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 *     http://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.configuration2.convert;
018
019import java.util.ArrayList;
020import java.util.Collection;
021import java.util.Collections;
022import java.util.Iterator;
023import java.util.List;
024
025import org.apache.commons.lang3.StringUtils;
026
027/**
028 * <p>
029 * A specialized implementation of {@code ListDelimiterHandler} which simulates the list delimiter handling as it was
030 * used by {@code PropertiesConfiguration} in Commons Configuration 1.x.
031 * </p>
032 * <p>
033 * This class mainly exists for compatibility reasons. It is intended to be used by applications which have to deal with
034 * properties files created by an older version of this library.
035 * </p>
036 * <p>
037 * In the 1.x series of Commons Configuration list handling was not fully consistent. The escaping of property values
038 * was done in a different way if they contained a list delimiter or not. From version 2.0 on, escaping is more
039 * stringent which might cause slightly different results when parsing properties files created by or for Configuration
040 * 1.x. If you encounter such problems, you can switch to this {@code ListDelimiterHandler} implementation rather than
041 * the default one. In other cases, this class should not be used!
042 * </p>
043 * <p>
044 * Implementation note: An instance of this class can safely be shared between multiple {@code Configuration} instances.
045 * </p>
046 *
047 * @since 2.0
048 */
049public class LegacyListDelimiterHandler extends AbstractListDelimiterHandler {
050    /** Constant for the escaping character. */
051    private static final String ESCAPE = "\\";
052
053    /** Constant for the escaped escaping character. */
054    private static final String DOUBLE_ESC = ESCAPE + ESCAPE;
055
056    /** Constant for a duplicated sequence of escaping characters. */
057    private static final String QUAD_ESC = DOUBLE_ESC + DOUBLE_ESC;
058
059    /** The list delimiter character. */
060    private final char delimiter;
061
062    /**
063     * Creates a new instance of {@code LegacyListDelimiterHandler} and sets the list delimiter character.
064     *
065     * @param listDelimiter the list delimiter character
066     */
067    public LegacyListDelimiterHandler(final char listDelimiter) {
068        delimiter = listDelimiter;
069    }
070
071    /**
072     * Gets the list delimiter character.
073     *
074     * @return the list delimiter character
075     */
076    public char getDelimiter() {
077        return delimiter;
078    }
079
080    /**
081     * {@inheritDoc} This implementation performs delimiter escaping for a single value (which is not part of a list).
082     */
083    @Override
084    public Object escape(final Object value, final ValueTransformer transformer) {
085        return escapeValue(value, false, transformer);
086    }
087
088    /**
089     * {@inheritDoc} This implementation performs a special encoding of backslashes at the end of a string so that they are
090     * not interpreted as escape character for a following list delimiter.
091     */
092    @Override
093    public Object escapeList(final List<?> values, final ValueTransformer transformer) {
094        if (!values.isEmpty()) {
095            final Iterator<?> it = values.iterator();
096            String lastValue = escapeValue(it.next(), true, transformer);
097            final StringBuilder buf = new StringBuilder(lastValue);
098            while (it.hasNext()) {
099                // if the last value ended with an escape character, it has
100                // to be escaped itself; otherwise the list delimiter will
101                // be escaped
102                if (lastValue.endsWith(ESCAPE) && countTrailingBS(lastValue) / 2 % 2 != 0) {
103                    buf.append(ESCAPE).append(ESCAPE);
104                }
105                buf.append(getDelimiter());
106                lastValue = escapeValue(it.next(), true, transformer);
107                buf.append(lastValue);
108            }
109            return buf.toString();
110        }
111        return null;
112    }
113
114    /**
115     * {@inheritDoc} This implementation simulates the old splitting algorithm. The string is split at the delimiter
116     * character if it is not escaped. If the delimiter character is not found, the input is returned unchanged.
117     */
118    @Override
119    protected Collection<String> splitString(final String s, final boolean trim) {
120        if (s.indexOf(getDelimiter()) < 0) {
121            return Collections.singleton(s);
122        }
123
124        final List<String> list = new ArrayList<>();
125
126        StringBuilder token = new StringBuilder();
127        int begin = 0;
128        boolean inEscape = false;
129        final char esc = ESCAPE.charAt(0);
130
131        while (begin < s.length()) {
132            final char c = s.charAt(begin);
133            if (inEscape) {
134                // last character was the escape marker
135                // can current character be escaped?
136                if (c != getDelimiter() && c != esc) {
137                    // no, also add escape character
138                    token.append(esc);
139                }
140                token.append(c);
141                inEscape = false;
142            } else if (c == getDelimiter()) {
143                // found a list delimiter -> add token and
144                // resetDefaultFileSystem buffer
145                String t = token.toString();
146                if (trim) {
147                    t = t.trim();
148                }
149                list.add(t);
150                token = new StringBuilder();
151            } else if (c == esc) {
152                // eventually escape next character
153                inEscape = true;
154            } else {
155                token.append(c);
156            }
157
158            begin++;
159        }
160
161        // Trailing delimiter?
162        if (inEscape) {
163            token.append(esc);
164        }
165        // Add last token
166        String t = token.toString();
167        if (trim) {
168            t = t.trim();
169        }
170        list.add(t);
171
172        return list;
173    }
174
175    /**
176     * {@inheritDoc} This is just a dummy implementation. It is never called.
177     */
178    @Override
179    protected String escapeString(final String s) {
180        return null;
181    }
182
183    /**
184     * Performs the escaping of backslashes in the specified properties value. Because a double backslash is used to escape
185     * the escape character of a list delimiter, double backslashes also have to be escaped if the property is part of a
186     * (single line) list. In addition, because the output is written into a properties file, each occurrence of a backslash
187     * again has to be doubled. This method is called by {@code escapeValue()}.
188     *
189     * @param value the value to be escaped
190     * @param inList a flag whether the value is part of a list
191     * @return the value with escaped backslashes as string
192     */
193    protected String escapeBackslashs(final Object value, final boolean inList) {
194        String strValue = String.valueOf(value);
195
196        if (inList && strValue.contains(DOUBLE_ESC)) {
197            strValue = StringUtils.replace(strValue, DOUBLE_ESC, QUAD_ESC);
198        }
199
200        return strValue;
201    }
202
203    /**
204     * Escapes the given property value. This method is called on saving the configuration for each property value. It
205     * ensures a correct handling of backslash characters and also takes care that list delimiter characters in the value
206     * are escaped.
207     *
208     * @param value the property value
209     * @param inList a flag whether the value is part of a list
210     * @param transformer the {@code ValueTransformer}
211     * @return the escaped property value
212     */
213    protected String escapeValue(final Object value, final boolean inList, final ValueTransformer transformer) {
214        String escapedValue = String.valueOf(transformer.transformValue(escapeBackslashs(value, inList)));
215        if (getDelimiter() != 0) {
216            escapedValue = StringUtils.replace(escapedValue, String.valueOf(getDelimiter()), ESCAPE + getDelimiter());
217        }
218        return escapedValue;
219    }
220
221    /**
222     * Returns the number of trailing backslashes. This is sometimes needed for the correct handling of escape characters.
223     *
224     * @param line the string to investigate
225     * @return the number of trailing backslashes
226     */
227    private static int countTrailingBS(final String line) {
228        int bsCount = 0;
229        for (int idx = line.length() - 1; idx >= 0 && line.charAt(idx) == '\\'; idx--) {
230            bsCount++;
231        }
232
233        return bsCount;
234    }
235}