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.io;
018
019import java.io.File;
020import java.net.MalformedURLException;
021import java.net.URI;
022import java.net.URL;
023import java.util.Arrays;
024import java.util.Map;
025
026import org.apache.commons.configuration2.ex.ConfigurationException;
027import org.apache.commons.lang3.ObjectUtils;
028import org.apache.commons.lang3.StringUtils;
029import org.apache.commons.logging.Log;
030import org.apache.commons.logging.LogFactory;
031
032/**
033 * <p>
034 * A utility class providing helper methods related to locating files.
035 * </p>
036 * <p>
037 * The methods of this class are used behind the scenes when retrieving configuration files based on different criteria,
038 * e.g. URLs, files, or more complex search strategies. They also implement functionality required by the default
039 * {@link FileSystem} implementations. Most methods are intended to be used internally only by other classes in the
040 * {@code io} package.
041 * </p>
042 *
043 * @since 2.0
044 */
045public final class FileLocatorUtils {
046    /**
047     * Constant for the default {@code FileSystem}. This file system is used by operations of this class if no specific file
048     * system is provided. An instance of {@link DefaultFileSystem} is used.
049     */
050    public static final FileSystem DEFAULT_FILE_SYSTEM = new DefaultFileSystem();
051
052    /**
053     * Constant for the default {@code FileLocationStrategy}. This strategy is used by the {@code locate()} method if the
054     * passed in {@code FileLocator} does not define its own location strategy. The default location strategy is roughly
055     * equivalent to the search algorithm used in version 1.x of <em>Commons Configuration</em> (there it was hard-coded
056     * though). It behaves in the following way when passed a {@code FileLocator}:
057     * <ul>
058     * <li>If the {@code FileLocator} has a defined URL, this URL is used as the file's URL (without any further
059     * checks).</li>
060     * <li>Otherwise, base path and file name stored in the {@code FileLocator} are passed to the current
061     * {@code FileSystem}'s {@code locateFromURL()} method. If this results in a URL, it is returned.</li>
062     * <li>Otherwise, if the locator's file name is an absolute path to an existing file, the URL of this file is
063     * returned.</li>
064     * <li>Otherwise, the concatenation of base path and file name is constructed. If this path points to an existing file,
065     * its URL is returned.</li>
066     * <li>Otherwise, a sub directory of the current user's home directory as defined by the base path is searched for the
067     * referenced file. If the file can be found there, its URL is returned.</li>
068     * <li>Otherwise, the base path is ignored, and the file name is searched in the current user's home directory. If the
069     * file can be found there, its URL is returned.</li>
070     * <li>Otherwise, a resource with the name of the locator's file name is searched in the classpath. If it can be found,
071     * its URL is returned.</li>
072     * <li>Otherwise, the strategy gives up and returns <b>null</b> indicating that the file cannot be resolved.</li>
073     * </ul>
074     */
075    public static final FileLocationStrategy DEFAULT_LOCATION_STRATEGY = initDefaultLocationStrategy();
076
077    /** Constant for the file URL protocol */
078    private static final String FILE_SCHEME = "file:";
079
080    /** The logger. */
081    private static final Log LOG = LogFactory.getLog(FileLocatorUtils.class);
082
083    /** Property key for the base path. */
084    private static final String PROP_BASE_PATH = "basePath";
085
086    /** Property key for the encoding. */
087    private static final String PROP_ENCODING = "encoding";
088
089    /** Property key for the file name. */
090    private static final String PROP_FILE_NAME = "fileName";
091
092    /** Property key for the file system. */
093    private static final String PROP_FILE_SYSTEM = "fileSystem";
094
095    /** Property key for the location strategy. */
096    private static final String PROP_STRATEGY = "locationStrategy";
097
098    /** Property key for the source URL. */
099    private static final String PROP_SOURCE_URL = "sourceURL";
100
101    /**
102     * Private constructor so that no instances can be created.
103     */
104    private FileLocatorUtils() {
105    }
106
107    /**
108     * Extends a path by another component. The given extension is added to the already existing path adding a separator if
109     * necessary.
110     *
111     * @param path the path to be extended
112     * @param ext the extension of the path
113     * @return the extended path
114     */
115    static String appendPath(final String path, final String ext) {
116        final StringBuilder fName = new StringBuilder();
117        fName.append(path);
118
119        // My best friend. Paranoia.
120        if (!path.endsWith(File.separator)) {
121            fName.append(File.separator);
122        }
123
124        //
125        // We have a relative path, and we have
126        // two possible forms here. If we have the
127        // "./" form then just strip that off first
128        // before continuing.
