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.net.ftp.parser;
019
020import java.text.ParseException;
021import java.util.List;
022import java.util.regex.Pattern;
023
024import org.apache.commons.net.ftp.FTPClientConfig;
025import org.apache.commons.net.ftp.FTPFile;
026
027/**
028 * Implementation FTPFileEntryParser and FTPFileListParser for standard Unix Systems.
029 * <p>
030 * This class is based on the logic of Daniel Savarese's DefaultFTPListParser, but adapted to use regular expressions and to fit the new FTPFileEntryParser
031 * interface.
032 * </p>
033 *
034 * @see org.apache.commons.net.ftp.FTPFileEntryParser FTPFileEntryParser (for usage instructions)
035 */
036public class UnixFTPEntryParser extends ConfigurableFTPFileEntryParserImpl {
037
038    static final String DEFAULT_DATE_FORMAT = "MMM d yyyy"; // Nov 9 2001
039
040    static final String DEFAULT_RECENT_DATE_FORMAT = "MMM d HH:mm"; // Nov 9 20:06
041
042    static final String NUMERIC_DATE_FORMAT = "yyyy-MM-dd HH:mm"; // 2001-11-09 20:06
043
044    // Suffixes used in Japanese listings after the numeric values
045    private static final String JA_MONTH = "\u6708";
046    private static final String JA_DAY = "\u65e5";
047    private static final String JA_YEAR = "\u5e74";
048
049    private static final String DEFAULT_DATE_FORMAT_JA = "M'" + JA_MONTH + "' d'" + JA_DAY + "' yyyy'" + JA_YEAR + "'"; // 6月 3日 2003年
050    private static final String DEFAULT_DATE_FORMAT_CN = "M'" + JA_MONTH + "' d yyyy"; // 6月 3 2003
051
052    private static final String DEFAULT_RECENT_DATE_FORMAT_JA = "M'" + JA_MONTH + "' d'" + JA_DAY + "' HH:mm"; // 8月 17日 20:10
053    private static final String DEFAULT_RECENT_DATE_FORMAT_CN = "M'" + JA_MONTH + "' d HH:mm"; // 8月 17 20:10
054
055    private static final Pattern TOTAL_PATTERN = Pattern.compile("^total \\d+$");
056
057    /**
058     * Some Linux distributions are now shipping an FTP server which formats file listing dates in an all-numeric format: {@code "yyyy-MM-dd HH:mm}. This
059     * is a very welcome development, and hopefully it will soon become the standard. However, since it is so new, for now, and possibly forever, we merely
060     * accommodate it, but do not make it the default.
061     * <p>
062     * For now end users may specify this format only via {@code UnixFTPEntryParser(FTPClientConfig)}. Steve Cohen - 2005-04-17
063     * </p>
064     */
065    public static final FTPClientConfig NUMERIC_DATE_CONFIG = new FTPClientConfig(FTPClientConfig.SYST_UNIX, NUMERIC_DATE_FORMAT, null);
066
067    /**
068     * Regular expression used by this parser.
069     * <p>
070     * Permissions:
071     * </p>
072     * <ul>
073     * <li>r the file is readable</li>
074     * <li>w the file is writable</li>
075     * <li>x the file is executable</li>
076     * <li>- the indicated permission is not granted</li>
077     * <li>L mandatory locking occurs</li> during access (the set-group-ID bit is on and the group execution bit is off)</li>
078     * <li>s the set-user-ID or set-group-ID bit is on, and the corresponding user or group execution bit is also on
079     * <li>S undefined bit-state (the set-user-ID bit is on and the user execution bit is off)</li>
080     * <li>t the 1000 (octal) bit, or sticky bit, is on [see chmod(1)], and execution is on</li>
081     * <li>T the 1000 bit is turned on, and execution is off (undefined bit-state)</li>
082     * <li>e z/OS external link bit.</li>
083     * <p>
084     * Final letter may be appended: + file has extended security attributes (e.g. ACL) Note: local listings on MacOSX also use '@'; this is not allowed for
085     * here as does not appear to be shown by FTP servers {@code @} file has extended attributes
086     * </p>
087     */
088    private static final String REGEX = "([bcdelfmpSs-])" // file type
089            + "(((r|-)(w|-)([xsStTL-]))((r|-)(w|-)([xsStTL-]))((r|-)(w|-)([xsStTL-])))\\+?" // permissions
090            + "\\s*" // separator TODO why allow it to be omitted??
