001///////////////////////////////////////////////////////////////////////////////////////////////
002// checkstyle: Checks Java source code and other text files for adherence to a set of rules.
003// Copyright (C) 2001-2024 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.imports;
021
022import java.net.URI;
023import java.util.Set;
024import java.util.regex.Pattern;
025
026import com.puppycrawl.tools.checkstyle.FileStatefulCheck;
027import com.puppycrawl.tools.checkstyle.api.AbstractCheck;
028import com.puppycrawl.tools.checkstyle.api.CheckstyleException;
029import com.puppycrawl.tools.checkstyle.api.DetailAST;
030import com.puppycrawl.tools.checkstyle.api.ExternalResourceHolder;
031import com.puppycrawl.tools.checkstyle.api.FullIdent;
032import com.puppycrawl.tools.checkstyle.api.TokenTypes;
033import com.puppycrawl.tools.checkstyle.utils.CommonUtil;
034
035/**
036 * <div>
037 * Controls what can be imported in each package and file. Useful for ensuring
038 * that application layering rules are not violated, especially on large projects.
039 * </div>
040 *
041 * <p>
042 * You can control imports based on the package name or based on the file name.
043 * When controlling packages, all files and sub-packages in the declared package
044 * will be controlled by this check. To specify differences between a main package
045 * and a sub-package, you must define the sub-package inside the main package.
046 * When controlling file, only the file name is considered and only files processed by
047 * <a href="https://checkstyle.org/config.html#TreeWalker">TreeWalker</a>.
048 * The file's extension is ignored.
049 * </p>
050 *
051 * <p>
052 * Short description of the behaviour:
053 * </p>
054 * <ul>
055 * <li>
056 * Check starts checking from the longest matching subpackage (later 'current subpackage') or
057 * the first file name match described inside import control file to package defined in class file.
058 * <ul>
059 * <li>
060 * The longest matching subpackage is found by starting with the root package and
061 * examining if any of the sub-packages or file definitions match the current
062 * class' package or file name.
063 * </li>
064 * <li>
065 * If a file name is matched first, that is considered the longest match and becomes
066 * the current file/subpackage.
067 * </li>
068 * <li>
069 * If another subpackage is matched, then it's subpackages and file names are examined
070 * for the next longest match and the process repeats recursively.
071 * </li>
072 * <li>
073 * If no subpackages or file names are matched, the current subpackage is then used.
074 * </li>
075 * </ul>
076 * </li>
077 * <li>
078 * Order of rules in the same subpackage/root are defined by the order of declaration
079 * in the XML file, which is from top (first) to bottom (last).
080 * </li>
081 * <li>
082 * If there is matching allow/disallow rule inside the current file/subpackage
083 * then the Check returns the first "allowed" or "disallowed" message.
084 * </li>
085 * <li>
086 * If there is no matching allow/disallow rule inside the current file/subpackage
087 * then it continues checking in the parent subpackage.
088 * </li>
089 * <li>
090 * If there is no matching allow/disallow rule in any of the files/subpackages,
091 * including the root level (import-control), then the import is disallowed by default.
092 * </li>
093 * </ul>
094 *
095 * <p>
096 * The DTD for an import control XML document is at
097 * <a href="https://checkstyle.org/dtds/import_control_1_4.dtd">
098 * https://checkstyle.org/dtds/import_control_1_4.dtd</a>.
099 * It contains documentation on each of the elements and attributes.
100 * </p>
101 *
102 * <p>
103 * The check validates a XML document when it loads the document. To validate against
104 * the above DTD, include the following document type declaration in your XML document:
105 * </p>
106 * <pre>
107 * &lt;!DOCTYPE import-control PUBLIC
108 *     "-//Checkstyle//DTD ImportControl Configuration 1.4//EN"
109 *     "https://checkstyle.org/dtds/import_control_1_4.dtd"&gt;
110 * </pre>
111 * <ul>
112 * <li>
113 * Property {@code file} - Specify the location of the file containing the
114 * import control configuration. It can be a regular file, URL or resource path.
115 * It will try loading the path as a URL first, then as a file, and finally as a resource.
116 * Type is {@code java.net.URI}.
117 * Default value is {@code null}.
118 * </li>
119 * <li>
120 * Property {@code path} - Specify the regular expression of file paths to which
121 * this check should apply. Files that don't match the pattern will not be checked.
122 * The pattern will be matched against the full absolute file path.
123 * Type is {@code java.util.regex.Pattern}.
124 * Default value is {@code ".*"}.
125 * </li>
126 * </ul>
127 *
128 * <p>
129 * Parent is {@code com.puppycrawl.tools.checkstyle.TreeWalker}
130 * </p>
131 *
132 * <p>
133 * Violation Message Keys:
134 * </p>
135 * <ul>
136 * <li>
137 * {@code import.control.disallowed}
138 * </li>
139 * <li>
140 * {@code import.control.missing.file}
141 * </li>
142 * <li>
143 * {@code import.control.unknown.pkg}
144 * </li>
145 * </ul>
146 *
147 * @since 4.0
148 */
149@FileStatefulCheck
150public class ImportControlCheck extends AbstractCheck implements ExternalResourceHolder {
151
152    /**
153     * A key is pointing to the warning message text in "messages.properties"
154     * file.
