X-Git-Url: https://git.deb.at/?a=blobdiff_plain;ds=sidebyside;f=cgi-bin%2Fsearch_packages.pl;h=060602a630138142dfe39b5f0b17402b397d316b;hb=d0086d64f40f548b3cdd63ca4223cc9bbaf70b76;hp=e22b8bd4bc812eba9bcd6bd2132c163181719d68;hpb=1d3541d031fd7eadfb034dc2ac6a644d3dddb7a8;p=deb%2Fpackages.git
diff --git a/cgi-bin/search_packages.pl b/cgi-bin/search_packages.pl
index e22b8bd..060602a 100755
--- a/cgi-bin/search_packages.pl
+++ b/cgi-bin/search_packages.pl
@@ -177,6 +177,7 @@ my $search_on_sources = 0;
my $st0 = new Benchmark;
my @results;
+my $too_much_hits;
if ($searchon eq 'sourcenames') {
$search_on_sources = 1;
}
@@ -187,51 +188,123 @@ my %archs = map { $_ => 1 } @archs;
print "DEBUG: suites=@suites, sections=@sections, archs=@archs
" if $debug > 2;
-if ($searchon eq 'names') {
+sub read_entry {
+ my ($hash, $key, $results) = @_;
+ my $result = $hash->{$key};
+ foreach (split /\000/, $result) {
+ my @data = split ( /\s/, $_, 7 );
+ print "DEBUG: Considering entry ".join( ':', @data)."
" if $debug > 2;
+ if ($suites{$data[0]} && ($archs{$data[1]} || $data[1] eq 'all')
+ && $sections{$data[2]}) {
+ print "DEBUG: Using entry ".join( ':', @data)."
" if $debug > 2;
+ push @$results, [ $key, @data ];
+ }
+ }
+}
+sub read_src_entry {
+ my ($hash, $key, $results) = @_;
+ my $result = $hash->{$key};
+
+ foreach (split /\000/, $result) {
+ my @data = split ( /\s/, $_, 5 );
+ print "DEBUG: Considering entry ".join( ':', @data)."
" if $debug > 2;
+ if ($suites{$data[0]} && $sections{$data[1]}) {
+ print "DEBUG: Using entry ".join( ':', @data)."
" if $debug > 2;
+ push @$results, [ $key, @data ];
+ }
+ }
+}
+sub do_names_search {
+ my ($keyword, $file, $postfix_file, $read_entry) = @_;
$keyword = lc $keyword unless $case_bool;
- my $obj = tie my %packages, 'DB_File', "$DBDIR/packages_small.db", O_RDONLY, 0666, $DB_BTREE
- or die "couldn't tie DB $DBDIR/packages_small.db: $!";
+ my $obj = tie my %packages, 'DB_File', "$DBDIR/$file", O_RDONLY, 0666, $DB_BTREE
+ or die "couldn't tie DB $DBDIR/$file: $!";
if ($exact) {
- my $result = $packages{$keyword};
- foreach (split /\000/, $result) {
- my @data = split ( /\s/, $_, 6 );
- #FIXME, should be done on db generation
- if ($data[2] =~ m,/,) {
- $data[2] =~ s,/.*$,,;
- } else {
- $data[2] = 'main';
- }
- print "DEBUG: Considering entry ".join( ':', @data)."
" if $debug > 2;
- if ($suites{$data[0]} && ($archs{$data[1]} || $data[1] eq 'all')
- && $sections{$data[2]}) {
- print "DEBUG: Using entry ".join( ':', @data)."
" if $debug > 2;
- push @results, [ $keyword, @data ];
- }
- }
+ &$read_entry( \%packages, $keyword, \@results );
} else {
- while (my ($pkg, $result) = each %packages) {
- #what's faster? I can't really see a difference
- (index($pkg, $keyword) >= 0) or next;
- #$pkg =~ /\Q$keyword\E/ or next;
- foreach (split /\000/, $packages{$pkg}) {
- my @data = split ( /\s/, $_, 6 );
- #FIXME, should be done on db generation
- if ($data[2] =~ m,/,) {
- $data[2] =~ s,/.*$,,;
- } else {
- $data[2] = 'main';
- }
- print "DEBUG: Considering entry ".join( ':', @data)."
" if $debug > 2;
- if ($suites{$data[0]} && ($archs{$data[1]} || $data[1] eq 'all')
- && $sections{$data[2]}) {
- print "DEBUG: Using entry ".join( ':', @data)."
