--- nutch-0.7/src/test/org/apache/nutch/util/TestStringUtil.java	2005-08-17 03:23:42.000000000 -0700
+++ nutch-0.7-patched/src/test/org/apache/nutch/util/TestStringUtil.java	2005-10-14 11:06:49.000000000 -0700
@@ -58,4 +58,10 @@
 
   }
 
+  public void testToValidXmlText() {
+      String tmp = StringUtil.toValidXmlText("1 < 2");
+      assertEquals("Failed escape less-than: " + tmp, tmp, "1 &lt; 2");
+      tmp = StringUtil.toValidXmlText("x\u000Cx");
+      assertEquals("Failed to drop illegal 0xC character: " + tmp, tmp, "xx");
+  }
 }
--- nutch-0.7/src/java/org/apache/nutch/util/StringUtil.java	2005-08-17 03:23:42.000000000 -0700
+++ nutch-0.7-patched/src/java/org/apache/nutch/util/StringUtil.java	2005-10-14 10:37:18.000000000 -0700
@@ -109,6 +109,146 @@
            (String) encodingAliases.get(canonicalName) : canonicalName; 
   }
 
+  /**
+   * Escapes a string so that it can be safely put into an XML text node.
+   * Please note that some characters cannot be serialized into an XML text
+   * (Such characters are dropped from the String returned). Refer to
+   * <a href="http://www.w3.org/TR/2000/REC-xml-20001006#charsets">XML
+   * specification</a> for more information.
+   *
+   * @param str The string to be escaped.
+   * <code>IllegalArgumentException</code> is thrown when an unescapable
+   * sequence of characters is encountered. Otherwise, the offending
+   * characters will be omitted in the output.
+   * @return A string that is safe to use in an XML element or attribute. The
+   * xml 5 'special characters' are entity encoded if present and characters
+   * outside of the legal range for xml documents will have been removed. 
+   * @author Dawid Weiss
+   */
+  public static String toValidXmlText(final String str)
+  {
+    return toValidXmlText(str, false);
+  }
+
+  /**
+   * Escapes a string so that it can be safely put into an XML text node.
+   * Please note that some characters cannot be serialized into an XML text.
+   * Refer to <a href="http://www.w3.org/TR/2000/REC-xml-20001006#charsets">XML
+   * specification</a> for more information.
+   *
+   * @param str The string to be escaped.
+   * @param exceptionOnUnescapable If true, 
+   * <code>IllegalArgumentException</code> is thrown when an unescapable
+   * sequence of characters is encountered. Otherwise, the offending
+   * characters will be omitted in the output.
+   * @return A string that is safe to use in an XML element or attribute. The
+   * xml 5 'special characters' are entity encoded if present and characters
+   * outside of the legal range for xml documents will have been removed 
+   * (if <code>exceptionOnUnescapable</code> is true.
+   * @author Dawid Weiss
+   */
+  public static String toValidXmlText(final String str,
+        final boolean exceptionOnUnescapable)
+  {
+      StringBuffer buffer = null;
+    
+      for (int i = 0; i < str.length(); i++)
+      {
+          char ch = str.charAt(i);
+          String entity;
+    
+          switch (ch)
+          {
+              case '<': // '<'
+                  entity = "&lt;";
+    
+                  break;
+    
+              case '>': // '>'
+                  entity = "&gt;";
+    
+                  break;
+    
+              case '&': // '&'
+                  entity = "&amp;";
+    
+                  break;
+    
+              case '\'':
+                  entity = "&apos;";
+    
+                  break;
+    
+              case '"':
+                  entity = "&quot;";
+    
+                  break;
+    
+              case 0x09: // valid xml characters
+              case 0x0a:
+              case 0x0d:
+                  entity = null;
+    
+                  break;
+    
+              default:
+    
+                  // check if valid XML characters
+                  if (
+                      ((ch >= 0x20) && (ch <= 0xD7FF)) ||
+                      ((ch >= 0xe000) && (ch <= 0xfffd)) ||
+                      ((ch >= 0x10000) && (ch <= 0x10ffff))
+                  )
+                  {
+                      entity = null;
+    
+                      break;
+                  }
+                  else
+                  {
+                      if (exceptionOnUnescapable)
+                      {
+                          throw new IllegalArgumentException(
+                              "Character is not within valid XML characters " +
+                              "(code: 0x" + Integer.toHexString(ch) +
+                              ", position: " + i + ")."
+                          );
+                      }
+                      else
+                      {
+                          // replace the character with an empty string.
+                          entity = "";
+    
+                          break;
+                      }
+                  }
+          }
+    
+          if (buffer == null)
+          {
+              if (entity != null)
+              {
+                  buffer = new StringBuffer(str.length() + 20);
+                  buffer.append(str.substring(0, i));
+                  buffer.append(entity);
+              }
+          }
+          else
+          {
+              if (entity == null)
+              {
+                  buffer.append(ch);
+              }
+              else
+              {
+                  buffer.append(entity);
+              }
+          }
+      }
+    
+      return (buffer != null) ? buffer.toString() : str;
+  }
+
   public static void main(String[] args) {
     if (args.length != 1)
       System.out.println("Usage: StringUtil <encoding name>");
--- nutch-0.7/src/java/org/apache/nutch/searcher/OpenSearchServlet.java	2005-08-17 03:23:42.000000000 -0700
+++ nutch-0.7-patched/src/java/org/apache/nutch/searcher/OpenSearchServlet.java	2005-10-14 11:04:28.000000000 -0700
@@ -15,6 +15,7 @@
  */
 
 package org.apache.nutch.searcher;
+import org.apache.nutch.util.StringUtil;
 
 import java.io.IOException;
 import java.net.URLEncoder;
@@ -252,14 +253,14 @@
   private static void addNode(Document doc, Node parent,
                               String name, String text) {
     Element child = doc.createElement(name);
-    child.appendChild(doc.createTextNode(text));
+    child.appendChild(doc.createTextNode(StringUtil.toValidXmlText(text)));
     parent.appendChild(child);
   }
 
   private static void addNode(Document doc, Node parent,
                               String ns, String name, String text) {
     Element child = doc.createElementNS((String)NS_MAP.get(ns), ns+":"+name);
-    child.appendChild(doc.createTextNode(text));
+    child.appendChild(doc.createTextNode(StringUtil.toValidXmlText(text)));
     parent.appendChild(child);
   }
 
@@ -269,6 +270,5 @@
     attribute.setValue(value);
     node.getAttributes().setNamedItem(attribute);
   }
-
 }