001// License: GPL. For details, see LICENSE file.
002package org.openstreetmap.josm.io;
003
004import static org.openstreetmap.josm.tools.I18n.tr;
005
006import java.io.BufferedInputStream;
007import java.io.BufferedReader;
008import java.io.ByteArrayOutputStream;
009import java.io.Closeable;
010import java.io.File;
011import java.io.FileInputStream;
012import java.io.IOException;
013import java.io.InputStream;
014import java.net.HttpURLConnection;
015import java.net.MalformedURLException;
016import java.net.URL;
017import java.nio.charset.StandardCharsets;
018import java.nio.file.Files;
019import java.nio.file.StandardCopyOption;
020import java.util.ArrayList;
021import java.util.Arrays;
022import java.util.Enumeration;
023import java.util.List;
024import java.util.Map;
025import java.util.concurrent.ConcurrentHashMap;
026import java.util.concurrent.TimeUnit;
027import java.util.zip.ZipEntry;
028import java.util.zip.ZipFile;
029
030import org.openstreetmap.josm.Main;
031import org.openstreetmap.josm.tools.HttpClient;
032import org.openstreetmap.josm.tools.Pair;
033import org.openstreetmap.josm.tools.Utils;
034
035/**
036 * Downloads a file and caches it on disk in order to reduce network load.
037 *
038 * Supports URLs, local files, and a custom scheme (<code>resource:</code>) to get
039 * resources from the current *.jar file. (Local caching is only done for URLs.)
040 * <p>
041 * The mirrored file is only downloaded if it has been more than 7 days since
042 * last download. (Time can be configured.)
043 * <p>
044 * The file content is normally accessed with {@link #getInputStream()}, but
045 * you can also get the mirrored copy with {@link #getFile()}.
046 */
047public class CachedFile implements Closeable {
048
049    /**
050     * Caching strategy.
051     */
052    public enum CachingStrategy {
053        /**
054         * If cached file on disk is older than a certain time (7 days by default),
055         * consider the cache stale and try to download the file again.
056         */
057        MaxAge,
058        /**
059         * Similar to MaxAge, considers the cache stale when a certain age is
060         * exceeded. In addition, a If-Modified-Since HTTP header is added.
061         * When the server replies "304 Not Modified", this is considered the same
062         * as a full download.
063         */
064        IfModifiedSince
065    }
066
067    protected String name;
068    protected long maxAge;
069    protected String destDir;
070    protected String httpAccept;
071    protected CachingStrategy cachingStrategy;
072
073    private boolean fastFail;
074    private HttpClient activeConnection;
075    protected File cacheFile;
076    protected boolean initialized;
077
078    public static final long DEFAULT_MAXTIME = -1L;
079    public static final long DAYS = TimeUnit.DAYS.toSeconds(1); // factor to get caching time in days
080
081    private final Map<String, String> httpHeaders = new ConcurrentHashMap<>();
082
083    /**
084     * Constructs a CachedFile object from a given filename, URL or internal resource.
085     *
086     * @param name can be:<ul>
087     *  <li>relative or absolute file name</li>
088     *  <li>{@code file:///SOME/FILE} the same as above</li>
089     *  <li>{@code http://...} a URL. It will be cached on disk.</li>
090     *  <li>{@code resource://SOME/FILE} file from the classpath (usually in the current *.jar)</li>
091     *  <li>{@code josmdir://SOME/FILE} file inside josm user data directory (since r7058)</li>
092     *  <li>{@code josmplugindir://SOME/FILE} file inside josm plugin directory (since r7834)</li></ul>
093     */
094    public CachedFile(String name) {
095        this.name = name;
096    }
097
098    /**
099     * Set the name of the resource.
100     * @param name can be:<ul>
101     *  <li>relative or absolute file name</li>
102     *  <li>{@code file:///SOME/FILE} the same as above</li>
103     *  <li>{@code http://...} a URL. It will be cached on disk.</li>
104     *  <li>{@code resource://SOME/FILE} file from the classpath (usually in the current *.jar)</li>
105     *  <li>{@code josmdir://SOME/FILE} file inside josm user data directory (since r7058)</li>
106     *  <li>{@code josmplugindir://SOME/FILE} file inside josm plugin directory (since r7834)</li></ul>
107     * @return this object
108     */
109    public CachedFile setName(String name) {
110        this.name = name;
111        return this;
112    }
113
114    /**
115     * Set maximum age of cache file. Only applies to URLs.
