Source code

001/*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  You may obtain a copy of the License at
008 *
009 *      https://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017
018package org.apache.commons.codec.net;
019
020import java.io.UnsupportedEncodingException;
021import java.nio.charset.Charset;
022import java.nio.charset.StandardCharsets;
023import java.nio.charset.UnsupportedCharsetException;
024import java.util.BitSet;
025
026import org.apache.commons.codec.DecoderException;
027import org.apache.commons.codec.EncoderException;
028import org.apache.commons.codec.StringDecoder;
029import org.apache.commons.codec.StringEncoder;
030
031/**
032 * Similar to the Quoted-Printable content-transfer-encoding defined in
033 * <a href="https://www.ietf.org/rfc/rfc1521.txt">RFC 1521</a> and designed to allow text containing mostly ASCII
034 * characters to be decipherable on an ASCII terminal without decoding.
035 * <p>
036 * <a href="https://www.ietf.org/rfc/rfc1522.txt">RFC 1522</a> describes techniques to allow the encoding of non-ASCII
037 * text in various portions of a RFC 822 [2] message header, in a manner which is unlikely to confuse existing message
038 * handling software.
039 * </p>
040 * <p>
041 * This class is conditionally thread-safe.
042 * The instance field for encoding blanks is mutable {@link #setEncodeBlanks(boolean)}
043 * but is not volatile, and accesses are not synchronized.
044 * If an instance of the class is shared between threads, the caller needs to ensure that suitable synchronization
045 * is used to ensure safe publication of the value between threads, and must not invoke
046 * {@link #setEncodeBlanks(boolean)} after initial setup.
047 * </p>
048 *
049 * @see <a href="https://www.ietf.org/rfc/rfc1522.txt">MIME (Multipurpose Internet Mail Extensions) Part Two: Message
050 *          Header Extensions for Non-ASCII Text</a>
051 *
052 * @since 1.3
053 */
054public class QCodec extends RFC1522Codec implements StringEncoder, StringDecoder {
055
056    /**
057     * BitSet of printable characters as defined in RFC 1522.
058     */
059    private static final BitSet PRINTABLE_CHARS = new BitSet(256);
060
061    // Static initializer for printable chars collection
062    static {
063        // alpha characters
064        PRINTABLE_CHARS.set(' ');
065        PRINTABLE_CHARS.set('!');
066        PRINTABLE_CHARS.set('"');
067        PRINTABLE_CHARS.set('#');
068        PRINTABLE_CHARS.set('$');
069        PRINTABLE_CHARS.set('%');
070        PRINTABLE_CHARS.set('&');
071        PRINTABLE_CHARS.set('\'');
072        PRINTABLE_CHARS.set('(');
073        PRINTABLE_CHARS.set(')');
074        PRINTABLE_CHARS.set('*');
075        PRINTABLE_CHARS.set('+');
076        PRINTABLE_CHARS.set(',');
077        PRINTABLE_CHARS.set('-');
078        PRINTABLE_CHARS.set('.');
079        PRINTABLE_CHARS.set('/');
080        for (int i = '0'; i <= '9'; i++) {
081            PRINTABLE_CHARS.set(i);
082        }
083        PRINTABLE_CHARS.set(':');
084        PRINTABLE_CHARS.set(';');
085        PRINTABLE_CHARS.set('<');
086        PRINTABLE_CHARS.set('>');
087        PRINTABLE_CHARS.set('@');
088        for (int i = 'A'; i <= 'Z'; i++) {
089            PRINTABLE_CHARS.set(i);
090        }
091        PRINTABLE_CHARS.set('[');
092        PRINTABLE_CHARS.set('\\');
093        PRINTABLE_CHARS.set(']');
094        PRINTABLE_CHARS.set('^');
095        PRINTABLE_CHARS.set('`');
096        for (int i = 'a'; i <= 'z'; i++) {
097            PRINTABLE_CHARS.set(i);
098        }
099        PRINTABLE_CHARS.set('{');
100        PRINTABLE_CHARS.set('|');
101        PRINTABLE_CHARS.set('}');
102        PRINTABLE_CHARS.set('~');
103    }
104    private static final byte UNDERSCORE = 95;
105
106    private boolean encodeBlanks;
107
108    /**
109     * Constructs a new instance.
110     */
111    public QCodec() {
112        this(StandardCharsets.UTF_8);
113    }
114
115    /**
116     * Constructs a new instance for the selection of a default Charset.
117     *
118     * @param charset
119     *            the default string Charset to use.
120     *
121     * @see Charset
122     * @since 1.7
123     */
124    public QCodec(final Charset charset) {
125        super(charset);
126    }
127
128    /**
129     * Constructs a new instance for the selection of a default Charset.
130     *
131     * @param charsetName
132     *            the Charset to use.
133     * @throws java.nio.charset.UnsupportedCharsetException
134     *             If the named Charset is unavailable.
135     * @since 1.7 throws UnsupportedCharsetException if the named Charset is unavailable
136     * @see Charset
137     */
138    public QCodec(final String charsetName) {
139        this(Charset.forName(charsetName));
140    }
141
142    /**
143     * Decodes a quoted-printable object into its original form. Escaped characters are converted back to their original
144     * representation.
145     *
146     * @param obj
147     *            quoted-printable object to convert into its original form.
148     * @return original object.
149     * @throws DecoderException
150     *             Thrown if the argument is not a {@code String}. Thrown if a failure condition is encountered
151     *             during the decode process.
