RtfParser


1   /**
2    * $Id: RtfParser.java 2377 2006-09-15 07:51:05Z xlv $
3    * $Name$
4    *
5    * Copyright 2006 by Mark Hall
6    *
7    * The contents of this file are subject to the Mozilla Public License Version 1.1
8    * (the "License"); you may not use this file except in compliance with the License.
9    * You may obtain a copy of the License at http://www.mozilla.org/MPL/
10   *
11   * Software distributed under the License is distributed on an "AS IS" basis,
12   * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
13   * for the specific language governing rights and limitations under the License.
14   *
15   * The Original Code is 'iText, a free JAVA-PDF library'.
16   *
17   * The Initial Developer of the Original Code is Bruno Lowagie. Portions created by
18   * the Initial Developer are Copyright (C) 1999-2006 by Bruno Lowagie.
19   * All Rights Reserved.
20   * Co-Developer of the code is Paulo Soares. Portions created by the Co-Developer
21   * are Copyright (C) 2000-2006 by Paulo Soares. All Rights Reserved.
22   *
23   * Contributor(s): all the names of the contributors are added in the source code
24   * where applicable.
25   *
26   * Alternatively, the contents of this file may be used under the terms of the
27   * LGPL license (the ?GNU LIBRARY GENERAL PUBLIC LICENSE?), in which case the
28   * provisions of LGPL are applicable instead of those above.  If you wish to
29   * allow use of your version of this file only under the terms of the LGPL
30   * License and not to allow others to use your version of this file under
31   * the MPL, indicate your decision by deleting the provisions above and
32   * replace them with the notice and other provisions required by the LGPL.
33   * If you do not delete the provisions above, a recipient may use your version
34   * of this file under either the MPL or the GNU LIBRARY GENERAL PUBLIC LICENSE.
35   *
36   * This library is free software; you can redistribute it and/or modify it
37   * under the terms of the MPL as stated above or under the terms of the GNU
38   * Library General Public License as published by the Free Software Foundation;
39   * either version 2 of the License, or any later version.
40   *
41   * This library is distributed in the hope that it will be useful, but WITHOUT
42   * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
43   * FOR A PARTICULAR PURPOSE. See the GNU Library general Public License for more
44   * details.
45   *
46   * If you didn't download this code from the following link, you should check if
47   * you aren't using an obsolete version:
48   * http://www.lowagie.com/iText/
49   */
50  package com.lowagie.text.rtf.direct;
51  
52  import java.awt.Color  ;
53  import java.io.IOException  ;
54  import java.io.Reader  ;
55  import java.util.Iterator  ;
56  
57  import com.lowagie.text.rtf.document.RtfDocument;
58  
59  /**
60   * The RtfParser allows the importing of RTF documents or
61   * RTF document fragments. The RTF document or fragment is tokenised,
62   * font and color definitions corrected and then added to
63   * the document being written.
64   * 
65   * @version $Revision: 2377 $
66   * @author Mark Hall (mhall@edu.uni-klu.ac.at)
67   */
68  public class RtfParser {
69      /**
70       * Currently the RTF document header is being parsed.
71       */
72      private static final int PARSER_IN_HEADER = 0;
73      /**
74       * Currently the RTF font table is being parsed.
75       */
76      private static final int PARSER_IN_FONT_TABLE = 1;
77      /**
78       * Currently the RTF color table is being parsed.
79       */
80      private static final int PARSER_IN_COLOR_TABLE = 2;
81      /**
82       * Currently the RTF info group is being parsed.
83       */
84      private static final int PARSER_IN_INFO_GROUP = 4;
85      /**
86       * Currently the RTF document content is being parsed.
87       */
88      private static final int PARSER_IN_DOCUMENT = 8;
89      
90      /**
91       * The RtfDocument to add the RTF document or fragment to.
92       */
93      private RtfDocument rtfDoc = null;
94      /**
95       * The RtfTokeniser to use for tokenising the RTF document or fragment.
