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 */
017package org.apache.commons.io;
018
019import java.nio.charset.Charset;
020import java.nio.charset.StandardCharsets;
021import java.nio.charset.UnsupportedCharsetException;
022import java.util.Collections;
023import java.util.SortedMap;
024import java.util.TreeMap;
025
026/**
027 * Charsets required of every implementation of the Java platform.
028 *
029 * From the Java documentation <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">
030 * Standard charsets</a>:
031 * <p>
032 * <cite>Every implementation of the Java platform is required to support the following character encodings. Consult
033 * the release documentation for your implementation to see if any other encodings are supported. Consult the release
034 * documentation for your implementation to see if any other encodings are supported. </cite>
035 * </p>
036 *
037 * <ul>
038 * <li>{@code US-ASCII}<br>
039 * Seven-bit ASCII, a.k.a. ISO646-US, a.k.a. the Basic Latin block of the Unicode character set.</li>
040 * <li>{@code ISO-8859-1}<br>
041 * ISO Latin Alphabet No. 1, a.k.a. ISO-LATIN-1.</li>
042 * <li>{@code UTF-8}<br>
043 * Eight-bit Unicode Transformation Format.</li>
044 * <li>{@code UTF-16BE}<br>
045 * Sixteen-bit Unicode Transformation Format, big-endian byte order.</li>
046 * <li>{@code UTF-16LE}<br>
047 * Sixteen-bit Unicode Transformation Format, little-endian byte order.</li>
048 * <li>{@code UTF-16}<br>
049 * Sixteen-bit Unicode Transformation Format, byte order specified by a mandatory initial byte-order mark (either order
050 * accepted on input, big-endian used on output.)</li>
051 * </ul>
052 *
053 * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
054 * @since 2.3
055 */
056public class Charsets {
057
058    //
059    // This class should only contain Charset instances for required encodings. This guarantees that it will load
060    // correctly and without delay on all Java platforms.
061    //
062
063    private static final SortedMap<String, Charset> STANDARD_CHARSET_MAP;
064
065    static {
066        final SortedMap<String, Charset> standardCharsetMap = new TreeMap<>(String.CASE_INSENSITIVE_ORDER);
067        standardCharsetMap.put(StandardCharsets.ISO_8859_1.name(), StandardCharsets.ISO_8859_1);
068        standardCharsetMap.put(StandardCharsets.US_ASCII.name(), StandardCharsets.US_ASCII);
069        standardCharsetMap.put(StandardCharsets.UTF_16.name(), StandardCharsets.UTF_16);
070        standardCharsetMap.put(StandardCharsets.UTF_16BE.name(), StandardCharsets.UTF_16BE);
071        standardCharsetMap.put(StandardCharsets.UTF_16LE.name(), StandardCharsets.UTF_16LE);
072        standardCharsetMap.put(StandardCharsets.UTF_8.name(), StandardCharsets.UTF_8);
073        STANDARD_CHARSET_MAP = Collections.unmodifiableSortedMap(standardCharsetMap);
074    }
075
076    /**
077     * CharEncodingISO Latin Alphabet No. 1, a.k.a. ISO-LATIN-1.
078     * <p>
079     * Every implementation of the Java platform is required to support this character encoding.
080     * </p>
081     *
082     * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
083     * @deprecated Use Java 7's {@link java.nio.charset.StandardCharsets}
084     */
085    @Deprecated
086    public static final Charset ISO_8859_1 = StandardCharsets.ISO_8859_1;
087
088    /**
089     * <p>
090     * Seven-bit ASCII, also known as ISO646-US, also known as the Basic Latin block of the Unicode character set.
091     * </p>
092     * <p>
093     * Every implementation of the Java platform is required to support this character encoding.
094     * </p>
095     *
096     * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
097     * @deprecated Use Java 7's {@link java.nio.charset.StandardCharsets}
098     */
099    @Deprecated
100    public static final Charset US_ASCII = StandardCharsets.US_ASCII;
101
102    /**
103     * <p>
104     * Sixteen-bit Unicode Transformation Format, The byte order specified by a mandatory initial byte-order mark
105     * (either order accepted on input, big-endian used on output)
106     * </p>
107     * <p>
108     * Every implementation of the Java platform is required to support this character encoding.
109     * </p>
110     *
111     * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
112     * @deprecated Use Java 7's {@link java.nio.charset.StandardCharsets}
113     */
114    @Deprecated
115    public static final Charset UTF_16 = StandardCharsets.UTF_16;
116
117    /**
118     * <p>
119     * Sixteen-bit Unicode Transformation Format, big-endian byte order.
120     * </p>
121     * <p>
122     * Every implementation of the Java platform is required to support this character encoding.
123     * </p>
124     *
125     * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
126     * @deprecated Use Java 7's {@link java.nio.charset.StandardCharsets}
127     */
128    @Deprecated
129    public static final Charset UTF_16BE = StandardCharsets.UTF_16BE;
130
131    /**
132     * <p>
133     * Sixteen-bit Unicode Transformation Format, little-endian byte order.
134     * </p>
135     * <p>
136     * Every implementation of the Java platform is required to support this character encoding.
137     * </p>
138     *
139     * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
140     * @deprecated Use Java 7's {@link java.nio.charset.StandardCharsets}
141     */
142    @Deprecated
143    public static final Charset UTF_16LE = StandardCharsets.UTF_16LE;
144
145    /**
146     * <p>
147     * Eight-bit Unicode Transformation Format.