116     * When this time has passed after the last download of the file, the
117     * cache is considered stale and a new download will be attempted.
118     * @param maxAge the maximum cache age in seconds
119     * @return this object
120     */
121    public CachedFile setMaxAge(long maxAge) {
122        this.maxAge = maxAge;
123        return this;
124    }
125
126    /**
127     * Set the destination directory for the cache file. Only applies to URLs.
128     * @param destDir the destination directory
129     * @return this object
130     */
131    public CachedFile setDestDir(String destDir) {
132        this.destDir = destDir;
133        return this;
134    }
135
136    /**
137     * Set the accepted MIME types sent in the HTTP Accept header. Only applies to URLs.
138     * @param httpAccept the accepted MIME types
139     * @return this object
140     */
141    public CachedFile setHttpAccept(String httpAccept) {
142        this.httpAccept = httpAccept;
143        return this;
144    }
145
146    /**
147     * Set the caching strategy. Only applies to URLs.
148     * @param cachingStrategy caching strategy
149     * @return this object
150     */
151    public CachedFile setCachingStrategy(CachingStrategy cachingStrategy) {
152        this.cachingStrategy = cachingStrategy;
153        return this;
154    }
155
156    /**
157     * Sets the http headers. Only applies to URL pointing to http or https resources
158     * @param headers that should be sent together with request
159     * @return this object
160     */
161    public CachedFile setHttpHeaders(Map<String, String> headers) {
162        this.httpHeaders.putAll(headers);
163        return this;
164    }
165
166    /**
167     * Sets whether opening HTTP connections should fail fast, i.e., whether a
168     * {@link HttpClient#setConnectTimeout(int) low connect timeout} should be used.
169     * @param fastFail whether opening HTTP connections should fail fast
170     */
171    public void setFastFail(boolean fastFail) {
172        this.fastFail = fastFail;
173    }
174
175    public String getName() {
176        return name;
177    }
178
179    /**
180     * Returns maximum age of cache file. Only applies to URLs.
181     * When this time has passed after the last download of the file, the
182     * cache is considered stale and a new download will be attempted.
183     * @return the maximum cache age in seconds
184     */
185    public long getMaxAge() {
186        return maxAge;
187    }
188
189    public String getDestDir() {
190        return destDir;
191    }
192
193    public String getHttpAccept() {
194        return httpAccept;
195    }
196
197    public CachingStrategy getCachingStrategy() {
198        return cachingStrategy;
199    }
200
201    /**
202     * Get InputStream to the requested resource.
203     * @return the InputStream
204     * @throws IOException when the resource with the given name could not be retrieved
205     */
206    public InputStream getInputStream() throws IOException {
207        File file = getFile();
208        if (file == null) {
209            if (name.startsWith("resource://")) {
210                InputStream is = getClass().getResourceAsStream(
211                        name.substring("resource:/".length()));
212                if (is == null)
213                    throw new IOException(tr("Failed to open input stream for resource ''{0}''", name));
214                return is;
215            } else {
216                throw new IOException("No file found for: "+name);
217            }
218        }
219        return new FileInputStream(file);
220    }
221
222    /**
223     * Get the full content of the requested resource as a byte array.
224     * @return the full content of the requested resource as byte array
225     * @throws IOException in case of an I/O error
226     */
227    public byte[] getByteContent() throws IOException {
228        try (InputStream is = getInputStream()) {
229            ByteArrayOutputStream buffer = new ByteArrayOutputStream();
230            int nRead;
231            byte[] data = new byte[8192];
232            while ((nRead = is.read(data, 0, data.length)) != -1) {
233                buffer.write(data, 0, nRead);
234            }
235            buffer.flush();
236            return buffer.toByteArray();
237        }
238    }
239
240    /**
241     * Returns {@link #getInputStream()} wrapped in a buffered reader.
242     * <p>
243     * Detects Unicode charset in use utilizing {@link UTFInputStreamReader}.
