DocumentParser


1   /*
2    * @(#)DocumentParser.java  1.28 03/12/19
3    *
4    * Copyright 2004 Sun Microsystems, Inc. All rights reserved.
5    * SUN PROPRIETARY/CONFIDENTIAL. Use is subject to license terms.
6    */
7   
8   package javax.swing.text.html.parser;
9   
10  import javax.swing.text.SimpleAttributeSet  ;
11  import javax.swing.text.html.HTMLEditorKit  ;
12  import javax.swing.text.html.HTML  ;
13  import javax.swing.text.ChangedCharSetException  ;
14  
15  import java.util.*;
16  import java.io.*;
17  import java.net.*;
18  
19  /**
20   * A Parser for HTML Documents (actually, you can specify a DTD, but
21   * you should really only use this class with the html dtd in swing).
22   * Reads an InputStream of HTML and
23   * invokes the appropriate methods in the ParserCallback class. This
24   * is the default parser used by HTMLEditorKit to parse HTML url's.
25   * <p>This will message the callback for all valid tags, as well as
26   * tags that are implied but not explicitly specified. For example, the
27   * html string (&lt;p&gt;blah) only has a p tag defined. The callback
28   * will see the following methods:
29   * <ol><li><i>handleStartTag(html, ...)</i></li>
30   *     <li><i>handleStartTag(head, ...)</i></li>
31   *     <li><i>handleEndTag(head)</i></li>
32   *     <li><i>handleStartTag(body, ...)</i></li>
33   *     <li>handleStartTag(p, ...)</i></li>
34   *     <li>handleText(...)</li>
35   *     <li><i>handleEndTag(p)</i></li>
36   *     <li><i>handleEndTag(body)</i></li>
37   *     <li><i>handleEndTag(html)</i></li>
38   * </ol>
39   * The items in <i>italic</i> are implied, that is, although they were not
40   * explicitly specified, to be correct html they should have been present
41   * (head isn't necessary, but it is still generated). For tags that
42   * are implied, the AttributeSet argument will have a value of
43   * <code>Boolean.TRUE</code> for the key
44   * <code>HTMLEditorKit.ParserCallback.IMPLIED</code>.
45   * <p>HTML.Attributes defines a type safe enumeration of html attributes.
46   * If an attribute key of a tag is defined in HTML.Attribute, the
47   * HTML.Attribute will be used as the key, otherwise a String will be used.
48   * For example &lt;p foo=bar class=neat&gt; has two attributes. foo is
49   * not defined in HTML.Attribute, where as class is, therefore the
50   * AttributeSet will have two values in it, HTML.Attribute.CLASS with
51   * a String value of 'neat' and the String key 'foo' with a String value of
52   * 'bar'.
53   * <p>The position argument will indicate the start of the tag, comment
54   * or text. Similiar to arrays, the first character in the stream has a
55   * position of 0. For tags that are
56   * implied the position will indicate
57   * the location of the next encountered tag. In the first example,
58   * the implied start body and html tags will have the same position as the
59   * p tag, and the implied end p, html and body tags will all have the same
60   * position.
61   * <p>As html skips whitespace the position for text will be the position
62   * of the first valid character, eg in the string '\n\n\nblah'
63   * the text 'blah' will have a position of 3, the newlines are skipped.
64   * <p>
65   * For attributes that do not have a value, eg in the html
66   * string <code>&lt;foo blah&gt;</code> the attribute <code>blah</code>
67   * does not have a value, there are two possible values that will be
68   * placed in the AttributeSet's value:
69   * <ul>
70   * <li>If the DTD does not contain an definition for the element, or the
71   *     definition does not have an explicit value then the value in the
72   *     AttributeSet will be <code>HTML.NULL_ATTRIBUTE_VALUE</code>.
73   * <li>If the DTD contains an explicit value, as in:
74   *     <code>&lt;!ATTLIST OPTION selected (selected) #IMPLIED&gt;</code>
75   *     this value from the dtd (in this case selected) will be used.
76   * </ul>
77   * <p>
78   * Once the stream has been parsed, the callback is notified of the most
79   * likely end of line string. The end of line string will be one of
80   * \n, \r or \r\n, which ever is encountered the most in parsing the
81   * stream.
82   *
83   * @version     1.28 12/19/03
84   * @author      Sunita Mani
85   */
86  public class DocumentParser extends javax.swing.text.html.parser.Parser   {
87  
88      private int inbody;
89      private int intitle;
90      private int inhead;
91      private int instyle;
92      private int inscript;
93      private boolean seentitle;
94      private HTMLEditorKit.ParserCallback   callback = null;
95      private boolean ignoreCharSet = false;
96      private static final boolean debugFlag = false;
97  
98      public DocumentParser(DTD   dtd) {
99      super(dtd);
100     }
101  
102     public void parse(Reader in,  HTMLEditorKit.ParserCallback   callback, boolean ignoreCharSet) throws IOException {
103     this.ignoreCharSet = ignoreCharSet;
104     this.callback = callback;
105     parse(in);
106     // end of line
107     callback.handleEndOfLineString(getEndOfLineString());
108     }
109 
110     /**
111      * Handle Start Tag.