155     */
156    public static final String MSG_MISSING_FILE = "import.control.missing.file";
157
158    /**
159     * A key is pointing to the warning message text in "messages.properties"
160     * file.
161     */
162    public static final String MSG_UNKNOWN_PKG = "import.control.unknown.pkg";
163
164    /**
165     * A key is pointing to the warning message text in "messages.properties"
166     * file.
167     */
168    public static final String MSG_DISALLOWED = "import.control.disallowed";
169
170    /**
171     * A part of message for exception.
172     */
173    private static final String UNABLE_TO_LOAD = "Unable to load ";
174
175    /**
176     * Specify the location of the file containing the import control configuration.
177     * It can be a regular file, URL or resource path. It will try loading the path
178     * as a URL first, then as a file, and finally as a resource.
179     */
180    private URI file;
181
182    /**
183     * Specify the regular expression of file paths to which this check should apply.
184     * Files that don't match the pattern will not be checked. The pattern will
185     * be matched against the full absolute file path.
186     */
187    private Pattern path = Pattern.compile(".*");
188    /** Whether to process the current file. */
189    private boolean processCurrentFile;
190
191    /** The root package controller. */
192    private PkgImportControl root;
193    /** The package doing the import. */
194    private String packageName;
195    /** The file name doing the import. */
196    private String fileName;
197
198    /**
199     * The package controller for the current file. Used for performance
200     * optimisation.
201     */
202    private AbstractImportControl currentImportControl;
203
204    @Override
205    public int[] getDefaultTokens() {
206        return getRequiredTokens();
207    }
208
209    @Override
210    public int[] getAcceptableTokens() {
211        return getRequiredTokens();
212    }
213
214    @Override
215    public int[] getRequiredTokens() {
216        return new int[] {TokenTypes.PACKAGE_DEF, TokenTypes.IMPORT, TokenTypes.STATIC_IMPORT, };
217    }
218
219    @Override
220    public void beginTree(DetailAST rootAST) {
221        currentImportControl = null;
222        final String fullFileName = getFilePath();
223        processCurrentFile = path.matcher(fullFileName).find();
224        fileName = CommonUtil.getFileNameWithoutExtension(fullFileName);
225    }
226
227    @Override
228    public void visitToken(DetailAST ast) {
229        if (processCurrentFile) {
230            if (ast.getType() == TokenTypes.PACKAGE_DEF) {
231                if (root == null) {
232                    log(ast, MSG_MISSING_FILE);
233                }
234                else {
235                    packageName = getPackageText(ast);
236                    currentImportControl = root.locateFinest(packageName, fileName);
237                    if (currentImportControl == null) {
238                        log(ast, MSG_UNKNOWN_PKG);
239                    }
240                }
241            }
242            else if (currentImportControl != null) {
243                final String importText = getImportText(ast);
244                final AccessResult access = currentImportControl.checkAccess(packageName, fileName,
245                        importText);
246                if (access != AccessResult.ALLOWED) {
247                    log(ast, MSG_DISALLOWED, importText);
248                }
249            }
250        }
251    }
252
253    @Override
254    public Set<String> getExternalResourceLocations() {
255        return Set.of(file.toASCIIString());
256    }
257
258    /**
259     * Returns package text.
260     *
261     * @param ast PACKAGE_DEF ast node
262     * @return String that represents full package name
263     */
264    private static String getPackageText(DetailAST ast) {
265        final DetailAST nameAST = ast.getLastChild().getPreviousSibling();
266        return FullIdent.createFullIdent(nameAST).getText();
267    }
268
269    /**
270     * Returns import text.
271     *
272     * @param ast ast node that represents import
273     * @return String that represents importing class
274     */
275    private static String getImportText(DetailAST ast) {
276        final FullIdent imp;
277        if (ast.getType() == TokenTypes.IMPORT) {
278            imp = FullIdent.createFullIdentBelow(ast);
279        }
280        else {
281            // know it is a static import
282            imp = FullIdent.createFullIdent(ast
283                    .getFirstChild().getNextSibling());
284        }
285        return imp.getText();
286    }
287
288    /**
289     * Setter to specify the location of the file containing the import control configuration.
290     * It can be a regular file, URL or resource path. It will try loading the path
291     * as a URL first, then as a file, and finally as a resource.
292     *
293     * @param uri the uri of the file to load.
294     * @throws IllegalArgumentException on error loading the file.
295     * @since 4.0
296     */
297    public void setFile(URI uri) {
298        // Handle empty param
299        if (uri != null) {
300            try {
301                root = ImportControlLoader.load(uri);
302                file = uri;
303            }
304            catch (CheckstyleException ex) {
305                throw new IllegalArgumentException(UNABLE_TO_LOAD + uri, ex);
306            }
307        }
308    }
309
310    /**
311     * Setter to specify the regular expression of file paths to which this check should apply.
312     * Files that don't match the pattern will not be checked. The pattern will be matched
313     * against the full absolute file path.
314     *
315     * @param pattern the file path regex this check should apply to.
316     * @since 7.5
317     */
318    public void setPath(Pattern pattern) {
319        path = pattern;
320    }
321}