zquery.pl.in

harvest是一个下载html网页得机器人

#!@PERL@ -w
#
# Query a z39.50 server with zoomsh.
# 5dec1999/kjl
#
# Usage:
#   for production:
#     put it into your /cgi-bin/ and use it via your browser as
#	http://localhost/cgi-bin/zquery.pl
#   for limited debugging:
#     zquery.pl [querystring]
#
# Installation:
# Put this script to your cgi-bin directory of the http daemon.
# Edit $::HTML_BASE to point to the directory where the supplemental html
# pages are.
# Additionally, you can modify print_html_* () to customize the appearance
# of the html pages.
#

use CGI 'param';
use CGI::Carp;
use CGI::Carp qw(fatalsToBrowser);
use URI::Escape;
use Time::HiRes qw(gettimeofday tv_interval);
use integer;
use strict;

$::DEBUG = param ('DEBUG');
$::DEBUG = 0 unless ($::DEBUG);

#############################
# configuration starts here #
#############################
# where am I installed?
# this script will find out its location when started as cgi script.
# If you want to use it under mod_perl, move this script to your perl
# modules directory and change the path.
##########################################
$::ME = $ENV { 'SCRIPT_NAME' };
$::ME = "/Harvest/cgi-bin/zquery.pl" unless $::ME;

# prefix to supplemental html pages
$::HTML_BASE = "../brokers";

# base URL to XML objects
$::DATA_URL = "../brokers/tengu.local/objects-xml" if ($::DEBUG);

# where is the zebrasrv and the database to use
$::ZURL = "localhost:9999/harvest";	 # how to contact z39.50 server

# some default values
# you can override these with the values of your query form
$::OP	    = "and";	# default boolean operator for queries ("and", "or")
$::N_LINES  = 3;	# default number of lines per hit to show
$::PER_PAGE = 10;	# default number of hits per page to show
# end of configuration
###############################################################################


# where is zoomsh
$::ZOOMSH = "@prefix@/bin/zoomsh";

# do some sanity checks
test_zebra ();

# parse arguments
# no arguments
# we must be called as a cgi script
$::query = param ('query') if ($#ARGV == -1);

# we found some arguments
$::query = join (" ", @ARGV) if ($#ARGV >= 0);

if (! $::query) {

    print_html_header ("Harvest Search");
    print_html_query_form ();

} else {

    # page to show
    $::page = param ('page');
    $::page = 0 if (! $::page);

    # number of hits to show per page
    $::per_page = param ('per_page');
    $::per_page = $::PER_PAGE if (! $::per_page);

    # number of lines per entry to show
    $::n_lines = param ('n_lines');
    $::n_lines = $::N_LINES if (! $::n_lines);

    # how to combine query words
    $::op = param ('op');
    $::op = $::OP if (! $::op);

    @::terms = split_query ($::query);

    # build query string in zoomsh notation
    $::z_query = build_query ($::query, $::op, @::terms);

    $::start = $::page * $::per_page;

    $::zoomq =  "$::ZOOMSH \"connect $::ZURL\"";
    $::zoomq .= " \"set preferredRecordSyntax XML\"";
    $::zoomq .= " \"search $::z_query\"";
    $::zoomq .= " \"show $::start $::per_page\" quit";

    print_html_header ("Search result for \"$::query\"");
    print_html_query_form ($::query);
    print "Query sent to Zebra server:<br>\n<b>$::z_query</b>\n<p>\n"
	if ($::DEBUG);
    print "<b>$::query:</b>";

    $::t0 = [gettimeofday];

    #($::num_hit, $::aref) = read_hash_zoomsh ();
    ($::num_hit, $::aref) = read_hash_zoomsh_xml ();
    if ($::num_hit <= 0) {
	print_html_nomatch ();
    } else {
	@::hits = @{$::aref};

	$::elapsed = tv_interval ($::t0);