112      */
113     protected void handleStartTag(TagElement   tag) {
114 
115     Element   elem = tag.getElement();
116     if (elem == dtd.body) {
117         inbody++;
118     } else if (elem == dtd.html) {
119     } else if (elem == dtd.head) {
120         inhead++;
121     } else if (elem == dtd.title) {
122         intitle++;
123     } else if (elem == dtd.style) {
124         instyle++;
125     } else if (elem == dtd.script) {
126             inscript++;
127     }   
128     if (debugFlag) {
129         if (tag.fictional()) {
130         debug("Start Tag: " + tag.getHTMLTag() + " pos: " + getCurrentPos());
131         } else {
132         debug("Start Tag: " + tag.getHTMLTag() + " attributes: " + 
133               getAttributes() + " pos: " + getCurrentPos());
134         }
135     }
136     if (tag.fictional()) {
137         SimpleAttributeSet   attrs = new SimpleAttributeSet  ();
138         attrs.addAttribute(HTMLEditorKit.ParserCallback.IMPLIED,
139                    Boolean.TRUE);
140         callback.handleStartTag(tag.getHTMLTag(), attrs,
141                     getBlockStartPosition());
142     } else {
143         callback.handleStartTag(tag.getHTMLTag(), getAttributes(),
144                     getBlockStartPosition());
145         flushAttributes();
146     }
147     }
148 
149 
150     protected void handleComment(char text[]) {
151     if (debugFlag) {
152         debug("comment: ->" + new String  (text) + "<-"
153           + " pos: " + getCurrentPos());
154     }
155     callback.handleComment(text, getBlockStartPosition());
156     }
157 
158     /**
159      * Handle Empty Tag.
160      */
161     protected void handleEmptyTag(TagElement   tag) throws ChangedCharSetException   {
162 
163     Element   elem = tag.getElement();
164     if (elem == dtd.meta && !ignoreCharSet) {
165         SimpleAttributeSet   atts = getAttributes();
166         if (atts != null) {
167         String   content = (String  )atts.getAttribute(HTML.Attribute.CONTENT);
168         if (content != null) {
169             if ("content-type".equalsIgnoreCase((String  )atts.getAttribute(HTML.Attribute.HTTPEQUIV))) {
170             if (!content.equalsIgnoreCase("text/html") && 
171                 !content.equalsIgnoreCase("text/plain")) {
172                 throw new ChangedCharSetException  (content, false);
173             }
174             } else if ("charset" .equalsIgnoreCase((String  )atts.getAttribute(HTML.Attribute.HTTPEQUIV))) {
175             throw new ChangedCharSetException  (content, true);
176             }
177         }
178         }
179     }
180     if (inbody != 0 || elem == dtd.meta || elem == dtd.base || elem == dtd.isindex || elem == dtd.style || elem == dtd.link) {
181         if (debugFlag) {
182         if (tag.fictional()) {
183             debug("Empty Tag: " + tag.getHTMLTag() + " pos: " + getCurrentPos());
184         } else {
185             debug("Empty Tag: " + tag.getHTMLTag() + " attributes: " 
186               + getAttributes() + " pos: " + getCurrentPos());
187         }
188         }
189         if (tag.fictional()) {
190         SimpleAttributeSet   attrs = new SimpleAttributeSet  ();
191         attrs.addAttribute(HTMLEditorKit.ParserCallback.IMPLIED,
192                    Boolean.TRUE);
193         callback.handleSimpleTag(tag.getHTMLTag(), attrs,
194                      getBlockStartPosition());
195         } else {
196         callback.handleSimpleTag(tag.getHTMLTag(), getAttributes(),
197                      getBlockStartPosition());
198         flushAttributes();
199         }
200     }
201     }
202 
203     /**
204      * Handle End Tag.
205      */
206     protected void handleEndTag(TagElement   tag) {
207     Element   elem = tag.getElement();
208     if (elem == dtd.body) {
209         inbody--;
210     } else if (elem == dtd.title) {
211         intitle--;
212         seentitle = true;
213     } else if (elem == dtd.head) {
214             inhead--;
215     } else if (elem == dtd.style) {
216             instyle--;
217     } else if (elem == dtd.script) {
218             inscript--;
219     }
220     if (debugFlag) {
221         debug("End Tag: " + tag.getHTMLTag() + " pos: " + getCurrentPos());
222     }
223     callback.handleEndTag(tag.getHTMLTag(), getBlockStartPosition());
224 
225     }
226 
227     /**
228      * Handle Text.
229      */
230     protected void handleText(char data[]) {
231     if (data != null) {
232         if (inscript != 0) {
233         callback.handleComment(data, getBlockStartPosition());
234         return;
235         }
236         if (inbody != 0 || ((instyle != 0) ||
237                 ((intitle != 0) && !seentitle))) {
238         if (debugFlag) {
239             debug("text:  ->" + new String  (data) + "<-" + " pos: " + getCurrentPos());
240         }
241         callback.handleText(data, getBlockStartPosition());
242         }
243     }
244     }
245 
246     /*
247      * Error handling.
248      */
249     protected void handleError(int ln, String   errorMsg) {
250     if (debugFlag) {
251         debug("Error: ->" + errorMsg + "<-" + " pos: " + getCurrentPos());
252     }
253     /* PENDING: need to improve the error string. */
254     callback.handleError(errorMsg, getCurrentPos());
255     }
256 
257 
258     /*
259      * debug messages
260      */
261     private void debug(String   msg) {
262     System.out.println(msg);
263     }
264 }
265
A to Z: JavaDoc & Examples Daily Java News & Articles Open Source Projects Open Source Codes Free Computer Books Remove Frame
Popular Tags