $data[1] eq 'all')
198 && $sections{$data[2]}) {
199 print "DEBUG: Using entry ".join( ':', @data)."<br>" if $debug > 2;
200 push @$results, [ $key, @data ];
205 my ($hash, $key, $results) = @_;
206 my $result = $hash->{$key};
208 foreach (split /\000/, $result) {
209 my @data = split ( /\s/, $_, 5 );
210 print "DEBUG: Considering entry ".join( ':', @data)."<br>" if $debug > 2;
211 if ($suites{$data[0]} && $sections{$data[1]}) {
212 print "DEBUG: Using entry ".join( ':', @data)."<br>" if $debug > 2;
213 push @$results, [ $key, @data ];
219 if ($searchon eq 'names') {
221 $keyword = lc $keyword unless $case_bool;
223 my $obj = tie my %packages, 'DB_File', "$DBDIR/packages_small.db", O_RDONLY, 0666, $DB_BTREE
224 or die "couldn't tie DB $DBDIR/packages_small.db: $!";
227 read_entry( \%packages, $keyword, \@results );
229 my ($key, $prefixes) = ($keyword, '');
231 my $p_obj = tie my %pref, 'DB_File', "$DBDIR/package_postfixes.db", O_RDONLY, 0666, $DB_BTREE
232 or die "couldn't tie postfix db $DBDIR/package_postfixes.db: $!";
233 $p_obj->seq( $key, $prefixes, R_CURSOR );
235 if ($prefixes =~ /^\001(\d+)/o) {
236 $too_much_hits += $1;
238 print "DEBUG: add word $key<br>" if $debug > 2;
240 foreach (split /\000/o, $prefixes) {
241 print "DEBUG: add word $_$key<br>" if $debug > 2;
245 } while (($p_obj->seq( $key, $prefixes, R_NEXT ) == 0)
246 && (index($key, $keyword) >= 0)
248 && (keys %pkgs < 100));
250 my $no_results = keys %pkgs;
251 if ($too_much_hits || ($no_results >= 100)) {
252 $too_much_hits += $no_results;
253 %pkgs = ( $keyword => 1 );
255 foreach my $pkg (sort keys %pkgs) {
256 read_entry( \%packages, $pkg, \@results );
259 } elsif ($searchon eq 'sourcenames') {
261 $keyword = lc $keyword unless $case_bool;
263 my $obj = tie my %packages, 'DB_File', "$DBDIR/sources_small.db", O_RDONLY, 0666, $DB_BTREE
264 or die "couldn't tie DB $DBDIR/sources_small.db: $!";
267 read_src_entry( \%packages, $keyword, \@results );
269 while (my ($pkg, $result) = each %packages) {
270 #what's faster? I can't really see a difference
271 (index($pkg, $keyword) >= 0) or next;
272 #$pkg =~ /\Q$keyword\E/ or next;
273 foreach (split /\000/, $result) {
274 my @data = split ( /\s/, $_, 5 );
275 print "DEBUG: Considering entry ".join( ':', @data)."<br>" if $debug > 2;
276 if ($suites{$data[0]} && $sections{$data[1]}) {
277 print "DEBUG: Using entry ".join( ':', @data)."<br>" if $debug > 2;
278 push @results, [ $pkg , @data ];
289 $regex = qr/\b\Q$keyword\E\b/o;
291 $regex = qr/\Q$keyword\E/o;
295 $regex = qr/\b\Q$keyword\E\b/io;
297 $regex = qr/\Q$keyword\E/io;
301 open DESC, '<', "$DBDIR/descriptions.txt" or die "couldn't open $DBDIR/descriptions.txt: $!";
303 $_ =~ $regex or next;
304 print "DEBUG: Matched line $.<br>" if $debug > 2;
309 my $obj = tie my %packages, 'DB_File', "$DBDIR/packages_small.db", O_RDONLY, 0666, $DB_BTREE
310 or die "couldn't tie DB $DBDIR/packages_small.db: $!";
311 my $obj = tie my %did2pkg, 'DB_File', "$DBDIR/descriptions_packages.db", O_RDONLY, 0666, $DB_BTREE
312 or die "couldn't tie DB $DBDIR/descriptions_packages.db: $!";
315 foreach my $l (@lines) {
316 my $result = $did2pkg{$l};
317 foreach (split /\000/o, $result) {
318 my @data = split /\s/, $_, 3;
319 next unless $archs{$data[2]};
320 $tmp_results{$data[0]}++;
323 foreach my $pkg (keys %tmp_results) {
324 read_entry( \%packages, $pkg, \@results );
328 my $st1 = new Benchmark;
329 my $std = timediff($st1, $st0);
330 print "DEBUG: Search took ".timestr($std)."<br>" if $debug;
332 if ($format eq 'html') {
333 my $suite_wording = $suites_enc eq "all" ? "all suites"
334 : "suite(s) <em>$suites_enc</em>";
335 my $section_wording = $sections_enc eq 'all' ? "all sections"
336 : "section(s) <em>$sections_enc</em>";
337 my $arch_wording = $archs_enc eq 'any' ? "all architectures"
338 : "architecture(s) <em>$archs_enc</em>";
339 if (($searchon eq "names") || ($searchon eq 'sourcenames')) {
340 my $source_wording = $search_on_sources ? "source " : "";
341 my $exact_wording = $exact ? "named" : "that names contain";
342 print "<p>You have searched for ${source_wording}packages $exact_wording <em>$keyword_enc</em> in $suite_wording, $section_wording, and $arch_wording.</p>";
344 my $exact_wording = $exact ? "" : " (including subword matching)";
345 print "<p>You have searched for <em>$keyword_enc</em> in packages names and descriptions in $suite_wording, $section_wording, and $arch_wording$exact_wording.</p>";
349 if ($too_much_hits) {
350 print "<p><strong>Your search was too wide so we will only display exact matches. At least <em>$too_much_hits</em> results have been omitted and will not be displayed. Please consider using a longer keyword or more keywords.</strong></p>";
354 if ($format eq 'html') {
355 my $keyword_esc = uri_escape( $keyword );
357 if (($searchon eq "names") || ($searchon eq 'sourcenames')) {
358 if (($suites_enc eq 'all')
359 && ($archs_enc eq 'any')
360 && ($sections_enc eq 'all')) {
361 print "<p><strong>Can't find that package.</strong></p>\n";
363 print "<p><strong>Can't find that package, at least not in that suite ".
