Source code

001////////////////////////////////////////////////////////////////////////////////
002// checkstyle: Checks Java source code for adherence to a set of rules.
003// Copyright (C) 2001-2020 the original author or authors.
004//
005// This library is free software; you can redistribute it and/or
006// modify it under the terms of the GNU Lesser General Public
007// License as published by the Free Software Foundation; either
008// version 2.1 of the License, or (at your option) any later version.
009//
010// This library is distributed in the hope that it will be useful,
011// but WITHOUT ANY WARRANTY; without even the implied warranty of
012// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
013// Lesser General Public License for more details.
014//
015// You should have received a copy of the GNU Lesser General Public
016// License along with this library; if not, write to the Free Software
017// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
018////////////////////////////////////////////////////////////////////////////////
019
020package com.puppycrawl.tools.checkstyle.checks.header;
021
022import java.io.File;
023import java.util.ArrayList;
024import java.util.Arrays;
025import java.util.List;
026import java.util.regex.Pattern;
027import java.util.regex.PatternSyntaxException;
028
029import com.puppycrawl.tools.checkstyle.StatelessCheck;
030import com.puppycrawl.tools.checkstyle.api.FileText;
031import com.puppycrawl.tools.checkstyle.utils.CommonUtil;
032
033/**
034 * <p>
035 * Checks the header of a source file against a header that contains a
036 * <a href="https://docs.oracle.com/en/java/javase/11/docs/api/java.base/java/util/regex/Pattern.html">
037 * regular expression</a> for each line of the source header.
038 * </p>
039 * <p>
040 * Rationale: In some projects <a href="https://checkstyle.org/config_header.html#Header">
041 * checking against a fixed header</a> is not sufficient, e.g. the header might
042 * require a copyright line where the year information is not static.
043 * </p>
044 * <p>
045 * For example, consider the following header:
046 * </p>
047 * <pre>
048 * line  1: ^/{71}$
049 * line  2: ^// checkstyle:$
050 * line  3: ^// Checks Java source code for adherence to a set of rules\.$
051 * line  4: ^// Copyright \(C\) \d\d\d\d  Oliver Burn$
052 * line  5: ^// Last modification by \$Author.*\$$
053 * line  6: ^/{71}$
054 * line  7:
055 * line  8: ^package
056 * line  9:
057 * line 10: ^import
058 * line 11:
059 * line 12: ^/\*\*
060 * line 13: ^ \*([^/]|$)
061 * line 14: ^ \*&#47;
062 * </pre>
063 * <p>
064 * Lines 1 and 6 demonstrate a more compact notation for 71 '/' characters.
065 * Line 4 enforces that the copyright notice includes a four digit year.
066 * Line 5 is an example how to enforce revision control keywords in a file header.
067 * Lines 12-14 is a template for javadoc (line 13 is so complicated to remove
068 * conflict with and of javadoc comment). Lines 7, 9 and 11 will be treated
069 * as '^$' and will forcefully expect the line to be empty.
070 * </p>
071 * <p>
072 * Different programming languages have different comment syntax rules,
073 * but all of them start a comment with a non-word character.
074 * Hence you can often use the non-word character class to abstract away
075 * the concrete comment syntax and allow checking the header for different
076 * languages with a single header definition. For example, consider the following
077 * header specification (note that this is not the full Apache license header):
078 * </p>
079 * <pre>
080 * line 1: ^#!
081 * line 2: ^&lt;\?xml.*&gt;$
082 * line 3: ^\W*$
083 * line 4: ^\W*Copyright 2006 The Apache Software Foundation or its licensors, as applicable\.$
084 * line 5: ^\W*Licensed under the Apache License, Version 2\.0 \(the "License"\);$
085 * line 6: ^\W*$
086 * </pre>
087 * <p>
088 * Lines 1 and 2 leave room for technical header lines, e.g. the "#!/bin/sh"
089 * line in Unix shell scripts, or the XML file header of XML files.
090 * Set the multiline property to "1, 2" so these lines can be ignored for
091 * file types where they do no apply. Lines 3 through 6 define the actual header content.
092 * Note how lines 2, 4 and 5 use escapes for characters that have special regexp semantics.
093 * </p>
094 * <p>
095 * In default configuration, if header is not specified, the default value
096 * of header is set to null and the check does not rise any violations.
097 * </p>
098 * <ul>
099 * <li>
100 * Property {@code headerFile} - Specify the name of the file containing the required header.
101 * Default value is {@code null}.
