From 5acdb022908bd1504f1f9fb4f766f458dc5ceaa0 Mon Sep 17 00:00:00 2001
From: Andrew Ruthven <andrew@etc.gen.nz>
Date: Fri, 30 Nov 2007 06:46:57 +1300
Subject: [PATCH] Clean out the bad UTF-8 characters.

---
 ChangeLog         |  1 +
 bin/mythtv-status | 16 ++++++++++++++++
 2 files changed, 17 insertions(+)

diff --git a/ChangeLog b/ChangeLog
index d521ac6..b7713e7 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,6 @@
 2007-11-28	Andrew Ruthven
 	Be more wary about processing what the backend has sent us.
+		This includes cleaning up some invalid UTF-8 characters.
 	Add support for reading XML from a file.
 	Be a bit more forgiving on the XML we're receiving.
 
diff --git a/bin/mythtv-status b/bin/mythtv-status
index 6d533b9..13090a1 100755
--- a/bin/mythtv-status
+++ b/bin/mythtv-status
@@ -305,6 +305,8 @@ sub load_xml {
   $parser->recover_silently(1)
     unless $verbose;
 
+  clean_xml(\$status);
+
   my $xml = eval { $parser->parse_string( $status ) };
 
   if ($@) {
@@ -327,6 +329,20 @@ sub load_perl_api {
   return $myth;
 }
 
+# We are sometimes passed dodgy XML from MythTV, make some attempts to clean
+# it.
+sub clean_xml {
+  my ($xml) = shift;
+
+  # Deal to invalid Unicode.
+  for my $bad ("&#xdbff;", "&#xdee9;") {
+    if ($$xml =~ s/$bad/?/g) {
+      warn "Found and replaced: $bad\n"
+        if $verbose;
+    }
+  }
+}
+
 sub process_xml {
   my ($block, $xml) = @_;
 
-- 
2.30.2