1 /*

     2 * Copyright (c) 2009 Nokia Corporation and/or its subsidiary(-ies).

     3 * All rights reserved.

     4 * This component and the accompanying materials are made available

     5 * under the terms of "Eclipse Public License v1.0"

     6 * which accompanies this distribution, and is available

     7 * at the URL "http://www.eclipse.org/legal/epl-v10.html".

     8 *

     9 * Initial Contributors:

    10 * Nokia Corporation - initial contribution.

    11 *

    12 * Contributors:

    13 *

    14 * Description: 

    15 *

    16 */

    17 /*

    18  * Licensed to the Apache Software Foundation (ASF) under one or more

    19  * contributor license agreements.  See the NOTICE file distributed with

    20  * this work for additional information regarding copyright ownership.

    21  * The ASF licenses this file to You under the Apache License, Version 2.0

    22  * (the "License"); you may not use this file except in compliance with

    23  * the License.  You may obtain a copy of the License at

    24  * 

    25  *      http://www.apache.org/licenses/LICENSE-2.0

    26  * 

    27  * Unless required by applicable law or agreed to in writing, software

    28  * distributed under the License is distributed on an "AS IS" BASIS,

    29  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

    30  * See the License for the specific language governing permissions and

    31  * limitations under the License.

    32  */

    33 

    34  /*

    35  * $Id: XMLDocumentHandler.hpp 568078 2007-08-21 11:43:25Z amassari $

    36  */

    37 

    38 

    39 #if !defined(XMLDOCUMENTHANDLER_HPP)

    40 #define XMLDOCUMENTHANDLER_HPP

    41 

    42 #include <xercesc/util/XercesDefs.hpp>

    43 #include <xercesc/util/RefVectorOf.hpp>

    44 #include <xercesc/framework/XMLAttr.hpp>

    45 

    46 XERCES_CPP_NAMESPACE_BEGIN

    47 

    48 class XMLElementDecl;

    49 class XMLEntityDecl;

    50 

    51 /**

    52   * This abstract class provides the interface for the scanner to return

    53   * XML document information up to the parser as it scans through the

    54   * document.

    55   *

    56   * The interface is very similar to org.sax.DocumentHandler, but

    57   * has some extra methods required to get all the data out.

    58   */

    59 class XMLPARSER_EXPORT XMLDocumentHandler

    60 {

    61 public:

    62     // -----------------------------------------------------------------------

    63     //  Constructors are hidden, just the virtual destructor is exposed

    64     // -----------------------------------------------------------------------

    65     /** @name Destructor */

    66     //@{

    67     virtual ~XMLDocumentHandler()

    68     {

    69     }

    70     //@}

    71 

    72     /** @name The document handler interface */

    73     //@{

    74     /** Receive notification of character data.

    75       *

    76       * <p>The scanner will call this method to report each chunk of

    77       * character data. The scanner may return all contiguous character

    78       * data in a single chunk, or they may split it into several

    79       * chunks; however, all of the characters in any single event

    80       * will come from the same external entity, so that the Locator

    81       * provides useful information.</p>

    82       *

    83       * <p>The parser must not attempt to read from the array

    84       * outside of the specified range.</p>

    85       *

    86       * @param  chars           The content (characters) between markup from the XML

    87       *                         document.

    88       * @param  length          The number of characters to read from the array.

    89       * @param  cdataSection    Indicates that this data is inside a CDATA

    90       *                         section.

    91       * @see #ignorableWhitespace

    92       * @see Locator

    93       */

    94     virtual void docCharacters

    95     (

    96         const   XMLCh* const    chars

    97         , const unsigned int    length

    98         , const bool            cdataSection

    99     ) = 0;

   100 

   101     /** Receive notification of comments in the XML content being parsed.

   102       *

   103       * This scanner will call this method for any comments found in the

   104       * content of the document.

   105       *

   106       * @param comment The text of the comment.

   107       */

   108     virtual void docComment

   109     (

   110         const   XMLCh* const    comment

   111     ) = 0;

   112 

   113     /** Receive notification of PI's parsed in the XML content.

