001/*
002 * Copyright (C) 2009 The Guava Authors
003 *
004 * Licensed under the Apache License, Version 2.0 (the "License");
005 * you may not use this file except in compliance with the License.
006 * You may obtain a copy of the License at
007 *
008 * http://www.apache.org/licenses/LICENSE-2.0
009 *
010 * Unless required by applicable law or agreed to in writing, software
011 * distributed under the License is distributed on an "AS IS" BASIS,
012 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
013 * See the License for the specific language governing permissions and
014 * limitations under the License.
015 */
016
017package com.google.common.escape;
018
019import static com.google.common.base.Preconditions.checkNotNull;
020
021import com.google.common.annotations.Beta;
022import com.google.common.annotations.GwtCompatible;
023
024import java.util.Map;
025
026/**
027 * A {@link CharEscaper} that uses an array to quickly look up replacement
028 * characters for a given {@code char} value. An additional safe range is
029 * provided that determines whether {@code char} values without specific
030 * replacements are to be considered safe and left unescaped or should be
031 * escaped in a general way.
032 *
033 * <p>A good example of usage of this class is for Java source code escaping
034 * where the replacement array contains information about special ASCII
035 * characters such as {@code \\t} and {@code \\n} while {@link #escapeUnsafe}
036 * is overridden to handle general escaping of the form {@code \\uxxxx}.
037 *
038 * <p>The size of the data structure used by {@link ArrayBasedCharEscaper} is
039 * proportional to the highest valued character that requires escaping.
040 * For example a replacement map containing the single character
041 * '{@code \}{@code u1000}' will require approximately 16K of memory. If you
042 * need to create multiple escaper instances that have the same character
043 * replacement mapping consider using {@link ArrayBasedEscaperMap}.
044 *
045 * @author Sven Mawson
046 * @author David Beaumont
047 * @since 15.0
048 */
049@Beta
050@GwtCompatible
051public abstract class ArrayBasedCharEscaper extends CharEscaper {
052  // The replacement array (see ArrayBasedEscaperMap).
053  private final char[][] replacements;
054  // The number of elements in the replacement array.
055  private final int replacementsLength;
056  // The first character in the safe range.
057  private final char safeMin;
058  // The last character in the safe range.
059  private final char safeMax;
060
061  /**
062   * Creates a new ArrayBasedCharEscaper instance with the given replacement map
063   * and specified safe range. If {@code safeMax < safeMin} then no characters
064   * are considered safe.
065   *
066   * <p>If a character has no mapped replacement then it is checked against the
067   * safe range. If it lies outside that, then {@link #escapeUnsafe} is
068   * called, otherwise no escaping is performed.
069   *
070   * @param replacementMap a map of characters to their escaped representations
071   * @param safeMin the lowest character value in the safe range
072   * @param safeMax the highest character value in the safe range
073   */
074  protected ArrayBasedCharEscaper(Map<Character, String> replacementMap,
075      char safeMin, char safeMax) {
076
077    this(ArrayBasedEscaperMap.create(replacementMap), safeMin, safeMax);
078  }
079
080  /**
081   * Creates a new ArrayBasedCharEscaper instance with the given replacement map
082   * and specified safe range. If {@code safeMax < safeMin} then no characters
083   * are considered safe. This initializer is useful when explicit instances of
084   * ArrayBasedEscaperMap are used to allow the sharing of large replacement
085   * mappings.
086   *
087   * <p>If a character has no mapped replacement then it is checked against the
088   * safe range. If it lies outside that, then {@link #escapeUnsafe} is
089   * called, otherwise no escaping is performed.
090   *
091   * @param escaperMap the mapping of characters to be escaped
092   * @param safeMin the lowest character value in the safe range
093   * @param safeMax the highest character value in the safe range
094   */
095  protected ArrayBasedCharEscaper(ArrayBasedEscaperMap escaperMap,
096      char safeMin, char safeMax) {
097
098    checkNotNull(escaperMap);  // GWT specific check (do not optimize)
099    this.replacements = escaperMap.getReplacementArray();
100    this.replacementsLength = replacements.length;
101    if (safeMax < safeMin) {
102      // If the safe range is empty, set the range limits to opposite extremes
103      // to ensure the first test of either value will (almost certainly) fail.
104      safeMax = Character.MIN_VALUE;
105      safeMin = Character.MAX_VALUE;
106    }
107    this.safeMin = safeMin;
108    this.safeMax = safeMax;
109  }
110
111  /*
112   * This is overridden to improve performance. Rough benchmarking shows that
113   * this almost doubles the speed when processing strings that do not require
114   * any escaping.
115   */
116  @Override
117  public final String escape(String s) {
118    checkNotNull(s);  // GWT specific check (do not optimize).
119    for (int i = 0; i < s.length(); i++) {
120      char c = s.charAt(i);
121      if ((c < replacementsLength && replacements[c] != null) ||
122          c > safeMax || c < safeMin) {
123        return escapeSlow(s, i);
124      }
125    }
126    return s;
127  }
128
129  /**
130   * Escapes a single character using the replacement array and safe range
131   * values. If the given character does not have an explicit replacement and
132   * lies outside the safe range then {@link #escapeUnsafe} is called.
133   */
134  @Override protected final char[] escape(char c) {
135    if (c < replacementsLength) {
136      char[] chars = replacements[c];
137      if (chars != null) {
138        return chars;
139      }
140    }
141    if (c >= safeMin && c <= safeMax) {
142      return null;
143    }
144    return escapeUnsafe(c);
145  }
146
147  /**
148   * Escapes a {@code char} value that has no direct explicit value in the
149   * replacement array and lies outside the stated safe range. Subclasses should
150   * override this method to provide generalized escaping for characters.
151   *
152   * <p>Note that arrays returned by this method must not be modified once they
153   * have been returned. However it is acceptable to return the same array
154   * multiple times (even for different input characters).
155   *
156   * @param c the character to escape
157   * @return the replacement characters, or {@code null} if no escaping was
158   *         required
159   */
160  // TODO(user,cpovirk): Rename this something better once refactoring done
161  protected abstract char[] escapeUnsafe(char c);
162}