102 * </li>
103 * <li>
104 * Property {@code charset} - Specify the character encoding to use when reading the headerFile.
105 * Default value is the charset property of the parent
106 * <a href="https://checkstyle.org/config.html#Checker">Checker</a> module.
107 * </li>
108 * <li>
109 * Property {@code header} - Define the required header specified inline.
110 * Individual header lines must be separated by the string {@code "\n"}
111 * (even on platforms with a different line separator).
112 * For header lines containing {@code "\n\n"} checkstyle will
113 * forcefully expect an empty line to exist. See examples below.
114 * Regular expressions must not span multiple lines.
115 * Default value is {@code null}.
116 * </li>
117 * <li>
118 * Property {@code multiLines} - Specify the line numbers to repeat (zero or more times).
119 * Default value is {@code {}}.
120 * </li>
121 * <li>
122 * Property {@code fileExtensions} - Specify the file type extension of files to process.
123 * Default value is {@code all files}.
124 * </li>
125 * </ul>
126 * <p>
127 * In default configuration the check does not rise any violations.
128 * Default values of properties are used.
129 * </p>
130 * <pre>
131 * &lt;module name="RegexpHeader"/&gt;
132 * </pre>
133 * <p>
134 * To configure the check to use header file {@code "config/java.header"} and
135 * {@code 10} and {@code 13} multi-lines:
136 * </p>
137 * <pre>
138 * &lt;module name="RegexpHeader"&gt;
139 *   &lt;property name="headerFile" value="config/java.header"/&gt;
140 *   &lt;property name="multiLines" value="10, 13"/&gt;
141 * &lt;/module&gt;
142 * </pre>
143 * <p>
144 * To configure the check to verify that each file starts with the header
145 * </p>
146 * <pre>
147 * ^// Copyright \(C\) (\d\d\d\d -)? 2004 MyCompany$
148 * ^// All rights reserved$
149 * </pre>
150 * <p>
151 * without the need for an external header file:
152 * </p>
153 * <pre>
154 * &lt;module name="RegexpHeader"&gt;
155 *   &lt;property
156 *     name="header"
157 *     value="^// Copyright \(C\) (\d\d\d\d -)? 2004 MyCompany$
158 *       \n^// All rights reserved$"/&gt;
159 * &lt;/module&gt;
160 * </pre>
161 * <p>
162 * For regex containing {@code "\n\n"}
163 * </p>
164 * <pre>
165 * &lt;module name="RegexpHeader"&gt;
166 *   &lt;property
167 *     name="header"
168 *     value="^package .*\n\n.*"/&gt;
169 * &lt;/module&gt;
170 * </pre>
171 * <p>
172 * {@code "\n\n"} will be treated as '^$' and will forcefully expect the line
173 * to be empty. For example -
174 * </p>
175 * <pre>
176 * package com.some.package;
177 * public class ThisWillFail { }
178 * </pre>
179 * <p>
180 * would fail for the regex above. Expected -
181 * </p>
182 * <pre>
183 * package com.some.package;
184 *
185 * public class ThisWillPass { }
186 * </pre>
187 * <p>
188 * <u>Note</u>: {@code ignoreLines} property has been removed from this check to simplify it.
189 * To make some line optional use "^.*$" regexp for this line.
190 * </p>
191 *
192 * @since 6.9
193 */
194@StatelessCheck
195public class RegexpHeaderCheck extends AbstractHeaderCheck {
196
197    /**
198     * A key is pointing to the warning message text in "messages.properties"
199     * file.
200     */
201    public static final String MSG_HEADER_MISSING = "header.missing";
202
203    /**
204     * A key is pointing to the warning message text in "messages.properties"
205     * file.
206     */
207    public static final String MSG_HEADER_MISMATCH = "header.mismatch";
208
209    /** Empty array to avoid instantiations. */
210    private static final int[] EMPTY_INT_ARRAY = new int[0];
211
212    /** Regex pattern for a blank line. **/
213    private static final String EMPTY_LINE_PATTERN = "^$";
214
215    /** Compiled regex pattern for a blank line. **/
216    private static final Pattern BLANK_LINE = Pattern.compile(EMPTY_LINE_PATTERN);
217
218    /** The compiled regular expressions. */
219    private final List<Pattern> headerRegexps = new ArrayList<>();
220
221    /** Specify the line numbers to repeat (zero or more times). */
222    private int[] multiLines = EMPTY_INT_ARRAY;
223
224    /**
225     * Setter to specify the line numbers to repeat (zero or more times).