" if $debug > 2;
- push @results, [ $pkg , @data ];
+ my ($key, $prefixes) = ($keyword, '');
+ my %pkgs;
+ my $p_obj = tie my %pref, 'DB_File', "$DBDIR/$postfix_file", O_RDONLY, 0666, $DB_BTREE
+ or die "couldn't tie postfix db $DBDIR/$postfix_file: $!";
+ $p_obj->seq( $key, $prefixes, R_CURSOR );
+ do {
+ if ($prefixes =~ /^\001(\d+)/o) {
+ $too_much_hits += $1;
+ } else {
+ print "DEBUG: add word $key
" if $debug > 2;
+ $pkgs{$key}++;
+ foreach (split /\000/o, $prefixes) {
+ print "DEBUG: add word $_$key
" if $debug > 2;
+ $pkgs{$_.$key}++;
}
}
+ } while (($p_obj->seq( $key, $prefixes, R_NEXT ) == 0)
+ && (index($key, $keyword) >= 0)
+ && !$too_much_hits
+ && (keys %pkgs < 100));
+
+ my $no_results = keys %pkgs;
+ if ($too_much_hits || ($no_results >= 100)) {
+ $too_much_hits += $no_results;
+ %pkgs = ( $keyword => 1 );
+ }
+ foreach my $pkg (sort keys %pkgs) {
+ &$read_entry( \%packages, $pkg, \@results );
+ }
+ }
+}
+
+if ($searchon eq 'names') {
+ do_names_search( $keyword, 'packages_small.db',
+ 'package_postfixes.db', \&read_entry );
+} elsif ($searchon eq 'sourcenames') {
+ do_names_search( $keyword, 'sources_small.db',
+ 'source_postfixes.db', \&read_src_entry );
+} else {
+
+ my @lines;
+ my $regex;
+ if ($case_bool) {
+ if ($exact) {
+ $regex = qr/\b\Q$keyword\E\b/o;
+ } else {
+ $regex = qr/\Q$keyword\E/o;
}
+ } else {
+ if ($exact) {
+ $regex = qr/\b\Q$keyword\E\b/io;
+ } else {
+ $regex = qr/\Q$keyword\E/io;
+ }
+ }
+
+ open DESC, '<', "$DBDIR/descriptions.txt" or die "couldn't open $DBDIR/descriptions.txt: $!";
+ while () {
+ $_ =~ $regex or next;
+ print "DEBUG: Matched line $.
" if $debug > 2;
+ push @lines, $.;
+ }
+ close DESC;
+
+ my $obj = tie my %packages, 'DB_File', "$DBDIR/packages_small.db", O_RDONLY, 0666, $DB_BTREE
+ or die "couldn't tie DB $DBDIR/packages_small.db: $!";
+ my $obj = tie my %did2pkg, 'DB_File', "$DBDIR/descriptions_packages.db", O_RDONLY, 0666, $DB_BTREE
+ or die "couldn't tie DB $DBDIR/descriptions_packages.db: $!";
+
+ my %tmp_results;
+ foreach my $l (@lines) {
+ my $result = $did2pkg{$l};
+ foreach (split /\000/o, $result) {
+ my @data = split /\s/, $_, 3;
+ next unless $archs{$data[2]};
+ $tmp_results{$data[0]}++;
+ }
+ }
+ foreach my $pkg (keys %tmp_results) {
+ read_entry( \%packages, $pkg, \@results );
}
}
@@ -256,6 +329,10 @@ if ($format eq 'html') {
}
}
+if ($too_much_hits) {
+print "Your search was too wide so we will only display exact matches. At least $too_much_hits results have been omitted and will not be displayed. Please consider using a longer keyword or more keywords.
";
+}
+
if (!@results) {
if ($format eq 'html') {
my $keyword_esc = uri_escape( $keyword );
@@ -300,11 +377,12 @@ my (%pkgs, %sect, %part, %desc, %binaries);
unless ($search_on_sources) {
foreach (@results) {
- my ($pkg_t, $suite, $arch, $section, $priority, $version, $desc) = @$_;
+ my ($pkg_t, $suite, $arch, $section, $subsection,
+ $priority, $version, $desc) = @$_;
my ($package) = $pkg_t =~ m/^(.+)/; # untaint
$pkgs{$package}{$suite}{$version}{$arch} = 1;
- $sect{$package}{$suite}{$version} = 'subsection';
+ $sect{$package}{$suite}{$version} = $subsection;
$part{$package}{$suite}{$version} = $section unless $section eq 'main';
$desc{$package}{$suite}{$version} = $desc;
@@ -368,14 +446,14 @@ unless ($search_on_sources) {
}
} else {
foreach (@results) {
- my ($package, $suite, $section, $version, $binaries);
+ my ($package, $suite, $section, $subsection, $priority,
+ $version, $binaries) = @$_;
$pkgs{$package}{$suite} = $version;
- $sect{$package}{$suite}{source} = 'subsection';
+ $sect{$package}{$suite}{source} = $subsection;
$part{$package}{$suite}{source} = $section unless $section eq 'main';
$binaries{$package}{$suite} = [ sort split( /\s*,\s*/, $binaries ) ];
-
}
if ($format eq 'html') {
@@ -398,12 +476,12 @@ unless ($search_on_sources) {
print "
Binary packages: ";
my @bp_links;
foreach my $bp (@{$binaries{$pkg}{$ver}}) {
- my $sect = find_section($bp, $ver, $part{$pkg}{$ver}{source}||'main') || '';
- $sect =~ s,^(non-free|contrib)/,,;
- $sect =~ s,^non-US.*$,non-US,,;
+ my $sect = 'section';
+
my $bp_link;
if ($sect) {
- $bp_link = sprintf "%s", $ver, $sect, uri_escape( $bp ), $bp;
+ $bp_link = sprintf( "%s",
+ $ver, $sect, uri_escape( $bp ), $bp );
} else {
$bp_link = $bp;
}
@@ -454,7 +532,9 @@ sub printindexline {
my $index_line;
if ($no_results > $results_per_page) {
- $index_line = prevlink($input,\%params)." | ".indexline( $input, \%params, $no_results)." | ".nextlink($input,\%params, $no_results);
+ $index_line = prevlink($input,\%params)." | ".
+ indexline( $input, \%params, $no_results)." | ".
+ nextlink($input,\%params, $no_results);
print "$index_line
";
}