kinman      2003/11/24 15:51:10

  Modified:    jasper2/src/share/org/apache/jasper/compiler
                        JspDocumentParser.java
  Log:
  - Fix 24779: Jasper fails to compile a valid JSP document (xml syntax) when
    an EL string in template text contains a "[]".  The fix to to buffer and
    concatenate template text strings and only start processing the texts when
    they are all in one piece.
  
    This, BTW, also fixes the bug that somestimes trims the white spaces that
    trails template texts.
  
  Revision  Changes    Path
  1.72      +42 -15    
jakarta-tomcat-jasper/jasper2/src/share/org/apache/jasper/compiler/JspDocumentParser.java
  
  Index: JspDocumentParser.java
  ===================================================================
  RCS file: 
/home/cvs/jakarta-tomcat-jasper/jasper2/src/share/org/apache/jasper/compiler/JspDocumentParser.java,v
  retrieving revision 1.71
  retrieving revision 1.72
  diff -u -r1.71 -r1.72
  --- JspDocumentParser.java    22 Nov 2003 02:09:34 -0000      1.71
  +++ JspDocumentParser.java    24 Nov 2003 23:51:10 -0000      1.72
  @@ -90,6 +90,7 @@
    * syntax.
    *
    * @author Jan Luehe
  + * @author Kin-man Chung
    */
   
   class JspDocumentParser
  @@ -110,6 +111,7 @@
       private JspCompilationContext ctxt;
       private PageInfo pageInfo;
       private String path;
  +    private StringBuffer charBuffer;
   
       // Node representing the XML element currently being parsed
       private Node current;
  @@ -287,6 +289,8 @@
           AttributesImpl nonTaglibAttrs = null;
           AttributesImpl nonTaglibXmlnsAttrs = null;
   
  +        processChars();
  +
           checkPrefixes(uri, qName, attrs);
   
           if (directivesOnly &&
  @@ -399,14 +403,33 @@
       /*
        * Receives notification of character data inside an element.
        *
  +     * The SAX does not call this method with all of the template text, but may
  +     * invoke this method with chunks of it.  This is a problem when we try
  +     * to determine if the text contains only whitespaces, or when we are
  +     * looking for an EL expression string.  Therefore it is necessary to
  +     * buffer and concatenate the chunks and process the concatenated text 
  +     * later (at beginTag and endTag)
  +     *
        * @param buf The characters
        * @param offset The start position in the character array
        * @param len The number of characters to use from the character array
        *
        * @throws SAXException
        */
  -    public void characters(char[] buf, int offset, int len)
  -        throws SAXException {
  +    public void characters(char[] buf, int offset, int len) {
  +
  +        if (charBuffer == null) {
  +            charBuffer = new StringBuffer();
  +        }
  +        charBuffer.append(buf, offset, len);
  +    }
  +
  +    private void processChars() throws SAXException {
  +
  +        if (charBuffer == null) {
  +            return;
  +        }
  +
           /*
            * JSP.6.1.1: All textual nodes that have only white space are to be
            * dropped from the document, except for nodes in a jsp:text element,
  @@ -418,11 +441,11 @@
           boolean isAllSpace = true;
           if (!(current instanceof Node.JspText)
               && !(current instanceof Node.NamedAttribute)) {
  -            for (int i = offset; i < offset + len; i++) {
  -                if (!(buf[i] == ' '
  -                    || buf[i] == '\n'
  -                    || buf[i] == '\r'
  -                    || buf[i] == '\t')) {
  +            for (int i = 0; i < charBuffer.length(); i++) {
  +                if (!(charBuffer.charAt(i) == ' '
  +                    || charBuffer.charAt(i) == '\n'
  +                    || charBuffer.charAt(i) == '\r'
  +                    || charBuffer.charAt(i) == '\t')) {
                       isAllSpace = false;
                       break;
                   }
  @@ -436,10 +459,10 @@
               int column = startMark.getColumnNumber();
   
               CharArrayWriter ttext = new CharArrayWriter();
  -            int limit = offset + len;
               int lastCh = 0;
  -            for (int i = offset; i < limit; i++) {
  -                int ch = buf[i];
  +            for (int i = 0; i < charBuffer.length(); i++) {
  +
  +                int ch = charBuffer.charAt(i);
                   if (ch == '\n') {
                       column = 1;
                       line++;
  @@ -463,7 +486,7 @@
                       boolean doubleQ = false;
                       lastCh = 0;
                       for (;; i++) {
  -                        if (i >= limit) {
  +                        if (i >= charBuffer.length()) {
                               throw new SAXParseException(
                                   Localizer.getMessage(
                                       "jsp.error.unterminated",
  @@ -471,7 +494,7 @@
                                   locator);
   
                           }
  -                        ch = buf[i];
  +                        ch = charBuffer.charAt(i);
                           if (ch == '\n') {
                               column = 1;
                               line++;
  @@ -519,6 +542,8 @@
           }
           startMark =
               new Mark(path, locator.getLineNumber(), locator.getColumnNumber());
  +
  +        charBuffer = null;
       }
   
       /*
  @@ -526,6 +551,8 @@
        */
       public void endElement(String uri, String localName, String qName)
           throws SAXException {
  +
  +        processChars();
   
           if (directivesOnly &&
               !(JSP_URI.equals(uri) && localName.startsWith(DIRECTIVE_ACTION))) {
  
  
  

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to