152     */
153    @Override
154    public Object decode(final Object obj) throws DecoderException {
155        if (obj == null) {
156            return null;
157        }
158        if (obj instanceof String) {
159            return decode((String) obj);
160        }
161        throw new DecoderException("Objects of type " + obj.getClass().getName() + " cannot be decoded using Q codec");
162    }
163
164    /**
165     * Decodes a quoted-printable string into its original form. Escaped characters are converted back to their original
166     * representation.
167     *
168     * @param str
169     *            quoted-printable string to convert into its original form.
170     * @return original string.
171     * @throws DecoderException
172     *             A decoder exception is thrown if a failure condition is encountered during the decode process.
173     */
174    @Override
175    public String decode(final String str) throws DecoderException {
176        try {
177            return decodeText(str);
178        } catch (final UnsupportedEncodingException e) {
179            throw new DecoderException(e.getMessage(), e);
180        }
181    }
182
183    @Override
184    protected byte[] doDecoding(final byte[] bytes) throws DecoderException {
185        if (bytes == null) {
186            return null;
187        }
188        boolean hasUnderscores = false;
189        for (final byte b : bytes) {
190            if (b == UNDERSCORE) {
191                hasUnderscores = true;
192                break;
193            }
194        }
195        if (hasUnderscores) {
196            final byte[] tmp = new byte[bytes.length];
197            for (int i = 0; i < bytes.length; i++) {
198                final byte b = bytes[i];
199                if (b != UNDERSCORE) {
200                    tmp[i] = b;
201                } else {
202                    tmp[i] = Utils.SPACE;
203                }
204            }
205            return QuotedPrintableCodec.decodeQuotedPrintable(tmp);
206        }
207        return QuotedPrintableCodec.decodeQuotedPrintable(bytes);
208    }
209
210    @Override
211    protected byte[] doEncoding(final byte[] bytes) {
212        if (bytes == null) {
213            return null;
214        }
215        final byte[] data = QuotedPrintableCodec.encodeQuotedPrintable(PRINTABLE_CHARS, bytes);
216        if (this.encodeBlanks) {
217            for (int i = 0; i < data.length; i++) {
218                if (data[i] == Utils.SPACE) {
219                    data[i] = UNDERSCORE;
220                }
221            }
222        }
223        return data;
224    }
225
226    /**
227     * Encodes an object into its quoted-printable form using the default Charset. Unsafe characters are escaped.
228     *
229     * @param obj
230     *            object to convert to quoted-printable form.
231     * @return quoted-printable object.
232     * @throws EncoderException
233     *             thrown if a failure condition is encountered during the encoding process.
234     */
235    @Override
236    public Object encode(final Object obj) throws EncoderException {
237        if (obj == null) {
238            return null;
239        }
240        if (obj instanceof String) {
241            return encode((String) obj);
242        }
243        throw new EncoderException("Objects of type " + obj.getClass().getName() + " cannot be encoded using Q codec");
244    }
245
246    /**
247     * Encodes a string into its quoted-printable form using the default Charset. Unsafe characters are escaped.
248     *
249     * @param sourceStr
250     *            string to convert to quoted-printable form.
251     * @return quoted-printable string.
252     * @throws EncoderException
253     *             thrown if a failure condition is encountered during the encoding process.
254     */
255    @Override
256    public String encode(final String sourceStr) throws EncoderException {
257        return encode(sourceStr, getCharset());
258    }
259
260    /**
261     * Encodes a string into its quoted-printable form using the specified Charset. Unsafe characters are escaped.
262     *
263     * @param sourceStr
264     *            string to convert to quoted-printable form.
265     * @param sourceCharset
266     *            the Charset for sourceStr.
267     * @return quoted-printable string.
268     * @throws EncoderException
269     *             thrown if a failure condition is encountered during the encoding process.
270     * @since 1.7
271     */
272    public String encode(final String sourceStr, final Charset sourceCharset) throws EncoderException {
273        return encodeText(sourceStr, sourceCharset);
274    }
275
276    /**
277     * Encodes a string into its quoted-printable form using the specified Charset. Unsafe characters are escaped.
278     *
279     * @param sourceStr
280     *            string to convert to quoted-printable form.
281     * @param sourceCharset
282     *            the Charset for sourceStr.
283     * @return quoted-printable string.
284     * @throws EncoderException
285     *             thrown if a failure condition is encountered during the encoding process.
286     */
287    public String encode(final String sourceStr, final String sourceCharset) throws EncoderException {
288        try {
289            return encodeText(sourceStr, sourceCharset);
290        } catch (final UnsupportedCharsetException e) {
291            throw new EncoderException(e.getMessage(), e);
292        }
293    }
294
295    @Override
296    protected String getEncoding() {
297        return "Q";
298    }
299
300    /**
301     * Tests if optional transformation of SPACE characters is to be used
302     *
303     * @return {@code true} if SPACE characters are to be transformed, {@code false} otherwise.
304     */
305    public boolean isEncodeBlanks() {
306        return this.encodeBlanks;
307    }
308
309    /**
310     * Defines whether optional transformation of SPACE characters is to be used
311     *
312     * @param b
313     *            {@code true} if SPACE characters are to be transformed, {@code false} otherwise.
314     */
315    public void setEncodeBlanks(final boolean b) {
316        this.encodeBlanks = b;
317    }
318}