X-Git-Url: https://git.deb.at/?a=blobdiff_plain;f=cgi-bin%2Fsearch_packages.pl;h=7aa7ffdcfa27122b7a678b81033afb89b1c02ba8;hb=bcb33dbb830672f6db19dfce97978da19d660291;hp=c3877fc04d6ae600602fc1dc05958b15f032c09d;hpb=11d34e2708df7b0bcdfe7d7bcbccdc5d10326457;p=deb%2Fpackages.git

diff --git a/cgi-bin/search_packages.pl b/cgi-bin/search_packages.pl
index c3877fc..7aa7ffd 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;
 }
@@ -187,28 +188,96 @@ my %archs = map { $_ => 1 } @archs;
 
 print "DEBUG: suites=@suites, sections=@sections, archs=@archs<br>" if $debug > 2;
 
+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)."<br>" if $debug > 2;
+	if ($suites{$data[0]} && ($archs{$data[1]} || $data[1] eq 'all')
+	    && $sections{$data[2]}) {
+	    print "DEBUG: Using entry ".join( ':', @data)."<br>" 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)."<br>" if $debug > 2;
+	if ($suites{$data[0]} && $sections{$data[1]}) {
+	    print "DEBUG: Using entry ".join( ':', @data)."<br>" if $debug > 2;
+	    push @$results, [ $key, @data ];
+	}
+    }
+}
+
+
 if ($searchon eq 'names') {
 
     $keyword = lc $keyword unless $case_bool;
     
-    my %packages;
-    tie %packages, 'DB_File', "$DBDIR/packages_small.db", O_RDONLY, 0666, $DB_BTREE
+    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 $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 {
-	    $data[2] = 'main';
+    if ($exact) {
+	read_entry( \%packages, $keyword, \@results );
+    } else {
+	my ($key, $prefixes) = ($keyword, '');
+	my %pkgs;
+	my $p_obj = tie my %pref, 'DB_File', "$DBDIR/package_postfixes.db", O_RDONLY, 0666, $DB_BTREE
+	    or die "couldn't tie postfix db $DBDIR/package_postfixes.db: $!";
+	$p_obj->seq( $key, $prefixes, R_CURSOR );
+	do {
+            if ($prefixes =~ /^\001(\d+)/o) {
+                $too_much_hits += $1;
+            } else {
+		print "DEBUG: add word $key<br>" if $debug > 2;
+		$pkgs{$key}++;
+		foreach (split /\000/o, $prefixes) {
+		    print "DEBUG: add word $_$key<br>" 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 );
 	}
-	print "DEBUG: Considering entry ".join( ':', @data)."<br>" if $debug > 2;
-	if ($suites{$data[0]} && ($archs{$data[1]} || $data[1] eq 'all')
-	    && $sections{$data[2]}) {
-	    print "DEBUG: Using entry ".join( ':', @data)."<br>" if $debug > 2;
-	    push @results, [ $keyword, @data ];
+	foreach my $pkg (sort keys %pkgs) {
+	    read_entry( \%packages, $pkg, \@results );
+	}
+    }
+} elsif ($searchon eq 'sourcenames') {
+ 
+    $keyword = lc $keyword unless $case_bool;
+    
+    my $obj = tie my %packages, 'DB_File', "$DBDIR/sources_small.db", O_RDONLY, 0666, $DB_BTREE
+	or die "couldn't tie DB $DBDIR/sources_small.db: $!";
+    
+    if ($exact) {
+	read_src_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/, $result) {
+		my @data = split ( /\s/, $_, 5 );
+		print "DEBUG: Considering entry ".join( ':', @data)."<br>" if $debug > 2;
+		if ($suites{$data[0]} && $sections{$data[1]}) {
+		    print "DEBUG: Using entry ".join( ':', @data)."<br>" if $debug > 2;
+		    push @results, [ $pkg , @data ];
+		}
+	    }
 	}
     }
 }
@@ -234,6 +303,10 @@ if ($format eq 'html') {
     }
 }
 
+if ($too_much_hits) {
+print "<p><strong>Your search was too wide so we will only display exact matches. At least <em>$too_much_hits</em> results have been omitted and will not be displayed. Please consider using a longer keyword or more keywords.</strong></p>";
+}
+
 if (!@results) {
     if ($format eq 'html') {
 	my $keyword_esc = uri_escape( $keyword );
@@ -278,11 +351,12 @@ 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} = $desc;
@@ -346,14 +420,14 @@ unless ($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 ($format eq 'html') {
@@ -376,12 +450,12 @@ unless ($search_on_sources) {
 		    print "<br>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) {
-			    $bp_link = sprintf "<a href=\"$ROOT/%s/%s/%s\">%s</a>", $ver, $sect, uri_escape( $bp ),  $bp;
+			    $bp_link = sprintf( "<a href=\"$ROOT/%s/%s/%s\">%s</a>",
+						$ver, $sect, uri_escape( $bp ),  $bp );
 			} else {
 			    $bp_link = $bp;
 			}
@@ -432,7 +506,9 @@ 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 "<p style=\"text-align:center\">$index_line</p>";
     }