X-Git-Url: https://git.deb.at/?a=blobdiff_plain;f=lib%2FPackages%2FSearch.pm;h=d04f44a3992507509be58892b857da7c9a693196;hb=f27f1fcc5bffc0696f694e962eb40dbb0c554855;hp=a88e5c973b648d7da24a52eb69eab55189e2842b;hpb=f480e4cc62ccc8ffa91357601649713a0a06f521;p=deb%2Fpackages.git

diff --git a/lib/Packages/Search.pm b/lib/Packages/Search.pm
index a88e5c9..d04f44a 100644
--- a/lib/Packages/Search.pm
+++ b/lib/Packages/Search.pm
@@ -43,7 +43,7 @@ package Packages::Search;
 use strict;
 use warnings;
 
-use CGI qw( -oldstyle_urls );
+#use CGI ();
 use POSIX;
 use HTML::Entities;
 use DB_File;
@@ -78,7 +78,7 @@ sub parse_params {
     my %params_ret = ( values => {}, errors => {} );
     my %params;
     if ($USE_PAGED_MODE) {
-        debug( "Use PAGED_MODE", 2 );
+        debug( "Use PAGED_MODE", 2 ) if DEBUG;
         %params = %$params_def;
         foreach (keys %page_params) {
             delete $params{$_};
@@ -90,7 +90,7 @@ sub parse_params {
 
     foreach my $param ( keys %params ) {
 	
-	debug( "Param <strong>$param</strong>", 2 );
+	debug( "Param <strong>$param</strong>", 2 ) if DEBUG;
 
 	my $p_value_orig = $cgi->param($param);
 
@@ -104,11 +104,11 @@ sub parse_params {
 
 	my @p_value = ($p_value_orig);
 
-	debug( "Value (Orig) ".($p_value_orig||""), 2 );
+	debug( "Value (Orig) ".($p_value_orig||""), 2 ) if DEBUG;
 
 	if ($params_def->{$param}{array} && defined $p_value_orig) {
 	    @p_value = split /$params_def->{$param}{array}/, $p_value_orig;
-	    debug( "Value (Array Split) ". join('##',@p_value), 2 );
+	    debug( "Value (Array Split) ". join('##',@p_value), 2 ) if DEBUG;
 	}
 
 	if ($params_def->{$param}{match} && defined $p_value_orig) {
@@ -118,7 +118,7 @@ sub parse_params {
 	}
 	@p_value = grep { defined $_ } @p_value;
 
-	debug( "Value (Match) ". join('##',@p_value), 2 );
+	debug( "Value (Match) ". join('##',@p_value), 2 ) if DEBUG;
 
 	unless (@p_value) {
 	    if (defined $params{$param}{default}) {
@@ -130,13 +130,14 @@ sub parse_params {
 	    }
 	}
 
-	debug( "Value (Default) ". join('##',@p_value), 2 );
+	debug( "Value (Default) ". join('##',@p_value), 2 ) if DEBUG;
 	my @p_value_no_replace = @p_value;
 
 	if ($params{$param}{replace} && @p_value) {
-	    @p_value = ();
 	    foreach my $pattern (keys %{$params{$param}{replace}}) {
-		foreach (@p_value_no_replace) {
+		my @p_value_tmp = @p_value;
+		@p_value = ();
+		foreach (@p_value_tmp) {
 		    if ($_ eq $pattern) {
 			my $replacement = $params{$param}{replace}{$_};
 			if (ref $replacement) {
@@ -151,7 +152,7 @@ sub parse_params {
 	    }
 	}
 	
-	debug( "Value (Final) ". join('##',@p_value), 2 );
+	debug( "Value (Final) ". join('##',@p_value), 2 ) if DEBUG;
 
 	if ($params_def->{$param}{array}) {
 	    $params_ret{values}{$param} = {
@@ -197,7 +198,7 @@ sub end {
     my $params = shift;
 