129        //
130        if (ext.startsWith("." + File.separator)) {
131            fName.append(ext.substring(2));
132        } else {
133            fName.append(ext);
134        }
135        return fName.toString();
136    }
137
138    /**
139     * Helper method for constructing a file object from a base path and a file name. This method is called if the base path
140     * passed to {@code getURL()} does not seem to be a valid URL.
141     *
142     * @param basePath the base path
143     * @param fileName the file name (must not be <b>null</b>)
144     * @return the resulting file
145     */
146    static File constructFile(final String basePath, final String fileName) {
147        final File file;
148
149        final File absolute = new File(fileName);
150        if (StringUtils.isEmpty(basePath) || absolute.isAbsolute()) {
151            file = absolute;
152        } else {
153            file = new File(appendPath(basePath, fileName));
154        }
155
156        return file;
157    }
158
159    /**
160     * Tries to convert the specified file to a URL. If this causes an exception, result is <b>null</b>.
161     *
162     * @param file the file to be converted
163     * @return the resulting URL or <b>null</b>
164     */
165    static URL convertFileToURL(final File file) {
166        return convertURIToURL(file.toURI());
167    }
168
169    /**
170     * Tries to convert the specified URI to a URL. If this causes an exception, result is <b>null</b>.
171     *
172     * @param uri the URI to be converted
173     * @return the resulting URL or <b>null</b>
174     */
175    static URL convertURIToURL(final URI uri) {
176        try {
177            return uri.toURL();
178        } catch (final MalformedURLException e) {
179            return null;
180        }
181    }
182
183    /**
184     * Creates a fully initialized {@code FileLocator} based on the specified URL.
185     *
186     * @param src the source {@code FileLocator}
187     * @param url the URL
188     * @return the fully initialized {@code FileLocator}
189     */
190    private static FileLocator createFullyInitializedLocatorFromURL(final FileLocator src, final URL url) {
191        final FileLocator.FileLocatorBuilder fileLocatorBuilder = fileLocator(src);
192        if (src.getSourceURL() == null) {
193            fileLocatorBuilder.sourceURL(url);
194        }
195        if (StringUtils.isBlank(src.getFileName())) {
196            fileLocatorBuilder.fileName(getFileName(url));
197        }
198        if (StringUtils.isBlank(src.getBasePath())) {
199            fileLocatorBuilder.basePath(getBasePath(url));
200        }
201        return fileLocatorBuilder.create();
202    }
203
204    /**
205     * Tries to convert the specified URL to a file object. If this fails, <b>null</b> is returned.
206     *
207     * @param url the URL
208     * @return the resulting file object
209     */
210    public static File fileFromURL(final URL url) {
211        return FileUtils.toFile(url);
212    }
213
214    /**
215     * Returns an uninitialized {@code FileLocatorBuilder} which can be used for the creation of a {@code FileLocator}
216     * object. This method provides a convenient way to create file locators using a fluent API as in the following example:
217     *
218     * <pre>
219     * FileLocator locator = FileLocatorUtils.fileLocator().basePath(myBasePath).fileName("test.xml").create();
220     * </pre>
221     *
222     * @return a builder object for defining a {@code FileLocator}
223     */
224    public static FileLocator.FileLocatorBuilder fileLocator() {
225        return fileLocator(null);
226    }
227
228    /**
229     * Returns a {@code FileLocatorBuilder} which is already initialized with the properties of the passed in
230     * {@code FileLocator}. This builder can be used to create a {@code FileLocator} object which shares properties of the
231     * original locator (e.g. the {@code FileSystem} or the encoding), but points to a different file. An example use case
232     * is as follows:
233     *
234     * <pre>
235     * FileLocator loc1 = ...
236     * FileLocator loc2 = FileLocatorUtils.fileLocator(loc1)
237     *     .setFileName("anotherTest.xml")
238     *     .create();
239     * </pre>
240     *
241     * @param src the source {@code FileLocator} (may be <b>null</b>)
242     * @return an initialized builder object for defining a {@code FileLocator}
243     */
244    public static FileLocator.FileLocatorBuilder fileLocator(final FileLocator src) {
245        return new FileLocator.FileLocatorBuilder(src);
246    }
247
248    /**
249     * Creates a new {@code FileLocator} object with the properties defined in the given map. The map must be conform to the
250     * structure generated by the {@link #put(FileLocator, Map)} method; unexpected data can cause
251     * {@code ClassCastException} exceptions. The map can be <b>null</b>, then an uninitialized {@code FileLocator} is
252     * returned.
