? TextEscaping.diff
Index: Formatter.java
===================================================================
RCS file: /home/cvspublic/jakarta-lucene-sandbox/contributions/highlighter/src/java/org/apache/lucene/search/highlight/Formatter.java,v
retrieving revision 1.3
diff -u -r1.3 Formatter.java
--- Formatter.java	26 Jul 2004 20:39:47 -0000	1.3
+++ Formatter.java	3 Feb 2005 15:26:01 -0000
@@ -30,4 +30,10 @@
 	 * @return
 	 */
 	String highlightTerm(String originalText, TokenGroup tokenGroup);
+
+	/**
+	 * @param originalText The section of non highlight text being output
+	 * @return
+	 */
+	String encodeText(String originalText);
 }
Index: GradientFormatter.java
===================================================================
RCS file: /home/cvspublic/jakarta-lucene-sandbox/contributions/highlighter/src/java/org/apache/lucene/search/highlight/GradientFormatter.java,v
retrieving revision 1.4
diff -u -r1.4 GradientFormatter.java
--- GradientFormatter.java	7 Jan 2005 22:34:24 -0000	1.4
+++ GradientFormatter.java	3 Feb 2005 15:26:01 -0000
@@ -117,11 +117,11 @@
     public String highlightTerm(String originalText, TokenGroup tokenGroup)
     {
         if (tokenGroup.getTotalScore() == 0)
-            return originalText;
+            return encodeText(originalText);
         float score = tokenGroup.getTotalScore();
         if (score == 0)
         {
-            return originalText;
+            return encodeText(originalText);
         }
         StringBuffer sb = new StringBuffer();
         sb.append("<font ");
@@ -138,11 +138,16 @@
             sb.append("\" ");
         }
         sb.append(">");
-        sb.append(originalText);
+        sb.append(encodeText(originalText));
         sb.append("</font>");
         return sb.toString();
     }
 
+    public String encodeText(String originalText)
+    {
+        return SimpleHTMLFormatter.htmlEncode(originalText);
+    }
+
     protected String getForegroundColorString(float score)
     {
         int rVal = getColorVal(fgRMin, fgRMax, score);
Index: Highlighter.java
===================================================================
RCS file: /home/cvspublic/jakarta-lucene-sandbox/contributions/highlighter/src/java/org/apache/lucene/search/highlight/Highlighter.java,v
retrieving revision 1.3
diff -u -r1.3 Highlighter.java
--- Highlighter.java	12 Aug 2004 22:09:37 -0000	1.3
+++ Highlighter.java	3 Feb 2005 15:26:01 -0000
@@ -163,7 +163,7 @@
 					String markedUpText=formatter.highlightTerm(tokenText, tokenGroup);
 					//store any whitespace etc from between this and last group
 					if (startOffset > lastEndOffset)
-						newText.append(text.substring(lastEndOffset, startOffset));
+						newText.append(formatter.encodeText(text.substring(lastEndOffset, startOffset)));
 					newText.append(markedUpText);
 					lastEndOffset=endOffset;
 					tokenGroup.clear();
@@ -198,14 +198,14 @@
 				String markedUpText=formatter.highlightTerm(tokenText, tokenGroup);
 				//store any whitespace etc from between this and last group
 				if (startOffset > lastEndOffset)
-					newText.append(text.substring(lastEndOffset, startOffset));
+					newText.append(formatter.encodeText(text.substring(lastEndOffset, startOffset)));
 				newText.append(markedUpText);
 				lastEndOffset=endOffset;						
 			}
 
 			// append text after end of last token
 			if (lastEndOffset < text.length())
-				newText.append(text.substring(lastEndOffset));
+				newText.append(formatter.encodeText(text.substring(lastEndOffset)));
 
 			currentFrag.textEndPos = newText.length();
 
Index: SimpleHTMLFormatter.java
===================================================================
RCS file: /home/cvspublic/jakarta-lucene-sandbox/contributions/highlighter/src/java/org/apache/lucene/search/highlight/SimpleHTMLFormatter.java,v
retrieving revision 1.2
diff -u -r1.2 SimpleHTMLFormatter.java
--- SimpleHTMLFormatter.java	26 Jul 2004 20:39:47 -0000	1.2
+++ SimpleHTMLFormatter.java	3 Feb 2005 15:26:01 -0000
@@ -52,10 +52,57 @@
 		{
 			returnBuffer=new StringBuffer();
 			returnBuffer.append(preTag);
-			returnBuffer.append(originalText);
+			returnBuffer.append(encodeText(originalText));
 			returnBuffer.append(postTag);
 			return returnBuffer.toString();
 		}
-		return originalText;
+		return encodeText(originalText);
+	}
+	
+	public String encodeText(String originalText)
+	{
+		return htmlEncode(originalText);
+	}
+	
+	/**
+	 * Encode string into HTML
+	 */
+	public final static String htmlEncode(String plainText) 
+	{
+		if (plainText == null)
+		{
+			return "";
+		}
+
+		StringBuffer result = new StringBuffer();
+
+		for (int index=0; index<plainText.length(); index++) 
+		{
+			char ch = plainText.charAt(index);
+
+			switch (ch) 
+			{
+			case '"':
+				result.append("&quot;");
+				break;
+
+			case '&':
+				result.append("&amp;");
+				break;
+
+			case '<':
+				result.append("&lt;");
+				break;
+
+			case '>':
+				result.append("&gt;");
+				break;
+
+			default:
+				result.append(ch);
+			}
+		}
+
+		return result.toString();
 	}
 }
Index: SpanGradientFormatter.java
===================================================================
RCS file: /home/cvspublic/jakarta-lucene-sandbox/contributions/highlighter/src/java/org/apache/lucene/search/highlight/SpanGradientFormatter.java,v
retrieving revision 1.1
diff -u -r1.1 SpanGradientFormatter.java
--- SpanGradientFormatter.java	7 Jan 2005 22:35:15 -0000	1.1
+++ SpanGradientFormatter.java	3 Feb 2005 15:26:02 -0000
@@ -43,11 +43,11 @@
 	public String highlightTerm(String originalText, TokenGroup tokenGroup)
     {
         if (tokenGroup.getTotalScore() == 0)
-            return originalText;
+            return encodeText(originalText);
         float score = tokenGroup.getTotalScore();
         if (score == 0)
         {
-            return originalText;
+            return encodeText(originalText);
         }
 
 		// try to size sb correctly
@@ -67,7 +67,7 @@
 			sb.append("; ");
 		}
 		sb.append("\">");
-		sb.append(originalText);
+		sb.append(encodeText(originalText));
 		sb.append("</span>");
         return sb.toString();
     }