X-Git-Url: http://git.indexdata.com/?p=irspy-moved-to-github.git;a=blobdiff_plain;f=lib%2FZOOM%2FIRSpy%2FRecord.pm;h=c213f54a3655add5c9742f51f389d0c683e406ec;hp=33699483337d6c402ca21c1d69fbb07fbf94cfc7;hb=3ab47c1a80bcf4b6e0f82a977ff865e1e1251d14;hpb=8c2e3dd4ca37ae405e0a5e5bb3a3d4f01abd0c87

diff --git a/lib/ZOOM/IRSpy/Record.pm b/lib/ZOOM/IRSpy/Record.pm
index 3369948..c213f54 100644
--- a/lib/ZOOM/IRSpy/Record.pm
+++ b/lib/ZOOM/IRSpy/Record.pm
@@ -1,4 +1,4 @@
-# $Id: Record.pm,v 1.7 2006-07-24 15:25:51 mike Exp $
+# $Id: Record.pm,v 1.15 2006-10-13 13:41:57 sondberg Exp $
 
 package ZOOM::IRSpy::Record;
 
@@ -6,6 +6,9 @@ use 5.008;
 use strict;
 use warnings;
 
+use Exporter 'import';
+our @EXPORT_OK = qw(xml_encode);
+
 use XML::LibXML;
 use XML::LibXML::XPathContext;
 
@@ -26,7 +29,7 @@ I<## To follow>
 
 sub new {
     my $class = shift();
-    my($target, $zeerex) = @_;
+    my($irspy, $target, $zeerex) = @_;
 
     if (!defined $zeerex) {
 	$zeerex = _empty_zeerex_record($target);
@@ -34,6 +37,7 @@ sub new {
 
     my $parser = new XML::LibXML();
     return bless {
+	irspy => $irspy,
 	target => $target,
 	parser => $parser,
 	zeerex => $parser->parse_string($zeerex)->documentElement(),
@@ -47,41 +51,65 @@ sub _empty_zeerex_record {
     ### Doesn't recognise SRU/SRW URLs
     my($host, $port, $db) = ZOOM::IRSpy::_parse_target_string($target);
 
+    my $xhost = xml_encode($host);
+    my $xport = xml_encode($port);
+    my $xdb = xml_encode($db);
     return <<__EOT__;
 <explain xmlns="http://explain.z3950.org/dtd/2.0/">
  <serverInfo protocol="Z39.50" version="1995">
-  <host>$host</host>
-  <port>$port</port>
-  <database>$db</database>
+  <host>$xhost</host>
+  <port>$xport</port>
+  <database>$xdb</database>
  </serverInfo>
 </explain>
 __EOT__
 }
 
 
+# I can't -- just can't, can't, can't -- believe that this function
+# isn't provided by one of the core XML modules.  But the evidence all
+# says that it's not: among other things, XML::Generator and
+# Template::Plugin both roll their own.  So I will do likewise.  D'oh!
+#
+sub xml_encode {
+    my ($text) = @_;
+    $text =~ s/&/&amp;/g;
+    $text =~ s/</&lt;/g;
+    $text =~ s/>/&gt;/g;
+    $text =~ s/['']/&apos;/g;
+    $text =~ s/[""]/&quot;/g;
+    return $text;
+}
+
+
 sub append_entry {
     my $this = shift();
     my($xpath, $frag) = @_;
 
-    print STDERR "this=$this, xpath='$xpath', frag='$frag'\n";
+    #print STDERR "this=$this, xpath='$xpath', frag='$frag'\n";
     my $root = $this->{zeerex}; # XML::LibXML::Element ISA XML::LibXML::Node
     my $xc = XML::LibXML::XPathContext->new($root);
     $xc->registerNs(zeerex => "http://explain.z3950.org/dtd/2.0/");
-    $xc->registerNs(irspy => "http://indexdata.com/irspy/1.0");
+    $xc->registerNs(irspy => $ZOOM::IRSpy::irspy_ns);
 
     my @nodes = $xc->findnodes($xpath);
     if (@nodes == 0) {
-	ZOOM::Log::log("irspy", "no matches for '$xpath': can't append");
-	return;
-    } elsif (@nodes > 1) {
-	ZOOM::Log::log("irspy", scalar(@nodes),
-		       " matches for '$xpath': using first");
+	# Make the node that we're inserting into, if possible.  A
+	# fully general version would work its way through each
+	# component of the XPath, but for now we just treat it as a
+	# single chunk to go inside the top-level node.
+	$this->_half_decent_appendWellBalancedChunk($root,
+						    "<$xpath></$xpath>");
+	@nodes = $xc->findnodes($xpath);
+	die("still no matches for '$xpath' after creating: can't append")
+	    if @nodes == 0;
     }
 
-    my $node = $nodes[0];
-    # $node ISA XML::LibXML::ElementXML::LibXML::Element
-    $this->_half_decent_appendWellBalancedChunk($node, $frag);
-    #print STDERR "POST: zeerex='$root' = \n", $root->toString(), "\n";
+    $this->{irspy}->log("warn",
+			scalar(@nodes), " matches for '$xpath': using first")
+	if @nodes > 1;
+
+    $this->_half_decent_appendWellBalancedChunk($nodes[0], $frag);
 }
 
 
@@ -101,7 +129,7 @@ sub append_entry {
 # namespace mapping for that node -- but that only affects pre-parsed
 # trees, and is no use for parsing.  Hence the following pair of lines
 # DOES NOT WORK:
-#	$node->setNamespace("http://indexdata.com/irspy/1.0", "irspy", 0);
+#	$node->setNamespace($ZOOM::IRSpy::irspy_ns, "irspy", 0);
 #	$node->appendWellBalancedChunk($frag);
 #
 # Instead I have to go the long way round, hence this method.  I have
@@ -115,7 +143,7 @@ sub _half_decent_appendWellBalancedChunk {
     my($node, $frag) = @_;
 
     if (1) {
-	$frag =~ s,>, xmlns:irspy="http://indexdata.com/irspy/1.0">,;
+	$frag =~ s,>, xmlns:irspy="$ZOOM::IRSpy::irspy_ns">,;
 	$node->appendWellBalancedChunk($frag);
 	return;
     }
@@ -132,7 +160,7 @@ sub _half_decent_appendWellBalancedChunk {
     die "mismatched XML start/end <$open>...<$close>"
 	if $close ne $tag;
     print STDERR "tag='$tag', attrs=[$attrs], content='$content'\n";
-    die "### no code yet to make DOM node";
+    die "## no code yet to make DOM node";
 }