091            + "(\\d+)" // link count
092            + "\\s+" // separator
093            + "(?:(\\S+(?:\\s\\S+)*?)\\s+)?" // owner name (optional spaces)
094            + "(?:(\\S+(?:\\s\\S+)*)\\s+)?" // group name (optional spaces)
095            + "(\\d+(?:,\\s*\\d+)?)" // size or n,m
096            + "\\s+" // separator
097            /*
098             * numeric or standard format date: yyyy-mm-dd (expecting hh:mm to follow) MMM [d]d [d]d MMM Use non-space for MMM to allow for languages such
099             * as German which use diacritics (e.g. umlaut) in some abbreviations. Japanese uses numeric day and month with suffixes to distinguish them [d]dXX
100             * [d]dZZ
101             */
102            + "(" + "(?:\\d+[-/]\\d+[-/]\\d+)" + // yyyy-mm-dd
103            "|(?:\\S{3}\\s+\\d{1,2})" + // MMM [d]d
104            "|(?:\\d{1,2}\\s+\\S{3})" + // [d]d MMM
105            "|(?:\\d{1,2}" + JA_MONTH + "\\s+\\d{1,2}" + JA_DAY + ")" +
106            "|(?:\\d{1,2}" + JA_MONTH + "\\s+\\d{1,2})"
107            + ")"
108            + "\\s+" // separator
109            /*
110             * year (for non-recent standard format) - yyyy or time (for numeric or recent standard format) [h]h:mm or Japanese year - yyyyXX
111             */
112            + "((?:\\d+(?::\\d+)?)|(?:\\d{4}" + JA_YEAR + "))" // (20)
113            + "\\s" // separator
114            + "(.*)"; // the rest (21)
115
116    /**
117     * Whether leading spaces are trimmed from file names this was the case for the original implementation.
118     */
119    final boolean trimLeadingSpaces; // package protected for access from test code
120
121    /**
122     * Constructs a new instance.
123     *
124     * @throws IllegalArgumentException Thrown if the regular expression is unparseable. Should not be seen under normal conditions.
125     *                                  If this exception is seen, this is a sign that {@code REGEX} is not a valid regular expression.
126     */
127    public UnixFTPEntryParser() {
128        this(null);
129    }
130
131    /**
132     * Constructs a new instance with something other than the default configuration.
133     *
134     * @param config The {@link FTPClientConfig configuration} object used to configure this parser.
135     * @throws IllegalArgumentException Thrown if the regular expression is unparseable. Should not be seen under normal conditions.
136     *                                  If this exception is seen, this is a sign that {@code REGEX} is not a valid regular expression.
137     * @since 1.4
138     */
139    public UnixFTPEntryParser(final FTPClientConfig config) {
140        this(config, false);
141    }
142
143    /**
144     * Constructs a new instance  with something other than the default configuration.
145     *
146     * @param config            The {@link FTPClientConfig configuration} object used to configure this parser.
147     * @param trimLeadingSpaces if {@code true}, trim leading spaces from file names
148     * @throws IllegalArgumentException Thrown if the regular expression is unparseable. Should not be seen under normal conditions.
149     *                                  If this exception is seen, this is a sign that {@code REGEX} is not a valid regular expression.
150     * @since 3.4
151     */
152    public UnixFTPEntryParser(final FTPClientConfig config, final boolean trimLeadingSpaces) {
153        super(REGEX);
154        configure(config);
155        this.trimLeadingSpaces = trimLeadingSpaces;
156    }
157
158    /**
159     * Gets a new default configuration to be used when this class is instantiated without a {@link FTPClientConfig FTPClientConfig} parameter being specified.
160     *
161     * @return the default configuration for this parser.
162     */
163    @Override
164    protected FTPClientConfig getDefaultConfiguration() {
165        return new FTPClientConfig(FTPClientConfig.SYST_UNIX, DEFAULT_DATE_FORMAT, DEFAULT_RECENT_DATE_FORMAT);
166    }
167
168    /**
169     * Parses a line of a Unix (standard) FTP server file listing and converts it into a usable format in the form of an {@code FTPFile} instance. If the
170     * file listing line doesn't describe a file, {@code null} is returned, otherwise a {@code FTPFile} instance representing the files in the
171     * directory is returned.
