Improve result sampling in case we have too generic keywords

Instead of either displaying the exact match if we have only
one keyword and nothing otherwise, lets try to give people
a sample of results, on the chance that might still be useful.

Mostly just display the first 100 package names found (i.e.
from suffix entries that had no "too many prefixes" value).
Allow suffix entries with "too many prefixes" value to specifiy
if the empty prefix was contained in the set (that effectivly
avoids a regression in the case "single keyword, exact match
exists").

If in doubt, prefer package names that begin with the first
keyword and only abort if we have more than 100 of those
(The worst case behaviour of this should be very limited
to due the "too many prefixes" constraints).

diff --git a/bin/parse-packages b/bin/parse-packages
index 2e713c6fbd93a596f19f853148ba1af21efcaeff..97edffcd19666bc4a438c7b1bfe1108a4ecfd132 100755 (executable)
--- a/bin/parse-packages
+++ b/bin/parse-packages
@@ -336,11 +336,10 @@ tie %package_postfixes_db, "DB_File", "$DBDIR/package_postfixes.db.new",
        or die "Error creating DB: $!";
 while (my ($k, $v) = each(%package_postfixes)) {
        $v =~ s/.$//s;
-       my $nr = $v;
-       $nr =~ s/[^\000]//g;
-       $nr = length($nr) + 1; # < number of hits
+       my $nr = ($v =~ tr/\000/\000/) + 1;
        if ($nr > $MAX_PACKAGE_POSTFIXES) {
-               $v = "\001" . $nr;
+           $v = ($v =~ /\^/) ? "^\001" . $nr
+               : "\001" . $nr;
        }
        $package_postfixes_db{$k} = $v;
 }

diff --git a/lib/Packages/Search.pm b/lib/Packages/Search.pm
index 04270f2301b9d29caa59e013d30843bb4f097f00..c33f1425322e0ddb10f0d18c51b2ef214919ea8b 100644 (file)
--- a/lib/Packages/Search.pm
+++ b/lib/Packages/Search.pm
@@ -164,14 +164,22 @@ sub do_names_search {
 
     my $first_keyword = lc shift @$keywords;
     @$keywords = map { lc $_ } @$keywords;
-        
+
     my ($key, $prefixes) = ($first_keyword, '');
-    my %pkgs;
+    my (%pkgs, %pkgs_min);
     $postfixes->seq( $key, $prefixes, R_CURSOR );
     while (index($key, $first_keyword) >= 0) {
-       if ($prefixes =~ /^\001(\d+)/o) {
-           debug( "$key has too many hits", 2 ) if DEBUG;
-           $too_many_hits += $1;
+       if ($prefixes =~ /^(\^)?\001(\d+)/o) {
+           debug("$key has too many hits", 2 ) if DEBUG;
+           $too_many_hits += $2;
+           if ($1) { # use the empty prefix
+               foreach my $k (@$keywords) {
+                   next unless $key =~ /\Q$k\E/;
+               }
+               debug("add key $key", 2) if DEBUG;
+               $pkgs{$key}++;
+               $pkgs_min{$key}++;
+           }
        } else {
          PREFIX:
            foreach (split /\000/o, $prefixes) {
@@ -180,18 +188,21 @@ sub do_names_search {
                foreach my $k (@$keywords) {
                    next PREFIX unless $word =~ /\Q$k\E/;
                }
-               debug( "add word $word", 2) if DEBUG;
+               debug("add word $word", 2) if DEBUG;
                $pkgs{$word}++;
+               $pkgs_min{$word}++ if $_ eq '';
            }
        }
        last if $postfixes->seq( $key, $prefixes, R_NEXT ) != 0;
-       last if $too_many_hits or keys %pkgs >= 100;
+       last if keys %pkgs_min >= 100;
     }
-    
-    my $no_results = keys %pkgs;
-    if ($too_many_hits || ($no_results >= 100)) {
-       $too_many_hits += $no_results;
-       %pkgs = ( $first_keyword => 1 ) unless @$keywords;
+
+    my $nr = keys %pkgs;
+    my $min_nr = keys %pkgs_min;
+    debug("nr=$nr min_nr=$min_nr too_many_hits=$too_many_hits", 1) if DEBUG;
+    if ($nr >= 100) {
+       $too_many_hits += $nr - $min_nr + 1;
+       %pkgs = %pkgs_min;
     }
     foreach my $pkg (sort keys %pkgs) {
        &$read_entry( $packages, $pkg, $results, $non_results, $opts );