diff options
Diffstat (limited to '_site/cgi-bin/find_sa.cgi')
| -rw-r--r-- | _site/cgi-bin/find_sa.cgi | 173 |
1 files changed, 173 insertions, 0 deletions
diff --git a/_site/cgi-bin/find_sa.cgi b/_site/cgi-bin/find_sa.cgi new file mode 100644 index 0000000..6b608a7 --- /dev/null +++ b/_site/cgi-bin/find_sa.cgi @@ -0,0 +1,173 @@ +#!/usr/bin/perl + +use strict; +use warnings; +use Storable qw(retrieve); +use Encode qw(decode_utf8); +use HTML::Escape qw(escape_html); +use Time::HiRes qw(gettimeofday tv_interval); +use BSD::Resource; + +# 1. Start Benchmark Timer +my $start_time = [gettimeofday]; +my $files_read = 0; # Track IO Activity + +# Decode search text +my $search_text = ''; +if (($ENV{QUERY_STRING} || '') =~ /^q=([^&]*)/) { + $search_text = decode_utf8($1 // ""); + $search_text =~ s/\P{Print}//g; + $search_text = substr($search_text, 0, 64); + $search_text =~ s/^\s+|\s+$//g; +} + +# We search using lowercase for the case-insensitive index +my $query = lc($search_text); +my $query_len = length($query); +my @results; + +if ($query_len >= 3 && -f 'sa.bin' && -f 'corpus.bin') { + open(my $fh_sa, '<', 'sa.bin') or die $!; + open(my $fh_cp, '<', 'corpus.bin') or die $!; + my $file_map = retrieve('file_map.dat'); + $files_read += 3; + + my $total_suffixes = (-s 'sa.bin') / 4; + + # Helper for binary search comparisons + sub compare_at { + my ($idx, $fh_sa, $fh_cp, $q, $len) = @_; + seek($fh_sa, $idx * 4, 0); + read($fh_sa, my $bin_off, 4); + my $off = unpack("L", $bin_off); + + seek($fh_cp, $off, 0); + read($fh_cp, my $text, $len); + return ($text cmp $q, $off); + } + + # Range Search: Find Left and Right boundaries + my ($low, $high) = (0, $total_suffixes - 1); + my $first_hit = -1; + while ($low <= $high) { + my $mid = int(($low + $high) / 2); + my ($cmp) = compare_at($mid, $fh_sa, $fh_cp, $query, $query_len); + if ($cmp >= 0) { $first_hit = $mid if $cmp == 0; $high = $mid - 1; } + else { $low = $mid + 1; } + } + + if ($first_hit != -1) { + ($low, $high) = ($first_hit, $total_suffixes - 1); + my $last_hit = $first_hit; + while ($low <= $high) { + my $mid = int(($low + $high) / 2); + my ($cmp) = compare_at($mid, $fh_sa, $fh_cp, $query, $query_len); + if ($cmp <= 0) { $last_hit = $mid if $cmp == 0; $low = $mid + 1; } + else { $high = $mid - 1; } + } + + # Collect unique file results + my %seen; + for my $i ($first_hit .. $last_hit) { + my (undef, $offset) = compare_at($i, $fh_sa, $fh_cp, $query, $query_len); + foreach my $m (@$file_map) { + if ($offset >= $m->{start} && $offset < $m->{end}) { + if (!$seen{$m->{path}}++) { + my $snip_start = ($offset - 30 < $m->{start}) ? $m->{start} : $offset - 30; + seek($fh_cp, $snip_start, 0); + read($fh_cp, my $raw_snip, 80); + + push @results, { + path => $m->{path}, + title => "Result: " . (split('/', $m->{path}))[-2], + snippet => "..." . escape_html($raw_snip) . "..." + }; + } + last; + } + } + last if @results >= 20; + } + } + close($fh_sa); + close($fh_cp); +} + +# 2. Calculate Metrics +my $end_time = [gettimeofday]; +my $elapsed = tv_interval($start_time, $end_time); + +my $rusage = getrusage(); +my $user_cpu = $rusage->utime; +my $system_cpu = $rusage->stime; +my $max_rss = $rusage->maxrss; + +# 3. Output +print "Content-Type: text/html\n\n"; + +my $list; +if ($search_text eq '') { + $list = "<p>Please enter a search term above.</p>"; +} elsif (@results == 0) { + $list = "<p>No results found for \"<b>" . escape_html($search_text) . "</b>\".</p>"; +} else { + $list = "<ul>"; + foreach my $res (@results) { + $list .= "<li><a href=\"/$res->{path}\">$res->{title}</a><br><small>$res->{snippet}</small></li>"; + } + $list .= "</ul>"; +} + +my $safe_search_text = escape_html($search_text); +my $year = (localtime)[5] + 1900; + +print <<"HTML"; +<!DOCTYPE html> +<html lang="en-us"> +<head> + <meta charset="utf-8"> + <meta name="viewport" content="width=device-width, initial-scale=1"> + <title>Search</title> + <link rel="stylesheet" href="/assets/css/main.css"> + <link rel="stylesheet" href="/assets/css/skeleton.css"> +</head> +<body> + <div id="nav-container" class="container"> + <ul id="navlist" class="left"> + <li><a href="/" class="link-decor-none">hme</a></li> + <li><a href="/log/" class="link-decor-none">log</a></li> + <li><a href="/projects/" class="link-decor-none">poc</a></li> + <li><a href="/about/" class="link-decor-none">abt</a></li> + <li class="active"><a href="/cgi-bin/find.cgi" class="link-decor-none">sws</a></li> + <li><a href="/feed.xml" class="link-decor-none">rss</a></li> + </ul> + </div> + <main class="container" id="main"> + <div class="container"> + <h2>Search</h2> + <form action="" method="GET"> + <input id="search-box" type="text" name="q" value="$safe_search_text"> + <input id="search-btn" type="submit" value="Search"> + </form> + $list + + <div style="background: #f4f4f4; padding: 10px; border-radius: 5px; font-family: monospace; font-size: 0.85em; margin-top: 20px; border: 1px solid #ddd;"> + <strong>Performance Metrics:</strong><br> + Total Time: @{[ sprintf("%.4f", $elapsed) ]} seconds<br> + User CPU: $user_cpu s<br> + System CPU: $system_cpu s<br> + Peak RAM: $max_rss KB<br> + Files Read: $files_read (IO Activity) + </div> + </div> + </main> + <div class="footer"> + <div class="container"> + <div class="twelve columns right container-2"> + <p id="footer-text">© ASCIIMX - $year</p> + </div> + </div> + </div> +</body> +</html> +HTML |