364 ( $search_on_sources ? "" : " and on that architecture" ).
370 print "<p>You have searched only for exact matches of the package name. You can try to search for <a href=\"$thisscript?exact=0&searchon=$searchon&suite=$suites_param&case=$case&section=$sections_param&keywords=$keyword_esc&arch=$archs_param\">package names that contain your search string</a>.</p>";
373 if (($suites_enc eq 'all')
374 && ($archs_enc eq 'any')
375 && ($sections_enc eq 'all')) {
376 print "<p><strong>Can't find that string.</strong></p>\n";
378 print "<p><strong>Can't find that string, at least not in that suite ($suites_enc, section $sections_enc) and on that architecture ($archs_enc).</strong></p>\n";
383 print "<p>You have searched only for words exactly matching your keywords. You can try to search <a href=\"$thisscript?subword=1&searchon=$searchon&suite=$suites_param&case=$case&section=$sections_param&keywords=$keyword_esc&arch=$archs_param\">allowing subword matching</a>.</p>";
386 print "<p>".( $printed ? "Or you" : "You" )." can try a different search on the <a href=\"$SEARCHPAGE#search_packages\">Packages search page</a>.</p>";
393 my (%pkgs, %sect, %part, %desc, %binaries);
395 unless ($search_on_sources) {
397 my ($pkg_t, $suite, $arch, $section, $subsection,
398 $priority, $version, $desc) = @$_;
400 my ($package) = $pkg_t =~ m/^(.+)/; # untaint
401 $pkgs{$package}{$suite}{$version}{$arch} = 1;
402 $sect{$package}{$suite}{$version} = $subsection;
403 $part{$package}{$suite}{$version} = $section unless $section eq 'main';
405 $desc{$package}{$suite}{$version} = $desc;
409 if ($format eq 'html') {
410 my ($start, $end) = multipageheader( scalar keys %pkgs );
413 foreach my $pkg (sort keys %pkgs) {
415 next if $count < $start or $count > $end;
416 printf "<h3>Package %s</h3>\n", $pkg;
418 foreach my $ver (@SUITES) {
419 if (exists $pkgs{$pkg}{$ver}) {
420 my @versions = version_sort keys %{$pkgs{$pkg}{$ver}};
422 if ($part{$pkg}{$ver}{$versions[0]}) {
423 $part_str = "[<span style=\"color:red\">$part{$pkg}{$ver}{$versions[0]}</span>]";
425 printf "<li><a href=\"$ROOT/%s/%s/%s\">%s</a> (%s): %s %s\n",
426 $ver, $sect{$pkg}{$ver}{$versions[0]}, $pkg, $ver, $sect{$pkg}{$ver}{$versions[0]}, $desc{$pkg}{$ver}{$versions[0]}, $part_str;
428 foreach my $v (@versions) {
429 printf "<br>%s: %s\n",
430 $v, join (" ", (sort keys %{$pkgs{$pkg}{$ver}{$v}}) );
437 } elsif ($format eq 'xml') {
438 require RDF::Simple::Serialiser;
439 my $rdf = new RDF::Simple::Serialiser;
440 $rdf->addns( debpkg => 'http://packages.debian.org/xml/01-debian-packages-rdf' );
442 foreach my $pkg (sort keys %pkgs) {
443 foreach my $ver (@DISTS) {
444 if (exists $pkgs{$pkg}{$ver}) {
445 my @versions = version_sort keys %{$pkgs{$pkg}{$ver}};
446 foreach my $version (@versions) {
447 my $id = "$ROOT/$ver/$sect{$pkg}{$ver}{$version}/$pkg/$version";
448 push @triples, [ $id, 'debpkg:package', $pkg ];
449 push @triples, [ $id, 'debpkg:version', $version ];
450 push @triples, [ $id, 'debpkg:section', $sect{$pkg}{$ver}{$version}, ];
451 push @triples, [ $id, 'debpkg:suite', $ver ];
452 push @triples, [ $id, 'debpkg:shortdesc', $desc{$pkg}{$ver}{$version} ];
453 push @triples, [ $id, 'debpkg:part', $part{$pkg}{$ver}{$version} || 'main' ];
454 foreach my $arch (sort keys %{$pkgs{$pkg}{$ver}{$version}}) {
455 push @triples, [ $id, 'debpkg:architecture', $arch ];
462 print $rdf->serialise(@triples);
