KickJava   Java API By Example, From Geeks To Geeks.

Java > Open Source Codes > org > apache > batik > xml > XMLStreamNormalizingReader


1 /*
2
3    Copyright 2002-2003 The Apache Software Foundation
4
5    Licensed under the Apache License, Version 2.0 (the "License");
6    you may not use this file except in compliance with the License.
7    You may obtain a copy of the License at
8
9        http://www.apache.org/licenses/LICENSE-2.0
10
11    Unless required by applicable law or agreed to in writing, software
12    distributed under the License is distributed on an "AS IS" BASIS,
13    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14    See the License for the specific language governing permissions and
15    limitations under the License.
16
17  */
18 package org.apache.batik.xml;
19
20 import java.io.IOException JavaDoc;
21 import java.io.InputStream JavaDoc;
22 import java.io.PushbackInputStream JavaDoc;
23 import java.io.Reader JavaDoc;
24
25 import org.apache.batik.util.io.StreamNormalizingReader;
26 import org.apache.batik.util.io.UTF16Decoder;
27
28 /**
29  * This class represents a normalizing reader with encoding detection
30  * management.
31  *
32  * @author <a HREF="mailto:stephane@hillion.org">Stephane Hillion</a>
33  * @version $Id: XMLStreamNormalizingReader.java,v 1.5 2004/08/18 07:16:00 vhardy Exp $
34  */
35 public class XMLStreamNormalizingReader extends StreamNormalizingReader {
36     
37     /**
38      * Creates a new XMLStreamNormalizingReader.
39      * @param is The input stream to read.
40      * @param encod The character encoding to use if the auto-detection fail.
41      */
42     public XMLStreamNormalizingReader(InputStream JavaDoc is, String JavaDoc encod)
43         throws IOException JavaDoc {
44         PushbackInputStream JavaDoc pbis = new PushbackInputStream JavaDoc(is, 128);
45         byte[] buf = new byte[4];
46
47         int len = pbis.read(buf);
48         if (len > 0) {
49             pbis.unread(buf, 0, len);
50         }
51
52         if (len == 4) {
53             switch (buf[0] & 0x00FF) {
54             case 0:
55                 if (buf[1] == 0x003c && buf[2] == 0x0000 && buf[3] == 0x003f) {
56                     charDecoder = new UTF16Decoder(pbis, true);
57                     return;
58                 }
59                 break;
60
61             case '<':
62                 switch (buf[1] & 0x00FF) {
63                 case 0:
64                     if (buf[2] == 0x003f && buf[3] == 0x0000) {
65                         charDecoder = new UTF16Decoder(pbis, false);
66                         return;
67                     }
68                     break;
69
70                 case '?':
71                     if (buf[2] == 'x' && buf[3] == 'm') {
72                         Reader JavaDoc r = XMLUtilities.createXMLDeclarationReader
73                             (pbis, "UTF8");
74                         String JavaDoc enc = XMLUtilities.getXMLDeclarationEncoding
75                             (r, "UTF-8");
76                         charDecoder = createCharDecoder(pbis, enc);
77                         return;
78                     }
79                 }
80                 break;
81
82             case 0x004C:
83                 if (buf[1] == 0x006f &&
84                     (buf[2] & 0x00FF) == 0x00a7 &&
85                     (buf[3] & 0x00FF) == 0x0094) {
86                     Reader JavaDoc r = XMLUtilities.createXMLDeclarationReader
87                         (pbis, "CP037");
88                     String JavaDoc enc = XMLUtilities.getXMLDeclarationEncoding
89                         (r, "EBCDIC-CP-US");
90                     charDecoder = createCharDecoder(pbis, enc);
91                     return;
92                 }
93                 break;
94
95             case 0x00FE:
96                 if ((buf[1] & 0x00FF) == 0x00FF) {
97                     charDecoder = createCharDecoder(pbis, "UTF-16");
98                     return;
99                 }
100                 break;
101
102             case 0x00FF:
103                 if ((buf[1] & 0x00FF) == 0x00FE) {
104                     charDecoder = createCharDecoder(pbis, "UTF-16");
105                     return;
106                 }
107             }
108         }
109
110         encod = (encod == null) ? "UTF-8" : encod;
111         charDecoder = createCharDecoder(pbis, encod);
112     }
113 }
114
Popular Tags