253     *
254     * @param map the map
255     * @return the new {@code FileLocator}
256     * @throws ClassCastException if the map contains invalid data
257     */
258    public static FileLocator fromMap(final Map<String, ?> map) {
259        final FileLocator.FileLocatorBuilder builder = fileLocator();
260        if (map != null) {
261            builder.basePath((String) map.get(PROP_BASE_PATH)).encoding((String) map.get(PROP_ENCODING)).fileName((String) map.get(PROP_FILE_NAME))
262                .fileSystem((FileSystem) map.get(PROP_FILE_SYSTEM)).locationStrategy((FileLocationStrategy) map.get(PROP_STRATEGY))
263                .sourceURL((URL) map.get(PROP_SOURCE_URL));
264        }
265        return builder.create();
266    }
267
268    /**
269     * Returns a {@code FileLocator} object based on the passed in one whose location is fully defined. This method ensures
270     * that all components of the {@code FileLocator} pointing to the file are set in a consistent way. In detail it behaves
271     * as follows:
272     * <ul>
273     * <li>If the {@code FileLocator} has already all components set which define the file, it is returned unchanged.
274     * <em>Note:</em> It is not checked whether all components are really consistent!</li>
275     * <li>{@link #locate(FileLocator)} is called to determine a unique URL pointing to the referenced file. If this is
276     * successful, a new {@code FileLocator} is created as a copy of the passed in one, but with all components pointing to
277     * the file derived from this URL.</li>
278     * <li>Otherwise, result is <b>null</b>.</li>
279     * </ul>
280     *
281     * @param locator the {@code FileLocator} to be completed
282     * @return a {@code FileLocator} with a fully initialized location if possible or <b>null</b>
283     */
284    public static FileLocator fullyInitializedLocator(final FileLocator locator) {
285        if (isFullyInitialized(locator)) {
286            // already fully initialized
287            return locator;
288        }
289
290        final URL url = locate(locator);
291        return url != null ? createFullyInitializedLocatorFromURL(locator, url) : null;
292    }
293
294    /**
295     * Gets the path without the file name, for example https://xyz.net/foo/bar.xml results in https://xyz.net/foo/
296     *
297     * @param url the URL from which to extract the path
298     * @return the path component of the passed in URL
299     */
300    static String getBasePath(final URL url) {
301        if (url == null) {
302            return null;
303        }
304
305        String s = url.toString();
306        if (s.startsWith(FILE_SCHEME) && !s.startsWith("file://")) {
307            s = "file://" + s.substring(FILE_SCHEME.length());
308        }
309
310        if (s.endsWith("/") || StringUtils.isEmpty(url.getPath())) {
311            return s;
312        }
313        return s.substring(0, s.lastIndexOf("/") + 1);
314    }
315
316    /**
317     * Tries to find a resource with the given name in the classpath.
318     *
319     * @param resourceName the name of the resource
320     * @return the URL to the found resource or <b>null</b> if the resource cannot be found
321     */
322    static URL getClasspathResource(final String resourceName) {
323        URL url = null;
324        // attempt to load from the context classpath
325        final ClassLoader loader = Thread.currentThread().getContextClassLoader();
326        if (loader != null) {
327            url = loader.getResource(resourceName);
328
329            if (url != null) {
330                LOG.debug("Loading configuration from the context classpath (" + resourceName + ")");
331            }
332        }
333
334        // attempt to load from the system classpath
335        if (url == null) {
336            url = ClassLoader.getSystemResource(resourceName);
337
338            if (url != null) {
339                LOG.debug("Loading configuration from the system classpath (" + resourceName + ")");
340            }
341        }
342        return url;
343    }
344
345    /**
346     * Tries to convert the specified base path and file name into a file object. This method is called e.g. by the save()
347     * methods of file based configurations. The parameter strings can be relative files, absolute files and URLs as well.
348     * This implementation checks first whether the passed in file name is absolute. If this is the case, it is returned.
349     * Otherwise further checks are performed whether the base path and file name can be combined to a valid URL or a valid
350     * file name. <em>Note:</em> The test if the passed in file name is absolute is performed using
351     * {@code java.io.File.isAbsolute()}. If the file name starts with a slash, this method will return <b>true</b> on Unix,
352     * but <b>false</b> on Windows. So to ensure correct behavior for relative file names on all platforms you should never
353     * let relative paths start with a slash. E.g. in a configuration definition file do not use something like that:
354     *
355     * <pre>
356     * &lt;properties fileName="/subdir/my.properties"/&gt;
357     * </pre>
358     *
359     * Under Windows this path would be resolved relative to the configuration definition file. Under Unix this would be
360     * treated as an absolute path name.
