LinkTag


1   // HTMLParser Library $Name: v1_5_20050313 $ - A java-based parser for HTML
2   // http://sourceforge.org/projects/htmlparser
3   // Copyright (C) 2004 Somik Raha
4   //
5   // Revision Control Information
6   //
7   // $Source: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/tags/LinkTag.java,v $
8   // $Author: derrickoswald $
9   // $Date: 2005/02/13 22:45:48 $
10  // $Revision: 1.53 $
11  //
12  // This library is free software; you can redistribute it and/or
13  // modify it under the terms of the GNU Lesser General Public
14  // License as published by the Free Software Foundation; either
15  // version 2.1 of the License, or (at your option) any later version.
16  //
17  // This library is distributed in the hope that it will be useful,
18  // but WITHOUT ANY WARRANTY; without even the implied warranty of
19  // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20  // Lesser General Public License for more details.
21  //
22  // You should have received a copy of the GNU Lesser General Public
23  // License along with this library; if not, write to the Free Software
24  // Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
25  //
26  
27  package org.htmlparser.tags;
28  
29  import org.htmlparser.Node;
30  import org.htmlparser.util.ParserUtils;
31  import org.htmlparser.util.SimpleNodeIterator;
32  
33  /**
34   * Identifies a link tag.
35   */
36  public class LinkTag extends CompositeTag
37  {
38      /**
39       * The set of names handled by this tag.
40       */
41      private static final String  [] mIds = new String  [] {"A"};
42  
43      /**
44       * The set of tag names that indicate the end of this tag.
45       */
46      private static final String  [] mEnders = new String  [] {"A", "P", "DIV", "TD", "TR", "FORM", "LI"};
47  
48      /**
49       * The set of end tag names that indicate the end of this tag.
50       */
51      private static final String  [] mEndTagEnders = new String  [] {"P", "DIV", "TD", "TR", "FORM", "LI", "BODY", "HTML"};
52  
53      /**
54       * The URL where the link points to
55       */
56      protected String   mLink;
57  
58      /**
59       * Set to true when the link was a mailto: URL.
60       */
61      private boolean mailLink;
62  
63      /**
64       * Set to true when the link was a javascript: URL.
65       */
66      private boolean javascriptLink;
67  
68      /**
69       * Constructor creates an LinkTag object, which basically stores the location
70       * where the link points to, and the text it contains.
71       * <p>
72       * In order to get the contents of the link tag, use the method linkData(),
73       * which returns an enumeration of nodes encapsulated within the link.
74       * <p>
75       * The following code will get all the images inside a link tag.
76       * <pre>
77       * Node node ;
78       * ImageTag imageTag;
79       * for (Enumeration e=linkTag.linkData();e.hasMoreElements();) {
80       *      node = (Node)e.nextElement();
81       *      if (node instanceof ImageTag) {
82       *          imageTag = (ImageTag)node;
83       *          // Process imageTag
84       *      }
85       * }
86       * </pre>
87       */
88      public LinkTag ()
89      {
90      }
91  
92      /**
93       * Return the set of names handled by this tag.
94       * @return The names to be matched that create tags of this type.
95       */
96      public String  [] getIds ()
97      {
98          return (mIds);
99      }
100 
101     /**
102      * Return the set of tag names that cause this tag to finish.
103      * @return The names of following tags that stop further scanning.
104      */
105     public String  [] getEnders ()
106     {
107         return (mEnders);
108     }
109 
110     /**
111      * Return the set of end tag names that cause this tag to finish.
112      * @return The names of following end tags that stop further scanning.
113      */
114     public String  [] getEndTagEnders ()
115     {
116         return (mEndTagEnders);
117     }
118 
119     /**
120      * Returns the accesskey attribute value, if any.
121      */
122     public String   getAccessKey()
123     {
124         return (getAttribute("ACCESSKEY"));
125     }
126 
127     /**
128      * Returns the url as a string, to which this link points.
129      * This string has had the "mailto:" and "javascript:" protocol stripped
130      * off the front (if those predicates return <code>true</code>) but not
131      * for other protocols. Don't ask me why, it's a legacy thing.