	# output
	print_num_hit ($::num_hit);
	printf (" in %.3f seconds\n", $::elapsed);

	print_hit ($::page, $::per_page, \@::hits, @::terms)
	    if ($::num_hit && $::num_hit > 0);

	# print pager navigation links
	print_html_page_index (uri_escape ($::query), $::num_hit,
			       $::page, $::per_page)
	    if ($::num_hit > $::per_page);
    }
}

print_html_footer ();
#############################################
# execution ends here			    #
#############################################

#############################################
# subroutines				    #
#############################################

#############################################
# html header				    #
#   args: title				    #
#############################################
sub print_html_header {

    my $title = shift;

    print << "EOF";
Content-type: text/html


<html>
<head>
<title>$title
</title>
<style><!--
div.pager A { color: black; }
//--></style>
</head>
<body
 bgcolor="#ffffff"
 text="#000000"
 link="#0000ff"
 vlink="#000080"
 alink="#ff0000"
>

EOF
}


#############################################
# html end tags				    #
#############################################
sub print_html_footer {

    print << "EOF";

<hr size=3 width=90% noshade>

<center>
<a href="http://harvest.sourceforge.net/"><b>About Harvest</b></a>
</center>

</body>
</html>
EOF
}


#############################################
# query form				    #
#   args: default querystring value	    #
#############################################
sub print_html_query_form {

    my $query = shift;

    $::op = $::OP unless ($::op);
    $::n_lines = $::N_LINES unless ($::n_lines);
    $::per_page = $::PER_PAGE unless ($::per_page);

    print << "EOF";
<table>
<tr>
<td><a href="$::HTML_BASE\/"><h1>Harvest</h1></a></td>
<td>&nbsp;&nbsp;&nbsp;</td>
<td>
<form action=$::ME method=get>

<nobr>
<input type=text name=query
EOF

    if ($query) {
	$query =~ s/\"/&quot;/g;
	print " value=\"$query\"";
    }

    print << "EOF";
 size=40 maxsize=200>
<input type=submit name=button value="Harvest Search">
<a href="$::HTML_BASE/help.html">Help</a>
</nobr>

<input type=hidden name=op value=$::op>
<input type=hidden name=n_lines value=$::n_lines>
<input type=hidden name=per_page value=$::per_page>
EOF

    print "<input type=hidden name=DEBUG value=$::DEBUG>\n" if ($::DEBUG);

    print << "EOF";
</form>
</td>
</tr>
</table>

EOF
}


#############################################
# print number of hits for a query	    #
# args: querystring, number_of_hits	    #
#############################################
sub print_num_hit {

    my $num_hit = shift;

    if ($num_hit == 1) {
	print " $num_hit match";
    } else {
	print " $num_hit matches";
    }

    print " found";
}


###################################################################
# split string to words or group of words when enclosed in quotes #
# args: string							  #
# returns: array with words, quoted words are kept in groups	  #
###################################################################
sub split_query {

    my $query = shift;
    my (@term, $i, $quoted, @res);

    $query =~ s/^\s+//g;    # remove leading whitespaces
    $query =~ s/\s+$//g;    # remove trailing whitespaces

    @term = split /\s+/, $query;
    $quoted = 0;
    $i = 0;
    foreach (@term) {
	if ($quoted) {
	    $res[$i] .= " $_";
	} else {
	    $res[$i] = $_;
	}