     use Data::Dumper;
-    debug( "end: ".Dumper($params) );
+    debug( "end: ".Dumper($params) ) if DEBUG;
     my $page = $params->{page}
     || DEFAULT_PAGE;
     my $res_per_page = $params->{number}
@@ -342,12 +343,12 @@ sub read_entry_all {
     my $result = $hash->{$key} || '';
     foreach (split /\000/o, $result) {
 	my @data = split ( /\s/o, $_, 8 );
-	debug( "Considering entry ".join( ':', @data), 2);
+	debug( "Considering entry ".join( ':', @data), 2) if DEBUG;
 	if ($opts->{h_archives}{$data[0]} && $opts->{h_suites}{$data[1]}
 	    && ($opts->{h_archs}{$data[2]} || $data[2] eq 'all'
 		|| $data[2] eq 'virtual')
 	    && ($opts->{h_sections}{$data[3]} || $data[3] eq 'v')) {
-	    debug( "Using entry ".join( ':', @data), 2);
+	    debug( "Using entry ".join( ':', @data), 2) if DEBUG;
 	    push @$results, [ $key, @data ];
 	} else {
 	    push @$non_results, [ $key, @data ];
@@ -362,32 +363,45 @@ sub read_entry {
 sub read_entry_simple {
     my ($hash, $key, $archives, $suite) = @_;
     my $result = $hash->{$key} || '';
-    my @data_fuzzy;
+    debug( "read_entry_simple: key=$key, archives=".
+	   join(" ",(keys %$archives)).", suite=$suite", 1);
+    my (@data_fuzzy, @data_virtual, @data_fuzzy_virtual);
     foreach (split /\000/o, $result) {
 	my @data = split ( /\s/o, $_, 8 );
-	debug( "Considering entry ".join( ':', @data), 2);
+	debug( "Considering entry ".join( ':', @data), 2) if DEBUG;
 	if ($data[1] eq $suite) {
-	    if ($archives->{$data[0]}) {
-		debug( "Using entry ".join( ':', @data), 2);
+	    if ($archives->{$data[0]}
+		&& ($data[2] ne 'virtual')) {
+		debug( "Using entry ".join( ':', @data), 2) if DEBUG;
 		return \@data;
-	    } elsif ($data[0] eq 'us') {
-		debug( "Fuzzy entry ".join( ':', @data), 2);
+	    } elsif ($archives->{$data[0]}) {
+		debug( "Virtual entry ".join( ':', @data), 2) if DEBUG;
+		@data_virtual = @data;
+	    } elsif (($data[0] eq 'us')
+		     && ($data[2] ne 'virtual')) {
+		debug( "Fuzzy entry ".join( ':', @data), 2) if DEBUG;
 		@data_fuzzy = @data;
+	    } elsif ($data[0] eq 'us') {
+		debug( "Virtual fuzzy entry ".join( ':', @data), 2) if DEBUG;
+		@data_fuzzy_virtual = @data;
 	    }
 	} 
     }
-    return \@data_fuzzy;
+    return \@data_virtual if @data_virtual;
+    return \@data_fuzzy if @data_fuzzy;
+    return \@data_fuzzy_virtual;
 }
 sub read_src_entry_all {
     my ($hash, $key, $results, $non_results, $opts) = @_;
     my $result = $hash->{$key} || '';
+    debug( "read_src_entry_all: key=$key", 1) if DEBUG;
     foreach (split /\000/o, $result) {
 	my @data = split ( /\s/o, $_, 6 );
-	debug( "Considering entry ".join( ':', @data), 2);
+	debug( "Considering entry ".join( ':', @data), 2) if DEBUG;
 	if ($opts->{h_archives}{$data[0]}
 	    && $opts->{h_suites}{$data[1]}
 	    && $opts->{h_sections}{$data[2]}) {
-	    debug( "Using entry ".join( ':', @data), 2);
+	    debug( "Using entry ".join( ':', @data), 2) if DEBUG;
 	    push @$results, [ $key, @data ];
 	} else {
 	    push @$non_results, [ $key, @data ];
@@ -400,44 +414,41 @@ sub read_src_entry {
     read_src_entry_all( $hash, $key, $results, \@non_results, $opts );
 }
 sub do_names_search {
-    my ($keyword, $packages, $postfixes, $read_entry, $opts) = @_;
-    my @results;
+    my ($keyword, $packages, $postfixes, $read_entry, $opts,
+	$results, $non_results) = @_;
 
-    $keyword = lc $keyword unless $opts->{case_bool};
+    $keyword = lc $keyword;
         