361     *
362     * @param basePath the base path
363     * @param fileName the file name (must not be <b>null</b>)
364     * @return the file object (<b>null</b> if no file can be obtained)
365     */
366    static File getFile(final String basePath, final String fileName) {
367        // Check if the file name is absolute
368        final File f = new File(fileName);
369        if (f.isAbsolute()) {
370            return f;
371        }
372
373        // Check if URLs are involved
374        URL url;
375        try {
376            url = new URL(new URL(basePath), fileName);
377        } catch (final MalformedURLException mex1) {
378            try {
379                url = new URL(fileName);
380            } catch (final MalformedURLException mex2) {
381                url = null;
382            }
383        }
384
385        if (url != null) {
386            return fileFromURL(url);
387        }
388
389        return constructFile(basePath, fileName);
390    }
391
392    /**
393     * Extract the file name from the specified URL.
394     *
395     * @param url the URL from which to extract the file name
396     * @return the extracted file name
397     */
398    static String getFileName(final URL url) {
399        if (url == null) {
400            return null;
401        }
402
403        final String path = url.getPath();
404
405        if (path.endsWith("/") || StringUtils.isEmpty(path)) {
406            return null;
407        }
408        return path.substring(path.lastIndexOf("/") + 1);
409    }
410
411    /**
412     * Obtains a non-<b>null</b> {@code FileSystem} object from the passed in {@code FileLocator}. If the passed in
413     * {@code FileLocator} has a {@code FileSystem} object, it is returned. Otherwise, result is the default
414     * {@code FileSystem}.
415     *
416     * @param locator the {@code FileLocator} (may be <b>null</b>)
417     * @return the {@code FileSystem} to be used for this {@code FileLocator}
418     */
419    static FileSystem getFileSystem(final FileLocator locator) {
420        return locator != null ? ObjectUtils.defaultIfNull(locator.getFileSystem(), DEFAULT_FILE_SYSTEM) : DEFAULT_FILE_SYSTEM;
421    }
422
423    /**
424     * Gets a non <b>null</b> {@code FileLocationStrategy} object from the passed in {@code FileLocator}. If the
425     * {@code FileLocator} is not <b>null</b> and has a {@code FileLocationStrategy} defined, this strategy is returned.
426     * Otherwise, result is the default {@code FileLocationStrategy}.
427     *
428     * @param locator the {@code FileLocator}
429     * @return the {@code FileLocationStrategy} for this {@code FileLocator}
430     */
431    static FileLocationStrategy getLocationStrategy(final FileLocator locator) {
432        return locator != null ? ObjectUtils.defaultIfNull(locator.getLocationStrategy(), DEFAULT_LOCATION_STRATEGY) : DEFAULT_LOCATION_STRATEGY;
433    }
434
435    /**
436     * Creates the default location strategy. This method creates a combined location strategy as described in the comment
437     * of the {@link #DEFAULT_LOCATION_STRATEGY} member field.
438     *
439     * @return the default {@code FileLocationStrategy}
440     */
441    private static FileLocationStrategy initDefaultLocationStrategy() {
442        final FileLocationStrategy[] subStrategies = {new ProvidedURLLocationStrategy(), new FileSystemLocationStrategy(), new AbsoluteNameLocationStrategy(),
443            new BasePathLocationStrategy(), new HomeDirectoryLocationStrategy(true), new HomeDirectoryLocationStrategy(false), new ClasspathLocationStrategy()};
444        return new CombinedLocationStrategy(Arrays.asList(subStrategies));
445    }
446
447    /**
448     * Returns a flag whether all components of the given {@code FileLocator} describing the referenced file are defined. In
449     * order to reference a file, it is not necessary that all components are filled in (for instance, the URL alone is
450     * sufficient). For some use cases however, it might be of interest to have different methods for accessing the
451     * referenced file. Also, depending on the filled out properties, there is a subtle difference how the file is accessed:
452     * If only the file name is set (and optionally the base path), each time the file is accessed a {@code locate()}
453     * operation has to be performed to uniquely identify the file. If however the URL is determined once based on the other
454     * components and stored in a fully defined {@code FileLocator}, it can be used directly to identify the file. If the
455     * passed in {@code FileLocator} is <b>null</b>, result is <b>false</b>.
