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 package org.apache.commons.text.translate;
18
19 import java.io.IOException;
20 import java.io.Writer;
21 import java.util.Arrays;
22 import java.util.Collections;
23 import java.util.EnumSet;
24
25 import org.apache.commons.lang3.ArrayUtils;
26
27 /**
28 * Translates XML numeric entities of the form &#[xX]?\d+;? to
29 * the specific code point.
30 *
31 * Note that the semicolon is optional.
32 *
33 * @since 1.0
34 */
35 public class NumericEntityUnescaper extends CharSequenceTranslator {
36
37 /** Enumerates NumericEntityUnescaper options for unescaping. */
38 public enum OPTION {
39
40 /**
41 * Requires a semicolon.
42 */
43 semiColonRequired,
44
45 /**
46 * Does not require a semicolon.
47 */
48 semiColonOptional,
49
50 /**
51 * Throws an exception if a semicolon is missing.
52 */
53 errorIfNoSemiColon
54 }
55
56 /** Default options. */
57 private static final EnumSet<OPTION> DEFAULT_OPTIONS = EnumSet
58 .copyOf(Collections.singletonList(OPTION.semiColonRequired));
59
60 /** EnumSet of OPTIONS, given from the constructor, read-only. */
61 private final EnumSet<OPTION> options;
62
63 /**
64 * Creates a UnicodeUnescaper.
65 *
66 * The constructor takes a list of options, only one type of which is currently
67 * available (whether to allow, error or ignore the semicolon on the end of a
68 * numeric entity to being missing).
69 *
70 * For example, to support numeric entities without a ';':
71 * new NumericEntityUnescaper(NumericEntityUnescaper.OPTION.semiColonOptional)
72 * and to throw an IllegalArgumentException when they're missing:
73 * new NumericEntityUnescaper(NumericEntityUnescaper.OPTION.errorIfNoSemiColon)
74 *
75 * Note that the default behavior is to ignore them.
76 *
77 * @param options to apply to this unescaper
78 */
79 public NumericEntityUnescaper(final OPTION... options) {
80 this.options = ArrayUtils.isEmpty(options) ? DEFAULT_OPTIONS : EnumSet.copyOf(Arrays.asList(options));
81 }
82
83 /**
84 * Tests whether the passed in option is currently set.
85 *
86 * @param option to check state of
87 * @return whether the option is set
88 */
89 public boolean isSet(final OPTION option) {
90 return options.contains(option);
91 }
92
93 /**
94 * {@inheritDoc}
95 */
96 @Override
97 public int translate(final CharSequence input, final int index, final Writer writer) throws IOException {
98 final int seqEnd = input.length();
99 // Uses -2 to ensure there is something after the &#
100 if (input.charAt(index) == '&' && index < seqEnd - 2 && input.charAt(index + 1) == '#') {
101 int start = index + 2;
102 boolean isHex = false;
103
104 final char firstChar = input.charAt(start);
105 if (firstChar == 'x' || firstChar == 'X') {
106 start++;
107 isHex = true;
108
109 // Check there's more than just an x after the &#
110 if (start == seqEnd) {
111 return 0;
112 }
113 }
114
115 int end = start;
116 // Note that this supports character codes without a ; on the end
117 while (end < seqEnd && (input.charAt(end) >= '0' && input.charAt(end) <= '9'
118 || input.charAt(end) >= 'a' && input.charAt(end) <= 'f'
119 || input.charAt(end) >= 'A' && input.charAt(end) <= 'F')) {
120 end++;
121 }
122
123 final boolean semiNext = end != seqEnd && input.charAt(end) == ';';
124
125 if (!semiNext) {
126 if (isSet(OPTION.semiColonRequired)) {
127 return 0;
128 }
129 if (isSet(OPTION.errorIfNoSemiColon)) {
130 throw new IllegalArgumentException("Semi-colon required at end of numeric entity");
131 }
132 }
133
134 final int entityValue;
135 try {
136 if (isHex) {
137 entityValue = Integer.parseInt(input.subSequence(start, end).toString(), 16);
138 } else {
139 entityValue = Integer.parseInt(input.subSequence(start, end).toString(), 10);
140 }
141 } catch (final NumberFormatException nfe) {
142 return 0;
143 }
144
145 if (entityValue > 0xFFFF) {
146 final char[] chrs = Character.toChars(entityValue);
147 writer.write(chrs[0]);
148 writer.write(chrs[1]);
149 } else {
150 writer.write(entityValue);
151 }
152
153 return 2 + end - start + (isHex ? 1 : 0) + (semiNext ? 1 : 0);
154 }
155 return 0;
156 }
157 }