1 /* Copyright 2002, 2003 Elliotte Rusty Harold
2 
3    This library is free software; you can redistribute it and/or modify
4    it under the terms of version 2.1 of the GNU Lesser General Public
5    License as published by the Free Software Foundation.
6 
7    This library is distributed in the hope that it will be useful,
8    but WITHOUT ANY WARRANTY; without even the implied warranty of
9    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
10    GNU Lesser General Public License for more details.
11 
12    You should have received a copy of the GNU Lesser General Public
13    License along with this library; if not, write to the
14    Free Software Foundation, Inc., 59 Temple Place, Suite 330,
15    Boston, MA 02111-1307  USA
16 
17    You can contact Elliotte Rusty Harold by sending e-mail to
18    elharo@ibiblio.org. Please include the word "XOM" in the
19    subject line. The XOM home page is located at http://www.xom.nu/
20 */
21 
22 package nu.xom.samples;
23 
24 import java.io.IOException;
25 import java.io.OutputStreamWriter;
26 import java.io.Writer;
27 
28 import nu.xom.Attribute;
29 import nu.xom.Builder;
30 import nu.xom.Document;
31 import nu.xom.Element;
32 import nu.xom.NodeFactory;
33 import nu.xom.Nodes;
34 import nu.xom.ParsingException;
35 
36 
37 /**
38  * <p>
39  *   Demonstrates filtered streaming via a <code>NodeFactory</code>
40  *   subclass.
41  * </p>
42  *
43  * @author Elliotte Rusty Harold
44  * @version 1.0
45  *
46  */
47 public class StreamingTextExtractor extends NodeFactory {
48 
49     private Writer out;
50     private Nodes empty = new Nodes();
51 
StreamingTextExtractor(Writer out)52     public StreamingTextExtractor(Writer out) {
53       if (out == null) {
54       throw new NullPointerException("Writer must be non-null.");
55       }
56       this.out = out;
57     }
58 
StreamingTextExtractor()59     public StreamingTextExtractor() {
60       this(new OutputStreamWriter(System.out));
61     }
62 
makeComment(String data)63     public Nodes makeComment(String data) {
64         return empty;
65     }
66 
makeText(String data)67     public Nodes makeText(String data) {
68         try {
69             out.write(data);
70         }
71         catch (IOException ex) {
72             System.err.println(ex);
73         }
74         return empty;
75     }
76 
makeRootElement(String name, String namespace)77     public Element makeRootElement(String name, String namespace) {
78         Element result = new Element(name, namespace);
79         return result;
80     }
81 
startMakingElement(String name, String namespace)82     public Element startMakingElement(String name, String namespace) {
83         return null;
84     }
85 
makeAttribute(String name, String namespace, String value, Attribute.Type type)86     public Nodes makeAttribute(String name, String namespace,
87       String value, Attribute.Type type) {
88         return empty;
89     }
90 
makeDocType(String rootElementName, String publicID, String systemID)91     public Nodes makeDocType(String rootElementName,
92       String publicID, String systemID) {
93         return empty;
94     }
95 
makeProcessingInstruction( String target, String data)96     public Nodes makeProcessingInstruction(
97       String target, String data) {
98         return empty;
99     }
100 
finishMakingDocument(Document doc)101     public void finishMakingDocument(Document doc) {
102         try {
103             out.flush();
104         }
105         catch (IOException ex) {
106            System.err.println(ex);
107         }
108     }
109 
main(String[] args)110     public static void main(String[] args) {
111 
112         if (args.length <= 0) {
113           System.out.println(
114             "Usage: java nu.xom.samples.StreamingTextExtractor URL"
115           );
116           return;
117         }
118 
119         try {
120           Builder parser = new Builder(new StreamingTextExtractor());
121           parser.build(args[0]);
122         }
123         catch (ParsingException ex) {
124           System.out.println(args[0] + " is not well-formed.");
125           System.out.println(ex.getMessage());
126         }
127         catch (IOException ex) {
128           System.out.println(
129            "Due to an IOException, the parser could not read "
130            + args[0]
131           );
132         }
133 
134     }
135 
136 }
137