96       */
97      private RtfTokeniser tokeniser = null;
98      /**
99       * The RtfImportHeader to store imported font and color mappings in.
100      */
101     private RtfImportHeader importHeader = null;
102     /**
103      * The RtfFontTableParser to use for parsing the font table.
104      */
105     private RtfFontTableParser fontTableParser = null;
106     /**
107      * The RtfColorTableParser to use for parsing the color table.
108      */
109     private RtfColorTableParser colorTableParser = null;
110     /**
111      * The current parser state.
112      */
113     private int state = PARSER_IN_HEADER;
114     
115     /**
116      * Imports a complete RTF document.
117      * 
118      * @param reader The Reader to read the RTF document from.
119      * @param rtfDoc The RtfDocument to add the imported document to.
120      * @throws IOException On I/O errors.
121      */
122     public void importRtfDocument(Reader   reader, RtfDocument rtfDoc) throws IOException   {
123         this.rtfDoc = rtfDoc;
124         this.state = PARSER_IN_HEADER;
125         this.importHeader = new RtfImportHeader(this.rtfDoc);
126         this.fontTableParser = new RtfFontTableParser(this.importHeader);
127         this.colorTableParser = new RtfColorTableParser(this.importHeader);
128         this.tokeniser = new RtfTokeniser(this, 0);
129         this.tokeniser.tokenise(reader);
130     }
131     
132     /**
133      * Imports an RTF fragment.
134      * 
135      * @param reader The Reader to read the RTF fragment from.
136      * @param rtfDoc The RTF document to add the RTF fragment to.
137      * @param importMappings The RtfImportMappings defining font and color mappings for the fragment.
138      * @throws IOException On I/O errors.
139      */
140     public void importRtfFragment(Reader   reader, RtfDocument rtfDoc, RtfImportMappings importMappings) throws IOException   {
141         this.rtfDoc = rtfDoc;
142         this.state = PARSER_IN_DOCUMENT;
143         this.importHeader = new RtfImportHeader(this.rtfDoc);
144         this.fontTableParser = new RtfFontTableParser(this.importHeader);
145         this.colorTableParser = new RtfColorTableParser(this.importHeader);
146         handleImportMappings(importMappings);
147         this.tokeniser = new RtfTokeniser(this, 1);
148         this.tokeniser.tokenise(reader);
149     }
150 
151     /**
152      * Imports the mappings defined in the RtfImportMappings into the
153      * RtfImportHeader of this RtfParser.
154      * 
155      * @param importMappings The RtfImportMappings to import.
156      */
157     private void handleImportMappings(RtfImportMappings importMappings) {
158         Iterator   it = importMappings.getFontMappings().keySet().iterator();
159         while(it.hasNext()) {
160             String   fontNr = (String  ) it.next();
161             this.importHeader.importFont(fontNr, (String  ) importMappings.getFontMappings().get(fontNr));
162         }
163         it = importMappings.getColorMappings().keySet().iterator();
164         while(it.hasNext()) {
165             String   colorNr = (String  ) it.next();
166             this.importHeader.importColor(colorNr, (Color  ) importMappings.getColorMappings().get(colorNr));
167         }
168     }
169     
170     /**
171      * Handles open group tokens.
172      * 
173      * @param groupLevel The current group nesting level.
174      */
175     public void handleOpenGroup(int groupLevel) {
176         if(this.state == PARSER_IN_DOCUMENT) {
177             this.rtfDoc.add(new RtfDirectContent("{"));
178         }
179     }
180     
181     /**
182      * Handles close group tokens. Depending on what is currently
183      * being parsed the parse state may change.
184      * 
185      * @param groupLevel The current group nesting level.