226     *
227     * @param list comma separated list of line numbers to repeat in header.
228     */
229    public void setMultiLines(int... list) {
230        if (list.length == 0) {
231            multiLines = EMPTY_INT_ARRAY;
232        }
233        else {
234            multiLines = new int[list.length];
235            System.arraycopy(list, 0, multiLines, 0, list.length);
236            Arrays.sort(multiLines);
237        }
238    }
239
240    @Override
241    protected void processFiltered(File file, FileText fileText) {
242        final int headerSize = getHeaderLines().size();
243        final int fileSize = fileText.size();
244
245        if (headerSize - multiLines.length > fileSize) {
246            log(1, MSG_HEADER_MISSING);
247        }
248        else {
249            int headerLineNo = 0;
250            int index;
251            for (index = 0; headerLineNo < headerSize && index < fileSize; index++) {
252                final String line = fileText.get(index);
253                boolean isMatch = isMatch(line, headerLineNo);
254                while (!isMatch && isMultiLine(headerLineNo)) {
255                    headerLineNo++;
256                    isMatch = headerLineNo == headerSize
257                            || isMatch(line, headerLineNo);
258                }
259                if (!isMatch) {
260                    log(index + 1, MSG_HEADER_MISMATCH, getHeaderLine(headerLineNo));
261                    break;
262                }
263                if (!isMultiLine(headerLineNo)) {
264                    headerLineNo++;
265                }
266            }
267            if (index == fileSize) {
268                // if file finished, but we have at least one non-multi-line
269                // header isn't completed
270                logFirstSinglelineLine(headerLineNo, headerSize);
271            }
272        }
273    }
274
275    /**
276     * Returns the line from the header. Where the line is blank return the regexp pattern
277     * for a blank line.
278     * @param headerLineNo header line number to return
279     * @return the line from the header
280     */
281    private String getHeaderLine(int headerLineNo) {
282        String line = getHeaderLines().get(headerLineNo);
283        if (line.isEmpty()) {
284            line = EMPTY_LINE_PATTERN;
285        }
286        return line;
287    }
288
289    /**
290     * Logs warning if any non-multiline lines left in header regexp.
291     * @param startHeaderLine header line number to start from
292     * @param headerSize whole header size
293     */
294    private void logFirstSinglelineLine(int startHeaderLine, int headerSize) {
295        for (int lineNum = startHeaderLine; lineNum < headerSize; lineNum++) {
296            if (!isMultiLine(lineNum)) {
297                log(1, MSG_HEADER_MISSING);
298                break;
299            }
300        }
301    }
302
303    /**
304     * Checks if a code line matches the required header line.
305     * @param line the code line
306     * @param headerLineNo the header line number.
307     * @return true if and only if the line matches the required header line.
308     */
309    private boolean isMatch(String line, int headerLineNo) {
310        return headerRegexps.get(headerLineNo).matcher(line).find();
311    }
312
313    /**
314     * Returns true if line is multiline header lines or false.
315     * @param lineNo a line number
316     * @return if {@code lineNo} is one of the repeat header lines.
317     */
318    private boolean isMultiLine(int lineNo) {
319        return Arrays.binarySearch(multiLines, lineNo + 1) >= 0;
320    }
321
322    @Override
323    protected void postProcessHeaderLines() {
324        final List<String> headerLines = getHeaderLines();
325        for (String line : headerLines) {
326            try {
327                if (line.isEmpty()) {
328                    headerRegexps.add(BLANK_LINE);
329                }
330                else {
331                    headerRegexps.add(Pattern.compile(line));
332                }
333            }
334            catch (final PatternSyntaxException ex) {
335                throw new IllegalArgumentException("line "
336                        + (headerRegexps.size() + 1)
337                        + " in header specification"
338                        + " is not a regular expression", ex);
339            }
340        }
341    }
342
343    /**
344     * Setter to define the required header specified inline.
345     * Individual header lines must be separated by the string {@code "\n"}
346     * (even on platforms with a different line separator).
347     * For header lines containing {@code "\n\n"} checkstyle will forcefully
348     * expect an empty line to exist. See examples below.
349     * Regular expressions must not span multiple lines.
350     *
351     * @param header the header value to validate and set (in that order)
352     */
353    @Override
354    public void setHeader(String header) {
355        if (!CommonUtil.isBlank(header)) {
356            if (!CommonUtil.isPatternValid(header)) {
357                throw new IllegalArgumentException("Unable to parse format: " + header);
358            }
359            super.setHeader(header);
360        }
361    }
362
363}