132      */
133     public String   getLink()
134     {
135         if (null == mLink)
136         {
137             mailLink=false;
138             javascriptLink = false;
139             mLink = extractLink ();
140 
141             int mailto = mLink.indexOf("mailto");
142             if (mailto==0)
143             {
144                 // yes it is
145                 mailto = mLink.indexOf(":");
146                 mLink = mLink.substring(mailto+1);
147                 mailLink = true;
148             }
149             int javascript = mLink.indexOf("javascript:");
150             if (javascript == 0)
151             {
152                 mLink = mLink.substring(11); // this magic number is "javascript:".length()
153                 javascriptLink = true;
154             }
155         }
156         return (mLink);
157     }
158 
159     /**
160      * Returns the text contained inside this link tag
161      */
162     public String   getLinkText()
163     {
164         String   ret;
165 
166         if (null != getChildren ())
167             ret = getChildren ().asString ();
168         else
169             ret = "";
170 
171         return (ret);
172     }
173 
174     /**
175      * Is this a mail address
176      * @return boolean true/false
177      */
178     public boolean isMailLink()
179     {
180         getLink (); // force an evaluation of the booleans
181         return (mailLink);
182     }
183 
184     /**
185      * Tests if the link is javascript
186      * @return flag indicating if the link is a javascript code
187      */
188     public boolean isJavascriptLink()
189     {
190         getLink (); // force an evaluation of the booleans
191         return (javascriptLink);
192     }
193 
194     /**
195      * Tests if the link is an FTP link.
196      *
197      * @return flag indicating if this link is an FTP link
198      */
199     public boolean isFTPLink() {
200         return getLink ().indexOf("ftp://")==0;
201     }
202 
203     /**
204      * Tests if the link is an IRC link.
205      * @return flag indicating if this link is an IRC link
206      */
207     public boolean isIRCLink() {
208         return getLink ().indexOf("irc://")==0;
209     }
210 
211     /**
212      * Tests if the link is an HTTP link.
213      *
214      * @return flag indicating if this link is an HTTP link
215      */
216     public boolean isHTTPLink()
217     {
218         return (!isFTPLink() && !isHTTPSLink() && !isJavascriptLink() && !isMailLink() && !isIRCLink());
219     }
220 
221     /**
222      * Tests if the link is an HTTPS link.
223      *
224      * @return flag indicating if this link is an HTTPS link
225      */
226     public boolean isHTTPSLink() {
227             return getLink ().indexOf("https://")==0;
228     }
229 
230         /**
231      * Tests if the link is an HTTP link or one of its variations (HTTPS, etc.).
232      *
233      * @return flag indicating if this link is an HTTP link or one of its variations (HTTPS, etc.)
234      */
235     public boolean isHTTPLikeLink() {
236             return isHTTPLink() || isHTTPSLink();
237     }
238 
239 
240     /**
241      * Insert the method's description here.
242      * Creation date: (8/3/2001 1:49:31 AM)
243      * @param newMailLink boolean
244      */
245     public void setMailLink(boolean newMailLink)
246     {
247         mailLink = newMailLink;
248     }
249 
250     /**
251      * Set the link as a javascript link.
252      *
253      * @param newJavascriptLink flag indicating if the link is a javascript code
254      */
255     public void setJavascriptLink(boolean newJavascriptLink)
256     {
257         javascriptLink = newJavascriptLink;
258     }
259 
260     /**
261      * Print the contents of this Link Node
262      */
263     public String   toString()
264     {
265         StringBuffer   sb = new StringBuffer  ();
266         sb.append("Link to : "+ getLink() + "; titled : "+getLinkText ()+"; begins at : "+getStartPosition ()+"; ends at : "+getEndPosition ()+ ", AccessKey=");
267         if (getAccessKey ()==null)
268             sb.append("null\n");
269         else
270             sb.append(getAccessKey ()+"\n");
271         if (null != getChildren ())
272         {
273             sb.append("  "+"LinkData\n");
274             sb.append("  "+"--------\n");
275 
276             Node node;
277             int i = 0;
278             for (SimpleNodeIterator e=children();e.hasMoreNodes();)
279             {
280                 node = e.nextNode();
281                 sb.append("   "+(i++)+ " ");
282                 sb.append(node.toString()+"\n");
283             }
284         }
285         sb.append("  "+"*** END of LinkData ***\n");
286         return sb.toString();
287     }
288 
289     public void setLink(String   link)
290     {
291         mLink = link;
292         setAttribute ("HREF", link);
293     }
294 
295     /**
296      * This method returns an enumeration of data that it contains
297      * @return Enumeration
298      * @deprecated Use children() instead.
299      */
300     public SimpleNodeIterator linkData() {
301         return children();
302     }
303 
304     /**
305      * Extract the link from the HREF attribute.
306      * @return The URL from the HREF attibute. This is absolute if the tag has
307      * a valid page.
308      */
309     public String   extractLink ()
310     {
311         String   ret;
312 
313         ret =  getAttribute ("HREF");
314         if (null != ret)
315         {
316             ret = ParserUtils.removeChars (ret,'\n');
317             ret = ParserUtils.removeChars (ret,'\r');
318         }
319         if (null != getPage ())
320             ret = getPage ().getAbsoluteURL (ret);
321 
322         return (ret);
323     }
324 }
325
A to Z: JavaDoc & Examples Daily Java News & Articles Open Source Projects Open Source Codes Free Computer Books Remove Frame
Popular Tags