186      */
187     public void handleCloseGroup(int groupLevel) {
188         if(this.state == PARSER_IN_DOCUMENT && groupLevel > 1) {
189             this.rtfDoc.add(new RtfDirectContent("}"));
190         } else if(this.state == PARSER_IN_INFO_GROUP && groupLevel == 2) {
191             this.state = PARSER_IN_DOCUMENT;
192         } else if(this.state == PARSER_IN_FONT_TABLE) {
193             this.fontTableParser.handleCloseGroup(groupLevel);
194             if(groupLevel == 2) {
195                 this.state = PARSER_IN_HEADER;
196             }
197         } else if(this.state == PARSER_IN_COLOR_TABLE) {
198             this.state = PARSER_IN_HEADER;
199         }
200     }
201     
202     /**
203      * Handles single control character tokens.
204      * 
205      * @param ctrlCharacter The control character to handle.
206      * @param groupLevel The current group nesting level.
207      */
208     public void handleCtrlCharacter(String   ctrlCharacter, int groupLevel) {
209         if(this.state == PARSER_IN_DOCUMENT) {
210             this.rtfDoc.add(new RtfDirectContent(ctrlCharacter));
211         }
212     }
213     
214     /**
215      * Handles control word tokens. Depending on the current
216      * state a control word can lead to a state change. When
217      * parsing the actual document contents, The font number,
218      * color number and background color number are remapped.
219      * 
220      * @param ctrlWord The control word to handle.
221      * @param groupLevel The current group nesting level.
222      */
223     public void handleCtrlWord(String   ctrlWord, int groupLevel) {
224         if(this.state == PARSER_IN_DOCUMENT) {
225             if(RtfColorTableParser.stringMatches(ctrlWord, "\\f")) {
226                 ctrlWord = "\\f" + this.importHeader.mapFontNr(ctrlWord.substring(2));
227             } else if(RtfColorTableParser.stringMatches(ctrlWord, "\\cf")) {
228                 ctrlWord = "\\cf" + this.importHeader.mapColorNr(ctrlWord.substring(3));
229             } else if(RtfColorTableParser.stringMatches(ctrlWord, "\\cb")) {
230                 ctrlWord = "\\cb" + this.importHeader.mapColorNr(ctrlWord.substring(3));
231             }
232             this.rtfDoc.add(new RtfDirectContent(ctrlWord));
233         } else if(this.state == PARSER_IN_FONT_TABLE) {
234             this.fontTableParser.handleCtrlWord(ctrlWord, groupLevel);
235         } else if(this.state == PARSER_IN_COLOR_TABLE) {
236             this.colorTableParser.handleCtrlWord(ctrlWord, groupLevel);
237         } else if(this.state == PARSER_IN_HEADER) {
238             if(ctrlWord.equals("\\info")) {
239                 this.state = PARSER_IN_INFO_GROUP;
240             } else if(ctrlWord.equals("\\fonttbl")) {
241                 this.state = PARSER_IN_FONT_TABLE;
242             } else if(ctrlWord.equals("\\colortbl")) {
243                 this.state = PARSER_IN_COLOR_TABLE;
244             }
245         }
246     }
247     
248     /**
249      * Handles text tokens. These are either handed on to the
250      * RtfColorTableParser or RtfFontTableParser or added directly
251      * to the document.
252      * 
253      * @param text The text token to handle.
254      * @param groupLevel The current group nesting level.
255      */
256     public void handleText(String   text, int groupLevel) {
257         if(this.state == PARSER_IN_DOCUMENT) {
258             this.rtfDoc.add(new RtfDirectContent(text));
259         } else if(this.state == PARSER_IN_FONT_TABLE) {
260             this.fontTableParser.handleText(text, groupLevel);
261         } else if(this.state == PARSER_IN_COLOR_TABLE) {
262             this.colorTableParser.handleText(text, groupLevel);
263         }
264     }
265 }
266
A to Z: JavaDoc & Examples Daily Java News & Articles Open Source Projects Open Source Codes Free Computer Books Remove Frame
Popular Tags