	$quoted = 1 if (/^\"/);

	if ($quoted) {
	    $quoted = 0 if (/\S+\"$/);
	}

	$i++ if (!$quoted);
    }

    for ($i = 0; $i <= $#res; $i++) {
	$res[$i] =~ s/\"$//g if ($res[$i] =~ s/^\"//g);
    }

    return @res;
}


#########################################
# highlight search terms		#
# args:					#
#   line to highlight			#
#   array with search terms		#
# returns:				#
#   line with highlighting		#
#########################################
sub highlight {

    my $line  = shift;
    my @terms = @_;

    foreach (@terms) {
	next if (/^\-|[\.\*\+\?]/);
	$line =~ s/(\b$_\b)/<b>$1<\/b>/ig;
    }

    return $line;
}


#########################################
# print hits				#
# arguments:				#
#   number of page			#
#   hits per page			#
#   hash of hit arrays			#
#   array of search terms		#
#########################################
sub print_hit {

    my $page     = shift;
    my $per_page = shift;
    my $aref     = shift;
    my @terms    = @_;

    my (@hits, $i, $entry, $key,
	$title, $last_modification_time, $file_size, $url, $full_text,
	@date);

    @hits = @{$aref};

    $entry = $page * $per_page;

    for $i (0 .. $#hits) {

	$entry++;

	$url			= $hits [$i] { 'url' };
	$last_modification_time = $hits [$i] { 'last-modification-time' };
	$file_size		= $hits [$i] { 'file-size' } / 1024;
	$file_size		= 1 if ($file_size == 0);
	$title			= $hits [$i] { 'title' };
	$full_text		= $hits [$i] { 'full-text' };
	$full_text		= highlight ($full_text, @terms);

	@date = gmtime ($last_modification_time);
	$last_modification_time = sprintf ("%d-%02d-%02d",
					   $date[5] + 1900, $date[4] + 1,
					   $date[3]);

	print "<p>\n";
	print "<b>$entry. </b>" if ($::DEBUG);
	print << "EOF";
<a href="$url"><b>$title</b></a><br>
$full_text<br>
EOF

	$url =~ s/^http\:\/\///g;
	print "$url <b>$file_size" . "KB $last_modification_time</b>\n";

	if ($::DEBUG) {
	    my $score    = $hits [$i] { 'score' };
	    my $filename = $hits [$i] { 'filename' };
	    $filename =~ s/\S+\/objects-xml\///g;
	    print "<br>\n";
	    print "Relevance score: $score\n<br>\n";
	    print "<a href=\"$::DATA_URL/$filename\">xsoif file: $filename</a>\n";
	}
    }
}


###############################
# print page-navigation links #
# args:			      #
#   querystring		      #
#   number of hits	      #
#   page number		      #
#   number of items per page  #
###############################
sub print_html_page_index {

    my $query    = shift;
    my $num_hit  = shift;
    my $page     = shift;
    my $per_page = shift;

    my ($url, $num_pages, $start, $i);

    $url =  "$::ME";
    $url .= "?query=$query";
    $url .= "&op=$::op";
    $url .= "&n_lines=$::n_lines";
    $url .= "&per_page=$per_page";
    $url .= "&DEBUG=$::DEBUG" if ($::DEBUG);

    $num_pages = $num_hit / $per_page;
    $num_pages++ if ($num_hit % $per_page);

    print "<center>\n";
    print "<div class=pager>\n";

    print "<a href=$url&page=" . ($page - 1) . ">Previous</a>&nbsp;\n"
	if ($page > 0);

    $start = $page - 5 if ($page > 5);
    $start = $start - ($page + 6 - $num_pages) if ($num_pages <= $page + 5);
    $start = 0 if (!defined $start || $start < 0);

    for ($i = $start; $i < $page; $i++) {
	print "<a href=$url&page=$i>" . ($i + 1) . "</a>&nbsp;\n";
    }

    print "<font color=red><b>";
    print $page + 1;
    print "</b></font>&nbsp;\n";

    for ($i = $page + 1; $i < $num_pages && $i < $start + 11; $i++) {
	print "<a href=$url&page=$i>" . ($i + 1) . "</a>&nbsp;\n";
    }

    print "<a href=$url&page=" . ($page + 1) . ">Next</a>\n"
	if ($page < $num_pages - 1);

    print "</div>\n";
    print "</center>\n";
}


#############################################
# print error page and exit		    #
# args: error message			    #
#############################################