456     *
457     * @param locator the {@code FileLocator} to be checked (may be <b>null</b>)
458     * @return a flag whether all components describing the referenced file are initialized
459     */
460    public static boolean isFullyInitialized(final FileLocator locator) {
461        if (locator == null) {
462            return false;
463        }
464        return locator.getBasePath() != null && locator.getFileName() != null && locator.getSourceURL() != null;
465    }
466
467    /**
468     * Checks whether the specified {@code FileLocator} contains enough information to locate a file. This is the case if a
469     * file name or a URL is defined. If the passed in {@code FileLocator} is <b>null</b>, result is <b>false</b>.
470     *
471     * @param locator the {@code FileLocator} to check
472     * @return a flag whether a file location is defined by this {@code FileLocator}
473     */
474    public static boolean isLocationDefined(final FileLocator locator) {
475        return locator != null && (locator.getFileName() != null || locator.getSourceURL() != null);
476    }
477
478    /**
479     * Locates the provided {@code FileLocator}, returning a URL for accessing the referenced file. This method uses a
480     * {@link FileLocationStrategy} to locate the file the passed in {@code FileLocator} points to. If the
481     * {@code FileLocator} contains itself a {@code FileLocationStrategy}, it is used. Otherwise, the default
482     * {@code FileLocationStrategy} is applied. The strategy is passed the locator and a {@code FileSystem}. The resulting
483     * URL is returned. If the {@code FileLocator} is <b>null</b>, result is <b>null</b>.
484     *
485     * @param locator the {@code FileLocator} to be resolved
486     * @return the URL pointing to the referenced file or <b>null</b> if the {@code FileLocator} could not be resolved
487     * @see #DEFAULT_LOCATION_STRATEGY
488     */
489    public static URL locate(final FileLocator locator) {
490        if (locator == null) {
491            return null;
492        }
493
494        return getLocationStrategy(locator).locate(getFileSystem(locator), locator);
495    }
496
497    /**
498     * Tries to locate the file referenced by the passed in {@code FileLocator}. If this fails, an exception is thrown. This
499     * method works like {@link #locate(FileLocator)}; however, in case of a failed location attempt an exception is thrown.
500     *
501     * @param locator the {@code FileLocator} to be resolved
502     * @return the URL pointing to the referenced file
503     * @throws ConfigurationException if the file cannot be resolved
504     */
505    public static URL locateOrThrow(final FileLocator locator) throws ConfigurationException {
506        final URL url = locate(locator);
507        if (url == null) {
508            throw new ConfigurationException("Could not locate: " + locator);
509        }
510        return url;
511    }
512
513    /**
514     * Stores the specified {@code FileLocator} in the given map. With the {@link #fromMap(Map)} method a new
515     * {@code FileLocator} with the same properties as the original one can be created.
516     *
517     * @param locator the {@code FileLocator} to be stored
518     * @param map the map in which to store the {@code FileLocator} (must not be <b>null</b>)
519     * @throws IllegalArgumentException if the map is <b>null</b>
520     */
521    public static void put(final FileLocator locator, final Map<String, Object> map) {
522        if (map == null) {
523            throw new IllegalArgumentException("Map must not be null!");
524        }
525
526        if (locator != null) {
527            map.put(PROP_BASE_PATH, locator.getBasePath());
528            map.put(PROP_ENCODING, locator.getEncoding());
529            map.put(PROP_FILE_NAME, locator.getFileName());
530            map.put(PROP_FILE_SYSTEM, locator.getFileSystem());
531            map.put(PROP_SOURCE_URL, locator.getSourceURL());
532            map.put(PROP_STRATEGY, locator.getLocationStrategy());
533        }
534    }
535
536    /**
537     * Convert the specified file into an URL. This method is equivalent to file.toURI().toURL(). It was used to work around
538     * a bug in the JDK preventing the transformation of a file into an URL if the file name contains a '#' character. See
539     * the issue CONFIGURATION-300 for more details. Now that we switched to JDK 1.4 we can directly use
540     * file.toURI().toURL().
541     *
542     * @param file the file to be converted into an URL
543     * @return a URL
544     * @throws MalformedURLException If the file protocol handler is not found (should not happen) or if an error occurred
545     *         while constructing the URL
546     */
547    static URL toURL(final File file) throws MalformedURLException {
548        return file.toURI().toURL();
549    }
550
551}