X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=zebra%2Fezeerex2pqfproperties.pl;h=a10c3f90839e5e72b5085b7215ec06b6e5c61f30;hb=eab4ac665d9829dc81c0cc114e707719e1d565fc;hp=09f6df411fcfc6a8b176bbe7a6d1403c697fb530;hpb=9caebcbf8b3e7c01bce7550c1c8e174953fa6318;p=irspy-moved-to-github.git

diff --git a/zebra/ezeerex2pqfproperties.pl b/zebra/ezeerex2pqfproperties.pl
index 09f6df4..a10c3f9 100755
--- a/zebra/ezeerex2pqfproperties.pl
+++ b/zebra/ezeerex2pqfproperties.pl
@@ -1,22 +1,143 @@
 #! /usr/bin/perl -w
 
-# $Id: ezeerex2pqfproperties.pl,v 1.2 2006-06-16 15:28:46 mike Exp $
+# $Id: ezeerex2pqfproperties.pl,v 1.4 2006-06-19 16:41:03 mike Exp $
+#
+# Run like this:
+#	./ezeerex2pqfproperties.pl zeerex.xml
 
 use strict;
 use warnings;
 use XML::LibXML;
 use XML::LibXML::XPathContext;
 
-my $text = join("", <>);
+my $text = join('', <>);
 my $parser = new XML::LibXML();
 my $doc = $parser->parse_string($text);
 my $root = $doc->getDocumentElement();
 my $xc = XML::LibXML::XPathContext->new($root);
-$xc->registerNs(zeerex => "http://explain.z3950.org/dtd/2.0/");
-print "root=$root, xc=$xc\n";
-my(@nodes) = $xc->findnodes('zeerex:serverInfo/zeerex:host');
-print "found ", scalar(@nodes), " values\n";
-for (my $i = 0; $i < @nodes; $i++) {
-    my $node = $nodes[$i];
-    print $i+1, ": ", $node->toString();
+$xc->registerNs(z => 'http://explain.z3950.org/dtd/2.0/');
+
+print_sets($xc);
+print_default_set($xc);
+print_indexes($xc);
+print_relations($xc);
+print_relation_modifiers($xc);
+print_positions($xc);
+print_structures($xc);
+print_truncations($xc);
+
+# We could limit the sets output to those that are actually used by an
+# SRU index: that way we could avoid defining
+#	set.bib1 = 1.2.840.10003.3.1
+# which is a Z39.50 attribute set that we don't need for CQL.  But
+# doing that would be a marginal gain.
+#
+sub print_sets {
+    my($xc) = @_;
+
+    my(@nodes) = $xc->findnodes('z:indexInfo/z:set');
+    foreach my $node (@nodes) {
+	my $name = $node->findvalue('@name');
+	my $identifier = $node->findvalue('@identifier');
+	print "set.$name = $identifier\n";
+    }
+}
+
+sub print_default_set {
+    my($xc) = @_;
+
+    my (@nodes) = $xc->findnodes('z:configInfo/' .
+				 'z:default[@type="contextSet"]');
+    foreach my $node (@nodes) {
+	### Look this up and render as a URI
+	print "set = ", $node->findvalue('.'), "\n";
+    }
+}
+
+sub print_indexes {
+    my($xc) = @_;
+
+    foreach my $node ($xc->findnodes('z:indexInfo/' .
+				     'z:index[@search="true"]')) {
+	my @pqf = $xc->findnodes("z:map/z:attr", $node);
+	die("no PQF mapping for index '" .
+	    $xc->findvalue("z:title", $node) . "'")
+	    if @pqf == 0;
+	my $ptype = $xc->findvalue('@type', $pqf[0]);
+	my $pval = $xc->findvalue(".", $pqf[0]);
+
+	foreach my $map ($xc->findnodes("z:map", $node)) {
+	    my $setname = $xc->findvalue('z:name/@set', $map);
+	    my $indexname = $xc->findvalue('z:name', $map);
+	    print "index.$setname.$indexname = $ptype=$pval\n"
+		if $indexname ne "";
+	}
+    }
+}
+
+# I don't think these are affected by the ZeeRex record
+sub print_relations {
+    my($xc) = @_;
+
+    print <<__EOT__;
+relation.< = 2=1
+relation.le = 2=2
+relation.eq = 2=3
+relation.exact = 2=3
+relation.ge = 2=4
+relation.> = 2=5
+relation.<> = 2=6
+relation.scr = 2=3
+__EOT__
+}
+
+# I don't think these are affected by the ZeeRex record
+sub print_relation_modifiers {
+    my($xc) = @_;
+
+    print <<__EOT__;
+relationModifier.relevant = 2=102
+relationModifier.fuzzy = 5=103
+relationModifier.stem = 2=101
+relationModifier.phonetic = 2=100
+relationModifier.regexp = 5=102
+__EOT__
+}
+
+# I don't think these are affected by the ZeeRex record
+sub print_positions {
+    my($xc) = @_;
+
+    print <<__EOT__;
+position.first = 3=1 6=1
+position.any = 3=3 6=1
+position.last = 3=4 6=1
+position.firstAndLast = 3=3 6=3
+__EOT__
+}
+
+# I don't think these are affected by the ZeeRex record
+sub print_structures {
+    my($xc) = @_;
+
+    print <<__EOT__;
+structure.exact = 4=108
+structure.all = 4=2
+structure.any = 4=2
+structure.* = 4=1
+__EOT__
+}
+
+# I don't think these are affected by the ZeeRex record
+sub print_truncations {
+    my($xc) = @_;
+
+    print <<__EOT__;
+truncation.right = 5=1
+truncation.left = 5=2
+truncation.both = 5=3
+truncation.none = 5=100
+truncation.regexp = 5=102
+truncation.z3958 = 5=104
+__EOT__
 }