archive-access-cvs

[Michael S. <sta...@us...>]

Update of /cvsroot/archive-access/archive-access/projects/nutch/src/java/org/archive/access/nutch
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv9411/src/java/org/archive/access/nutch

Modified Files:
	NutchwaxOpenSearchServlet.java 
Log Message:

* src/java/org/archive/access/nutch/NutchwaxOpenSearchServlet.java 
    Use same code as new version 2 patch that I put up into NUTCH-110.


Index: NutchwaxOpenSearchServlet.java
===================================================================
RCS file: /cvsroot/archive-access/archive-access/projects/nutch/src/java/org/archive/access/nutch/NutchwaxOpenSearchServlet.java,v
retrieving revision 1.4
retrieving revision 1.5
diff -C2 -d -r1.4 -r1.5
*** NutchwaxOpenSearchServlet.java	13 Oct 2005 15:53:36 -0000	1.4
--- NutchwaxOpenSearchServlet.java	15 Oct 2005 01:18:56 -0000	1.5
***************
*** 268,272 ****
                                String name, String text) {
      Element child = doc.createElement(name);
!     child.appendChild(doc.createTextNode(getLegalXml(text)));
      parent.appendChild(child);
    }
--- 268,272 ----
                                String name, String text) {
      Element child = doc.createElement(name);
!     child.appendChild(doc.createTextNode(toValidXmlText(text)));
      parent.appendChild(child);
    }
***************
*** 275,279 ****
                                String ns, String name, String text) {
      Element child = doc.createElementNS((String)NS_MAP.get(ns), ns+":"+name);
!     child.appendChild(doc.createTextNode(getLegalXml(text)));
      parent.appendChild(child);
    }
--- 275,279 ----
                                String ns, String name, String text) {
      Element child = doc.createElementNS((String)NS_MAP.get(ns), ns+":"+name);
!     child.appendChild(doc.createTextNode(toValidXmlText(text)));
      parent.appendChild(child);
    }
***************
*** 282,330 ****
                                     String name, String value) {
      Attr attribute = doc.createAttribute(name);
!     attribute.setValue(getLegalXml(value));
      node.getAttributes().setNamedItem(attribute);
    }
  
!   /*
!    * Ensure string is legal xml.
!    * First look to see if string has illegal characters.  If it doesn't,
!    * just return it.  Otherwise, create new string with illegal characters
!    * @param text String to verify.
!    * @return Passed <code>text</code> or a new string with illegal
!    * characters removed if any found in <code>text</code>.
!    * @see http://www.w3.org/TR/2000/REC-xml-20001006#NT-Char
     */
!   private static String getLegalXml(final String text) {
!       if (text == null) {
!           return null;
!       }
!       boolean allLegal = true;
!       for (int i = 0; i < text.length(); i++) {
!         if (!isLegalXml(text.charAt(i))) {
!             allLegal = false; 
!             break;
!         }
!       }
!       return allLegal? text: createLegalXml(text);
    }
  
!   private static String createLegalXml(final String text) {
!       if (text == null) {
!           return null;
!       }
!       StringBuffer buffer = new StringBuffer(text.length());
!       for (int i = 0; i < text.length(); i++) {
!         char c = text.charAt(i);
!         if (isLegalXml(c)) {
!           buffer.append(c);
!          }
        }
!       return buffer.toString();
!   }
!   
!   private static boolean isLegalXml(final char c) {
!     return c == 0x9 || c == 0xa || c == 0xd || (c >= 0x20 && c <= 0xd7ff)
!         || (c >= 0xe000 && c <= 0xfffd) || (c >= 0x10000 && c <= 0x10ffff);
    }
  }
- 
--- 282,427 ----
                                     String name, String value) {
      Attr attribute = doc.createAttribute(name);
!     attribute.setValue(value);
      node.getAttributes().setNamedItem(attribute);
    }
  
!   /**
!    * 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;
    }
  }