148     * </p>
149     * <p>
150     * Every implementation of the Java platform is required to support this character encoding.
151     * </p>
152     *
153     * @see <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">Standard charsets</a>
154     * @deprecated Use Java 7's {@link java.nio.charset.StandardCharsets}
155     */
156    @Deprecated
157    public static final Charset UTF_8 = StandardCharsets.UTF_8;
158
159    /**
160     * Tests whether the given non-null Charset has an alias of the given name.
161     *
162     * @param charset a non-null Charset.
163     * @param charsetName The name to test.
164     * @return whether the given non-null charset name is a UTF-8 alias.
165     * @since 2.20.0
166     */
167    public static boolean isAlias(final Charset charset, final String charsetName) {
168        return charsetName != null && (charset.name().equalsIgnoreCase(charsetName) || charset.aliases().stream().anyMatch(charsetName::equalsIgnoreCase));
169    }
170
171    /**
172     * Tests whether a given encoding is UTF-8. If the given charset is null, then check the platform's default encoding.
173     *
174     * @param charset If the given charset is null, then check the platform's default encoding.
175     * @return whether a given encoding is UTF-8.
176     * @since 2.20.0
177     */
178    public static boolean isUTF8(final Charset charset) {
179        return isUTF8Alias(toCharset(charset).name());
180    }
181
182    /**
183     * Tests whether the given non-null charset name is a UTF-8 alias.
184     *
185     * @param charsetName a non-null charset name.
186     * @return whether the given non-null charset name is a UTF-8 alias.
187     */
188    private static boolean isUTF8Alias(final String charsetName) {
189        return isAlias(StandardCharsets.UTF_8, charsetName);
190    }
191
192    /**
193     * Constructs a sorted map from canonical charset names to charset objects required of every implementation of the
194     * Java platform.
195     * <p>
196     * From the Java documentation <a href="https://docs.oracle.com/javase/8/docs/api/java/nio/charset/Charset.html">
197     * Standard charsets</a>:
198     * </p>
199     *
200     * @return An immutable, case-insensitive map from canonical charset names to charset objects.
201     * @see Charset#availableCharsets()
202     * @since 2.5
203     */
204    public static SortedMap<String, Charset> requiredCharsets() {
205        return STANDARD_CHARSET_MAP;
206    }
207
208    /**
209     * Returns the given Charset or the {@link Charset#defaultCharset() default Charset} if the given Charset is null.
210     *
211     * @param charset
212     *            A charset or null.
213     * @return the given Charset or the default Charset if the given Charset is null
214     * @see Charset#defaultCharset()
215     */
216    public static Charset toCharset(final Charset charset) {
217        return charset == null ? Charset.defaultCharset() : charset;
218    }
219
220    /**
221     * Returns the given charset if non-null, otherwise return defaultCharset.
222     *
223     * @param charset The charset to test, may be null.
224     * @param defaultCharset The charset to return if charset is null, may be null.
225     * @return a Charset.
226     * @since 2.12.0
227     */
228    public static Charset toCharset(final Charset charset, final Charset defaultCharset) {
229        return charset == null ? defaultCharset : charset;
230    }
231
232    /**
233     * Returns a Charset for the named charset. If the name is null, return the {@link Charset#defaultCharset() default Charset}.
234     *
235     * @param charsetName The name of the requested charset, may be null.
236     * @return a Charset for the named charset.
237     * @throws UnsupportedCharsetException If the named charset is unavailable (unchecked exception).
238     * @see Charset#defaultCharset()
239     */
240    public static Charset toCharset(final String charsetName) throws UnsupportedCharsetException {
241        return toCharset(charsetName, Charset.defaultCharset());
242    }
243
244    /**
245     * Returns a Charset for the named charset. If the name is null, return the given default Charset.
246     *
247     * @param charsetName The name of the requested charset, may be null.
248     * @param defaultCharset The charset to return if charsetName is null, may be null.
249     * @return a Charset for the named charset.
250     * @throws UnsupportedCharsetException If the named charset is unavailable (unchecked exception).
251     * @since 2.12.0
252     */
253    public static Charset toCharset(final String charsetName, final Charset defaultCharset) throws UnsupportedCharsetException {
254        return charsetName == null ? defaultCharset : Charset.forName(charsetName);
255    }
256
257    /**
258     * Returns a Charset for the named charset or the {@code defaultCharset}.
259     * <p>
260     * If {@code charsetName} cannot load a charset, return {@code defaultCharset}. Therefore, this method should never fail and always return a Charset.
261     * </p>
262     *
263     * @param charsetName    The name of the requested charset, may be null.
264     * @param defaultCharset The charset to return if charsetName is null or there is a problem, may be null which returns {@link Charset#defaultCharset()}.
265     * @return a Charset for the named charset or {@code defaultCharset} if any errors occur.
266     * @see Charset#defaultCharset()
267     * @since 2.20.0
268     */
269    public static Charset toCharsetDefault(final String charsetName, final Charset defaultCharset) {
270        try {
271            return toCharset(charsetName);
272        } catch (final RuntimeException ignored) {
273            return toCharset(defaultCharset);
274        }
275    }
276
277    /**
278     * Construct a new instance.
279     *
280     * @deprecated Will be private in 4.0
281     */
282    @Deprecated
283    public Charsets() {
284        // empty
285    }
286}