1 /*
2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. 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.commons.codec.net;
19
20 import java.io.UnsupportedEncodingException;
21
22 import org.apache.commons.codec.DecoderException;
23 import org.apache.commons.codec.EncoderException;
24 import org.apache.commons.codec.binary.StringUtils;
25
26 /**
27 * <p>
28 * Implements methods common to all codecs defined in RFC 1522.
29 * </p>
30 *
31 * <p>
32 * <a href="http://www.ietf.org/rfc/rfc1522.txt">RFC 1522</a>
33 * describes techniques to allow the encoding of non-ASCII text in
34 * various portions of a RFC 822 [2] message header, in a manner which
35 * is unlikely to confuse existing message handling software.
36 * </p>
37
38 * @see <a href="http://www.ietf.org/rfc/rfc1522.txt">
39 * MIME (Multipurpose Internet Mail Extensions) Part Two:
40 * Message Header Extensions for Non-ASCII Text</a>
41 * </p>
42 *
43 * @author Apache Software Foundation
44 * @since 1.3
45 * @version $Id: RFC1522Codec.java 1170351 2011-09-13 21:09:09Z ggregory $
46 */
47 abstract class RFC1522Codec {
48
49 /**
50 * Separator.
51 */
52 protected static final char SEP = '?';
53
54 /**
55 * Prefix
56 */
57 protected static final String POSTFIX = "?=";
58
59 /**
60 * Postfix
61 */
62 protected static final String PREFIX = "=?";
63
64 /**
65 * Applies an RFC 1522 compliant encoding scheme to the given string of text with the
66 * given charset. This method constructs the "encoded-word" header common to all the
67 * RFC 1522 codecs and then invokes {@link #doEncoding(byte [])} method of a concrete
68 * class to perform the specific encoding.
69 *
70 * @param text a string to encode
71 * @param charset a charset to be used
72 *
73 * @return RFC 1522 compliant "encoded-word"
74 *
75 * @throws EncoderException thrown if there is an error condition during the Encoding
76 * process.
77 * @throws UnsupportedEncodingException thrown if charset is not supported
78 *
79 * @see <a href="http://download.oracle.com/javase/1.5.0/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
80 */
81 protected String encodeText(final String text, final String charset)
82 throws EncoderException, UnsupportedEncodingException
83 {
84 if (text == null) {
85 return null;
86 }
87 StringBuffer buffer = new StringBuffer();
88 buffer.append(PREFIX);
89 buffer.append(charset);
90 buffer.append(SEP);
91 buffer.append(getEncoding());
92 buffer.append(SEP);
93 byte [] rawdata = doEncoding(text.getBytes(charset));
94 buffer.append(StringUtils.newStringUsAscii(rawdata));
95 buffer.append(POSTFIX);
96 return buffer.toString();
97 }
98
99 /**
100 * Applies an RFC 1522 compliant decoding scheme to the given string of text. This method
101 * processes the "encoded-word" header common to all the RFC 1522 codecs and then invokes
102 * {@link #doEncoding(byte [])} method of a concrete class to perform the specific decoding.
103 *
104 * @param text a string to decode
105 * @return A new decoded String or <code>null</code> if the input is <code>null</code>.
106 *
107 * @throws DecoderException thrown if there is an error condition during the decoding
108 * process.
109 * @throws UnsupportedEncodingException thrown if charset specified in the "encoded-word"
110 * header is not supported
111 */
112 protected String decodeText(final String text)
113 throws DecoderException, UnsupportedEncodingException
114 {
115 if (text == null) {
116 return null;
117 }
118 if ((!text.startsWith(PREFIX)) || (!text.endsWith(POSTFIX))) {
119 throw new DecoderException("RFC 1522 violation: malformed encoded content");
120 }
121 int terminator = text.length() - 2;
122 int from = 2;
123 int to = text.indexOf(SEP, from);
124 if (to == terminator) {
125 throw new DecoderException("RFC 1522 violation: charset token not found");
126 }
127 String charset = text.substring(from, to);
128 if (charset.equals("")) {
129 throw new DecoderException("RFC 1522 violation: charset not specified");
130 }
131 from = to + 1;
132 to = text.indexOf(SEP, from);
133 if (to == terminator) {
134 throw new DecoderException("RFC 1522 violation: encoding token not found");
135 }
136 String encoding = text.substring(from, to);
137 if (!getEncoding().equalsIgnoreCase(encoding)) {
138 throw new DecoderException("This codec cannot decode " +
139 encoding + " encoded content");
140 }
141 from = to + 1;
142 to = text.indexOf(SEP, from);
143 byte[] data = StringUtils.getBytesUsAscii(text.substring(from, to));
144 data = doDecoding(data);
145 return new String(data, charset);
146 }
147
148 /**
149 * Returns the codec name (referred to as encoding in the RFC 1522)
150 *
151 * @return name of the codec
152 */
153 protected abstract String getEncoding();
154
155 /**
156 * Encodes an array of bytes using the defined encoding scheme
157 *
158 * @param bytes Data to be encoded
159 *
160 * @return A byte array containing the encoded data
161 *
162 * @throws EncoderException thrown if the Encoder encounters a failure condition
163 * during the encoding process.
164 */
165 protected abstract byte[] doEncoding(byte[] bytes) throws EncoderException;
166
167 /**
168 * Decodes an array of bytes using the defined encoding scheme
169 *
170 * @param bytes Data to be decoded
171 *
172 * @return a byte array that contains decoded data
173 *
174 * @throws DecoderException A decoder exception is thrown if a Decoder encounters a
175 * failure condition during the decode process.
176 */
177 protected abstract byte[] doDecoding(byte[] bytes) throws DecoderException;
178 }