172     *
173     * @param entry A line of text from the file listing
174     * @return An FTPFile instance corresponding to the supplied entry
175     */
176    @Override
177    public FTPFile parseFTPEntry(final String entry) {
178        final FTPFile file = new FTPFile();
179        file.setRawListing(entry);
180        final int type;
181        boolean isDevice = false;
182        if (matches(entry)) {
183            final String typeStr = group(1);
184            final String hardLinkCount = group(15);
185            final String usr = group(16);
186            final String grp = group(17);
187            final String fileSize = group(18);
188            final String datestr = group(19) + " " + group(20);
189            String name = group(21);
190            if (trimLeadingSpaces) {
191                name = name.replaceFirst("^\\s+", "");
192            }
193            try {
194                if (group(19).contains(JA_MONTH) && group(19).contains(JA_DAY)) { // special processing for Japanese format
195                    final FTPTimestampParserImpl jaParser = new FTPTimestampParserImpl();
196                    jaParser.configure(new FTPClientConfig(FTPClientConfig.SYST_UNIX, DEFAULT_DATE_FORMAT_JA, DEFAULT_RECENT_DATE_FORMAT_JA));
197                    file.setTimestamp(jaParser.parseTimestamp(datestr));
198                } else if (group(19).contains(JA_MONTH) && !group(19).contains(JA_DAY)) {
199                    final FTPTimestampParserImpl jaParser = new FTPTimestampParserImpl();
200                    jaParser.configure(new FTPClientConfig(FTPClientConfig.SYST_UNIX, DEFAULT_DATE_FORMAT_CN, DEFAULT_RECENT_DATE_FORMAT_CN));
201                    file.setTimestamp(jaParser.parseTimestamp(datestr));
202                } else {
203                    file.setTimestamp(super.parseTimestamp(datestr));
204                }
205            } catch (final ParseException e) {
206                // intentionally do nothing
207            }
208            // A 'whiteout' file is an ARTIFICIAL entry in any of several types of
209            // 'translucent' filesystems, of which a 'union' filesystem is one.
210            // bcdelfmpSs-
211            switch (typeStr.charAt(0)) {
212            case 'd':
213                type = FTPFile.DIRECTORY_TYPE;
214                break;
215            case 'e': // NET-39 => z/OS external link
216                type = FTPFile.SYMBOLIC_LINK_TYPE;
217                break;
218            case 'l':
219                type = FTPFile.SYMBOLIC_LINK_TYPE;
220                break;
221            case 'b':
222            case 'c':
223                isDevice = true;
224                type = FTPFile.FILE_TYPE; // TODO change this if DEVICE_TYPE implemented
225                break;
226            case 'f':
227            case '-':
228                type = FTPFile.FILE_TYPE;
229                break;
230            default: // e.g. ? and w = whiteout
231                type = FTPFile.UNKNOWN_TYPE;
232            }
233            file.setType(type);
234            int g = 4;
235            for (int access = 0; access < 3; access++, g += 4) {
236                // Use != '-' to avoid having to check for suid and sticky bits
237                file.setPermission(access, FTPFile.READ_PERMISSION, !group(g).equals("-"));
238                file.setPermission(access, FTPFile.WRITE_PERMISSION, !group(g + 1).equals("-"));
239                final String execPerm = group(g + 2);
240                file.setPermission(access, FTPFile.EXECUTE_PERMISSION, !execPerm.equals("-") && !Character.isUpperCase(execPerm.charAt(0)));
241            }
242            if (!isDevice) {
243                try {
244                    file.setHardLinkCount(Integer.parseInt(hardLinkCount));
245                } catch (final NumberFormatException e) {
246                    // intentionally do nothing
247                }
248            }
249            file.setUser(usr);
250            file.setGroup(grp);
251            try {
252                file.setSize(Long.parseLong(fileSize));
253            } catch (final NumberFormatException e) {
254                // intentionally do nothing
255            }
256            // oddball cases like symbolic links, file names
257            // with spaces in them.
258            if (type == FTPFile.SYMBOLIC_LINK_TYPE) {
259                final int end = name.indexOf(" -> ");
260                // Give up if no link indicator is present
261                if (end == -1) {
262                    file.setName(name);
263                } else {
264                    file.setName(name.substring(0, end));
265                    file.setLink(name.substring(end + 4));
266                }
267            } else {
268                file.setName(name);
269            }
270            return file;
271        }
272        return null;
273    }
274
275    /**
276     * Preparses the list to discard "total nnn" lines.
277     */
278    @Override
279    public List<String> preParse(final List<String> original) {
280        // NET-389
281        original.removeIf(entry -> TOTAL_PATTERN.matcher(entry).matches());
282        return original;
283    }
284
285}