   114       *

   115       * The scanner will call this method for any PIs it finds within the

   116       * content of the document.

   117       *

   118       * @param  target  The name of the PI.

   119       * @param  data    The body of the PI. This may be an empty string since

   120       *                 the body is optional.

   121       */

   122     virtual void docPI

   123     (

   124         const   XMLCh* const    target

   125         , const XMLCh* const    data

   126     ) = 0;

   127 

   128     /** Receive notification after the scanner has parsed the end of the

   129       * document.

   130       *

   131       * The scanner will call this method when the current document has been

   132       * fully parsed. The handler may use this opportunity to do something with

   133       * the data, clean up temporary data, etc...

   134       */

   135     virtual void endDocument() = 0;

   136 

   137     /** Receive notification of the end of an element.

   138       *

   139       * This method is called when scanner encounters the end of element tag.

   140       * There will be a corresponding startElement() event for every

   141       * endElement() event, but not necessarily the other way around. For

   142       * empty tags, there is only a startElement() call.

   143       *

   144       * @param  elemDecl The name of the element whose end tag was just

   145       *                     parsed.

   146       * @param  uriId       The ID of the URI in the URI pool (only valid if

   147       *                     name spaces is enabled)

   148       * @param  isRoot      Indicates if this is the root element.

   149       * @param  prefixName  The string representing the prefix name

   150       */

   151     virtual void endElement

   152     (

   153         const   XMLElementDecl& elemDecl

   154         , const unsigned int    uriId

   155         , const bool            isRoot

   156         , const XMLCh* const    prefixName = 0

   157     ) = 0;

   158 

   159     /** Receive notification when a referenced entity's content ends

   160       *

   161       * This method is called when scanner encounters the end of an entity

   162       * reference.

   163       *

   164       * @param  entDecl  The name of the entity reference just scanned.

   165       */

   166     virtual void endEntityReference

   167     (

   168         const   XMLEntityDecl&  entDecl

   169     ) = 0;

   170 

   171     /** Receive notification of ignorable whitespace in element content.

   172       *

   173       * <p>Validating Parsers must use this method to report each chunk

   174       * of ignorable whitespace (see the W3C XML 1.0 recommendation,

   175       * section 2.10): non-validating parsers may also use this method

   176       * if they are capable of parsing and using content models.</p>

   177       *

   178       * <p>The scanner may return all contiguous whitespace in a single

   179       * chunk, or it may split it into several chunks; however, all of

   180       * the characters in any single event will come from the same

   181       * external entity, so that the Locator provides useful

   182       * information.</p>

   183       *

   184       * <p>The parser must not attempt to read from the array

   185       * outside of the specified range.</p>

   186       *

   187       * @param  chars       The whitespace characters from the XML document.

   188       * @param  length      The number of characters to read from the array.

   189       * @param  cdataSection Indicates that this data is inside a CDATA

   190       *                     section.

   191       * @see #characters

   192       */

   193     virtual void ignorableWhitespace

   194     (

   195         const   XMLCh* const    chars

   196         , const unsigned int    length

   197         , const bool            cdataSection

   198     ) = 0;

   199 

   200     /** Reset the document handler's state, if required

   201       *

   202       * This method is used to give the registered document handler a

   203       * chance to reset itself. Its called by the scanner at the start of

   204       * every parse.

   205       */

   206     virtual void resetDocument() = 0;

   207 

   208     /** Receive notification of the start of a new document

   209       *

   210       * This method is the first callback called the scanner at the

   211       * start of every parse. This is before any content is parsed.

   212       */

   213     virtual void startDocument() = 0;

   214 

   215     /** Receive notification of a new start tag

   216       *

   217       * This method is called when scanner encounters the start of an element tag.

   218       * All elements must always have a startElement() tag. Empty tags will

   219       * only have the startElement() tag and no endElement() tag.

   220       *

   221       * @param  elemDecl The name of the element whose start tag was just

   222       *                     parsed.