466 my ($package, $suite, $section, $subsection, $priority,
467 $version, $binaries) = @$_;
469 $pkgs{$package}{$suite} = $version;
470 $sect{$package}{$suite}{source} = $subsection;
471 $part{$package}{$suite}{source} = $section unless $section eq 'main';
473 $binaries{$package}{$suite} = [ sort split( /\s*,\s*/, $binaries ) ];
476 if ($format eq 'html') {
477 my ($start, $end) = multipageheader( scalar keys %pkgs );
480 foreach my $pkg (sort keys %pkgs) {
482 next if ($count < $start) or ($count > $end);
483 printf "<h3>Source package %s</h3>\n", $pkg;
485 foreach my $ver (@SUITES) {
486 if (exists $pkgs{$pkg}{$ver}) {
488 if ($part{$pkg}{$ver}{source}) {
489 $part_str = "[<span style=\"color:red\">$part{$pkg}{$ver}{source}</span>]";
491 printf "<li><a href=\"$ROOT/%s/source/%s\">%s</a> (%s): %s %s", $ver, $pkg, $ver, $sect{$pkg}{$ver}{source}, $pkgs{$pkg}{$ver}, $part_str;
493 print "<br>Binary packages: ";
495 foreach my $bp (@{$binaries{$pkg}{$ver}}) {
496 my $sect = 'section';
500 $bp_link = sprintf( "<a href=\"$ROOT/%s/%s/%s\">%s</a>",
501 $ver, $sect, uri_escape( $bp ), $bp );
505 push @bp_links, $bp_link;
507 print join( ", ", @bp_links );
513 } elsif ($format eq 'xml') {
514 require RDF::Simple::Serialiser;
515 my $rdf = new RDF::Simple::Serialiser;
516 $rdf->addns( debpkg => 'http://packages.debian.org/xml/01-debian-packages-rdf' );
518 foreach my $pkg (sort keys %pkgs) {
519 foreach my $ver (@SUITES) {
520 if (exists $pkgs{$pkg}{$ver}) {
521 my $id = "$ROOT/$ver/source/$pkg";
523 push @triples, [ $id, 'debpkg:package', $pkg ];
524 push @triples, [ $id, 'debpkg:type', 'source' ];
525 push @triples, [ $id, 'debpkg:section', $sect{$pkg}{$ver}{source} ];
526 push @triples, [ $id, 'debpkg:version', $pkgs{$pkg}{$ver} ];
527 push @triples, [ $id, 'debpkg:part', $part{$pkg}{$ver}{source} || 'main' ];
529 foreach my $bp (@{$binaries{$pkg}{$ver}}) {
530 push @triples, [ $id, 'debpkg:binary', $bp ];
535 print $rdf->serialise(@triples);
539 if ($format eq 'html') {
540 &printindexline( scalar keys %pkgs );
547 my $no_results = shift;
550 if ($no_results > $results_per_page) {
552 $index_line = prevlink($input,\%params)." | ".
553 indexline( $input, \%params, $no_results)." | ".
554 nextlink($input,\%params, $no_results);
556 print "<p style=\"text-align:center\">$index_line</p>";
560 sub multipageheader {
561 my $no_results = shift;
564 if ($results_per_page =~ /^all$/i) {
567 $results_per_page = $no_results;
569 $start = Packages::Search::start( \%params );
570 $end = Packages::Search::end( \%params );
571 if ($end > $no_results) { $end = $no_results; }
574 print "<p>Found <em>$no_results</em> matching packages,";
575 if ($end == $start) {
576 print " displaying package $end.</p>";
578 print " displaying packages $start to $end.</p>";
581 printindexline( $no_results );
583 if ($no_results > 100) {
584 print "<p>Results per page: ";
587 if ($results_per_page == $_) {
588 push @resperpagelinks, $_;
590 push @resperpagelinks, resperpagelink($input,\%params,$_);
593 if ($params{values}{number}{final} =~ /^all$/i) {
594 push @resperpagelinks, "all";
596 push @resperpagelinks, resperpagelink($input, \%params,"all");
598 print join( " | ", @resperpagelinks )."</p>";
600 return ( $start, $end );
608 <p style="text-align:right;font-size:small;font-stlye:italic"><a href="$SEARCHPAGE">Packages search page</a></p>
613 print $input->end_html;