- Don't let page number fields through.

2 files changed, 25 insertions, 1 deletions

diff --git a/ChangeLog b/ChangeLog
index a7daeb3..b152bf1 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,9 +1,9 @@
 Version 0.9.7
-
  - Don't let most characters below 0x0020 through. Old RTF's seem
    to have these for strange reasons.
  - Send data in chunks so massive amounts of picture data don't
    slow down the conversion.
+ - Don't let page number fields through.
 
 Version 0.9.6
  - Add character styles
diff --git a/src/xmlcomposer.cpp b/src/xmlcomposer.cpp
index 6a731ff..e1ae89e 100644
--- a/src/xmlcomposer.cpp
+++ b/src/xmlcomposer.cpp
@@ -79,6 +79,22 @@ void tokenize(const string& str, StringArray& tokens,
     }
 }
 
+void trim(string& str, const char* whitespace = " \t")
+{
+	string::size_type pos = str.find_last_not_of(whitespace);
+	if(pos != string::npos)
+	{
+		str.erase(pos + 1);
+		pos = str.find_first_not_of(whitespace);
+		if(pos != string::npos)
+			str.erase(0, pos);
+	}
+	else
+	{
+		str.erase(str.begin(), str.end());
+	}
+}
+
 /* ----------------------------------------------------------------------------------
  *  CONSTRUCTION
  */
@@ -1074,6 +1090,7 @@ ON_CONTROLWORD(Info)
 // Field Analyser -------------------------------------------------------------------
 
 static const string kFieldHyperlink = "HYPERLINK";
+static const string kFieldPage = "PAGE";
 
 ON_CONTROLWORD(Field)
 {
@@ -1086,6 +1103,7 @@ ON_CONTROLWORD(Field)
 	{
 		StringArray codes;
 		tokenize(fieldCode, codes, " ", '"');
+		trim(fieldCode);
 
 		if(codes.size() >= 2 && codes[0].compare(kFieldHyperlink) == 0)
 		{
@@ -1106,6 +1124,12 @@ ON_CONTROLWORD(Field)
 				AN_ATTRIBUTE(kAtTo, codes[1]);
 			}
 		}
+
+		// Page numbers don't make sense, don't let through
+		else if(fieldCode.compare(kFieldPage) == 0)
+		{
+			AN_DESTINATION(Null);
+		}
 	}
 
 	// The internal link data