   223       * @param  uriId       The ID of the URI in the URI pool (only valid if

   224       *                     name spaces is enabled)

   225       * @param  prefixName  The string representing the prefix name

   226       * @param  attrList    List of attributes in the element

   227       * @param  attrCount   Count of the attributes in the element

   228       * @param  isEmpty     Indicates if the element is empty, in which case

   229       *                     you should not expect an endElement() event.

   230       * @param  isRoot      Indicates if this is the root element.

   231       */

   232     virtual void startElement

   233     (

   234         const   XMLElementDecl&         elemDecl

   235         , const unsigned int            uriId

   236         , const XMLCh* const            prefixName

   237         , const RefVectorOf<XMLAttr>&   attrList

   238         , const unsigned int            attrCount

   239         , const bool                    isEmpty

   240         , const bool                    isRoot

   241     ) = 0;

   242 

   243     /** Receive notification when the scanner hits an entity reference.

   244       *

   245       * This is currently useful only to DOM parser configurations as SAX

   246       * does not provide any api to return this information.

   247       *

   248       * @param  entDecl  The name of the entity that was referenced.

   249       */

   250     virtual void startEntityReference(const XMLEntityDecl& entDecl) = 0;

   251 

   252     /** Receive notification of an XML declaration

   253       *

   254       * Currently neither DOM nor SAX provide API's to return back this

   255       * information.

   256       *

   257       * @param  versionStr      The value of the <code>version</code> pseudoattribute

   258       *                         of the XML decl.

   259       * @param  encodingStr     The value of the <code>encoding</code> pseudoattribute

   260       *                         of the XML decl.

   261       * @param  standaloneStr   The value of the <code>standalone</code>

   262       *                         pseudoattribute of the XML decl.

   263       * @param  autoEncodingStr The encoding string auto-detected by the

   264       *                         scanner. In absence of any 'encoding' attribute in the

   265       *                         XML decl, the XML standard specifies how a parser can

   266       *                         auto-detect. If there is no <code>encodingStr</code>

   267       *                         this is what will be used to try to decode the file.

   268       */

   269     virtual void XMLDecl

   270     (

   271         const   XMLCh* const    versionStr

   272         , const XMLCh* const    encodingStr

   273         , const XMLCh* const    standaloneStr

   274         , const XMLCh* const    autoEncodingStr

   275     ) = 0;

   276 

   277     /** Receive notification of the name and namespace of the type that validated 

   278       * the element corresponding to the most recent endElement event.

   279       * This event will be fired immediately after the

   280       * endElement() event that signifies the end of the element

   281       * to which it applies; no other events will intervene.

   282       * This method is <em>EXPERIMENTAL</em> and may change, disappear 

   283       * or become pure virtual at any time.

   284       *

   285       * This corresponds to a part of the information required by DOM Core

   286       * level 3's TypeInfo interface.

   287       *

   288       * @param  typeName        local name of the type that actually validated

   289       *                         the content of the element corresponding to the

   290       *                         most recent endElement() callback

   291       * @param  typeURI         namespace of the type that actually validated

   292       *                         the content of the element corresponding to the

   293       *                         most recent endElement() callback

   294       * @deprecated

   295       */

   296     virtual void elementTypeInfo

   297     (

   298         const   XMLCh* const /* typeName */

   299         , const XMLCh* const /* typeURI */

   300     ) { /* non pure virtual to permit backward compatibility of implementations.  */  };

   301     //@}

   302 

   303 

   304 

   305 protected :

   306     // -----------------------------------------------------------------------

   307     //  Hidden Constructors

   308     // -----------------------------------------------------------------------

   309     XMLDocumentHandler()

   310     {

   311     }

   312 

   313 

   314 private:

   315     // -----------------------------------------------------------------------

   316     //  Unimplemented constructors and operators

   317     // -----------------------------------------------------------------------

   318     XMLDocumentHandler(const XMLDocumentHandler&);

   319     XMLDocumentHandler& operator=(const XMLDocumentHandler&);

   320 };

   321 

   322 XERCES_CPP_NAMESPACE_END

   323 

   324 #endif