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 * https://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
18 package org.apache.commons.net.ftp.parser;
19
20 import java.text.ParseException;
21 import java.util.List;
22 import java.util.regex.Pattern;
23
24 import org.apache.commons.net.ftp.FTPClientConfig;
25 import org.apache.commons.net.ftp.FTPFile;
26
27 /**
28 * Implementation FTPFileEntryParser and FTPFileListParser for standard Unix Systems.
29 * <p>
30 * This class is based on the logic of Daniel Savarese's DefaultFTPListParser, but adapted to use regular expressions and to fit the new FTPFileEntryParser
31 * interface.
32 * </p>
33 *
34 * @see org.apache.commons.net.ftp.FTPFileEntryParser FTPFileEntryParser (for usage instructions)
35 */
36 public class UnixFTPEntryParser extends ConfigurableFTPFileEntryParserImpl {
37
38 static final String DEFAULT_DATE_FORMAT = "MMM d yyyy"; // Nov 9 2001
39
40 static final String DEFAULT_RECENT_DATE_FORMAT = "MMM d HH:mm"; // Nov 9 20:06
41
42 static final String NUMERIC_DATE_FORMAT = "yyyy-MM-dd HH:mm"; // 2001-11-09 20:06
43
44 // Suffixes used in Japanese listings after the numeric values
45 private static final String JA_MONTH = "\u6708";
46 private static final String JA_DAY = "\u65e5";
47 private static final String JA_YEAR = "\u5e74";
48
49 private static final String DEFAULT_DATE_FORMAT_JA = "M'" + JA_MONTH + "' d'" + JA_DAY + "' yyyy'" + JA_YEAR + "'"; // 6月 3日 2003年
50 private static final String DEFAULT_DATE_FORMAT_CN = "M'" + JA_MONTH + "' d yyyy"; // 6月 3 2003
51
52 private static final String DEFAULT_RECENT_DATE_FORMAT_JA = "M'" + JA_MONTH + "' d'" + JA_DAY + "' HH:mm"; // 8月 17日 20:10
53 private static final String DEFAULT_RECENT_DATE_FORMAT_CN = "M'" + JA_MONTH + "' d HH:mm"; // 8月 17 20:10
54
55 private static final Pattern TOTAL_PATTERN = Pattern.compile("^total \\d+$");
56
57 /**
58 * 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
59 * 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
60 * accommodate it, but do not make it the default.
61 * <p>
62 * For now end users may specify this format only via {@code UnixFTPEntryParser(FTPClientConfig)}. Steve Cohen - 2005-04-17
63 * </p>
64 */
65 public static final FTPClientConfig NUMERIC_DATE_CONFIG = new FTPClientConfig(FTPClientConfig.SYST_UNIX, NUMERIC_DATE_FORMAT, null);
66
67 /**
68 * Regular expression used by this parser.
69 * <p>
70 * Permissions:
71 * </p>
72 * <ul>
73 * <li>r the file is readable</li>
74 * <li>w the file is writable</li>
75 * <li>x the file is executable</li>
76 * <li>- the indicated permission is not granted</li>
77 * <li>L mandatory locking occurs</li> during access (the set-group-ID bit is on and the group execution bit is off)</li>
78 * <li>s the set-user-ID or set-group-ID bit is on, and the corresponding user or group execution bit is also on
79 * <li>S undefined bit-state (the set-user-ID bit is on and the user execution bit is off)</li>
80 * <li>t the 1000 (octal) bit, or sticky bit, is on [see chmod(1)], and execution is on</li>
81 * <li>T the 1000 bit is turned on, and execution is off (undefined bit-state)</li>
82 * <li>e z/OS external link bit.</li>
83 * <p>
84 * 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
85 * here as does not appear to be shown by FTP servers {@code @} file has extended attributes
86 * </p>
87 */
88 private static final String REGEX = "([bcdelfmpSs-])" // file type
89 + "(((r|-)(w|-)([xsStTL-]))((r|-)(w|-)([xsStTL-]))((r|-)(w|-)([xsStTL-])))\\+?" // permissions
90 + "\\s*" // separator TODO why allow it to be omitted??
91 + "(\\d+)" // link count
92 + "\\s+" // separator
93 + "(?:(\\S+(?:\\s\\S+)*?)\\s+)?" // owner name (optional spaces)
94 + "(?:(\\S+(?:\\s\\S+)*)\\s+)?" // group name (optional spaces)
95 + "(\\d+(?:,\\s*\\d+)?)" // size or n,m
96 + "\\s+" // separator
97 /*
98 * 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
99 * 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 }