-    if ($opts->{exact}) {
-	&$read_entry( $packages, $keyword, \@results, $opts );
-    } else {
-	my ($key, $prefixes) = ($keyword, '');
-	my %pkgs;
-	$postfixes->seq( $key, $prefixes, R_CURSOR );
-	while (index($key, $keyword) >= 0) {
-            if ($prefixes =~ /^\001(\d+)/o) {
-                $too_many_hits += $1;
-            } else {
-		foreach (split /\000/o, $prefixes) {
-		    $_ = '' if $_ eq '^';
-		    debug( "add word $_$key", 2);
-		    $pkgs{$_.$key}++;
-		}
+    my ($key, $prefixes) = ($keyword, '');
+    my %pkgs;
+    $postfixes->seq( $key, $prefixes, R_CURSOR );
+    while (index($key, $keyword) >= 0) {
+	if ($prefixes =~ /^\001(\d+)/o) {
+	    debug( "$key has too many hits", 2 ) if DEBUG;
+	    $too_many_hits += $1;
+	} else {
+	    foreach (split /\000/o, $prefixes) {
+		$_ = '' if $_ eq '^';
+		debug( "add word $_$key", 2) if DEBUG;
+		$pkgs{$_.$key}++;
 	    }
-	    last if $postfixes->seq( $key, $prefixes, R_NEXT ) != 0;
-	    last if $too_many_hits or keys %pkgs >= 100;
-	}
-        
-        my $no_results = keys %pkgs;
-        if ($too_many_hits || ($no_results >= 100)) {
-	    $too_many_hits += $no_results;
-	    %pkgs = ( $keyword => 1 );
-	}
-	foreach my $pkg (sort keys %pkgs) {
-	    &$read_entry( $packages, $pkg, \@results, $opts );
 	}
+	last if $postfixes->seq( $key, $prefixes, R_NEXT ) != 0;
+	last if $too_many_hits or keys %pkgs >= 100;
+    }
+    
+    my $no_results = keys %pkgs;
+    if ($too_many_hits || ($no_results >= 100)) {
+	$too_many_hits += $no_results;
+	%pkgs = ( $keyword => 1 );
+    }
+    foreach my $pkg (sort keys %pkgs) {
+	&$read_entry( $packages, $pkg, $results, $non_results, $opts );
     }
-    return \@results;
 }
 sub do_fulltext_search {
-    my ($keyword, $file, $did2pkg, $packages, $read_entry, $opts) = @_;
+    my ($keyword, $file, $did2pkg, $packages, $read_entry, $opts,
+	$results, $non_results) = @_;
 
 # NOTE: this needs to correspond with parse-packages!
     $keyword =~ tr [A-Z] [a-z];
@@ -449,17 +460,19 @@ sub do_fulltext_search {
 
     my $numres = 0;
     my %tmp_results;
-    open DESC, '<', "$file"
+    # fgrep is seriously faster than using perl
+    open DESC, '-|', 'fgrep', '-n', '--', $keyword, $file
 	or die "couldn't open $file: $!";
     while (<DESC>) {
-	next if (index $_, $keyword) < 0;
-	debug( "Matched line $.: $_", 2);
-	my $result = $did2pkg->{$.};
+	/^(\d+)/;
+	my $nr = $1;
+	debug( "Matched line $_", 2) if DEBUG;
+	my $result = $did2pkg->{$nr};
 	foreach (split /\000/o, $result) {
 	    my @data = split /\s/, $_, 3;
-	    debug ("Considering $data[0], arch = $data[2]", 3);
-	    next unless $data[2] eq 'all' || $opts->{h_archs}{$data[2]};
-	    debug ("Ok", 3);
+#	    debug ("Considering $data[0], arch = $data[2]", 3) if DEBUG;
+#	    next unless $data[2] eq 'all' || $opts->{h_archs}{$data[2]};
+#	    debug ("Ok", 3) if DEBUG;
 	    $numres++ unless $tmp_results{$data[0]}++;
 	}
 	last if $numres > 100;
@@ -469,10 +482,9 @@ sub do_fulltext_search {
 
     my @results;
     foreach my $pkg (keys %tmp_results) {
-	&$read_entry( $packages, $pkg, \@results, $opts );
+	&$read_entry( $packages, $pkg, $results, $non_results, $opts );
     }
-    return \@results;
-}
+ }
 
 sub find_binaries {
     my ($pkg, $archive, $suite, $src2bin) = @_;
@@ -482,11 +494,11 @@ sub find_binaries {
     foreach (split /\000/o, $bins) {
 	my @data = split /\s/, $_, 5;
 
-	debug( "find_binaries: considering @data", 3 );
+	debug( "find_binaries: considering @data", 3 ) if DEBUG;
 	if (($data[0] eq $archive)
 	    && ($data[1] eq $suite)) {
 	    $bins{$data[2]}++;
-	    debug( "find_binaries: using @data", 3 );
+	    debug( "find_binaries: using @data", 3 ) if DEBUG;
 	}
     }