244     *
245     * @return buffered reader
246     * @throws IOException if any I/O error occurs
247     * @since 9411
248     */
249    public BufferedReader getContentReader() throws IOException {
250        return new BufferedReader(UTFInputStreamReader.create(getInputStream()));
251    }
252
253    /**
254     * Get local file for the requested resource.
255     * @return The local cache file for URLs. If the resource is a local file,
256     * returns just that file.
257     * @throws IOException when the resource with the given name could not be retrieved
258     */
259    public synchronized File getFile() throws IOException {
260        if (initialized)
261            return cacheFile;
262        initialized = true;
263        URL url;
264        try {
265            url = new URL(name);
266            if ("file".equals(url.getProtocol())) {
267                cacheFile = new File(name.substring("file:/".length() - 1));
268                if (!cacheFile.exists()) {
269                    cacheFile = new File(name.substring("file://".length() - 1));
270                }
271            } else {
272                cacheFile = checkLocal(url);
273            }
274        } catch (MalformedURLException e) {
275            if (name.startsWith("resource://")) {
276                return null;
277            } else if (name.startsWith("josmdir://")) {
278                cacheFile = new File(Main.pref.getUserDataDirectory(), name.substring("josmdir://".length()));
279            } else if (name.startsWith("josmplugindir://")) {
280                cacheFile = new File(Main.pref.getPluginsDirectory(), name.substring("josmplugindir://".length()));
281            } else {
282                cacheFile = new File(name);
283            }
284        }
285        if (cacheFile == null)
286            throw new IOException("Unable to get cache file for "+name);
287        return cacheFile;
288    }
289
290    /**
291     * Looks for a certain entry inside a zip file and returns the entry path.
292     *
293     * Replies a file in the top level directory of the ZIP file which has an
294     * extension <code>extension</code>. If more than one files have this
295     * extension, the last file whose name includes <code>namepart</code>
296     * is opened.
297     *
298     * @param extension  the extension of the file we're looking for
299     * @param namepart the name part
300     * @return The zip entry path of the matching file. Null if this cached file
301     * doesn't represent a zip file or if there was no matching
302     * file in the ZIP file.
303     */
304    public String findZipEntryPath(String extension, String namepart) {
305        Pair<String, InputStream> ze = findZipEntryImpl(extension, namepart);
306        if (ze == null) return null;
307        return ze.a;
308    }
309
310    /**
311     * Like {@link #findZipEntryPath}, but returns the corresponding InputStream.
312     * @param extension  the extension of the file we're looking for
313     * @param namepart the name part
314     * @return InputStream to the matching file. Null if this cached file
315     * doesn't represent a zip file or if there was no matching
316     * file in the ZIP file.
317     * @since 6148
318     */
319    public InputStream findZipEntryInputStream(String extension, String namepart) {
320        Pair<String, InputStream> ze = findZipEntryImpl(extension, namepart);
321        if (ze == null) return null;
322        return ze.b;
323    }
324
325    private Pair<String, InputStream> findZipEntryImpl(String extension, String namepart) {
326        File file = null;
327        try {
328            file = getFile();
329        } catch (IOException ex) {
330            Main.warn(ex, false);
331        }
332        if (file == null)
333            return null;
334        Pair<String, InputStream> res = null;
335        try {
336            ZipFile zipFile = new ZipFile(file, StandardCharsets.UTF_8);
337            ZipEntry resentry = null;
338            Enumeration<? extends ZipEntry> entries = zipFile.entries();
339            while (entries.hasMoreElements()) {
340                ZipEntry entry = entries.nextElement();
341                // choose any file with correct extension. When more than one file, prefer the one which matches namepart
342                if (entry.getName().endsWith('.' + extension) && (resentry == null || entry.getName().indexOf(namepart) >= 0)) {
343                    resentry = entry;
344                }
345            }
346            if (resentry != null) {
347                InputStream is = zipFile.getInputStream(resentry);
348                res = Pair.create(resentry.getName(), is);
349            } else {
350                Utils.close(zipFile);
351            }
352        } catch (IOException e) {
353            if (file.getName().endsWith(".zip")) {
354                Main.warn(e, tr("Failed to open file with extension ''{2}'' and namepart ''{3}'' in zip file ''{0}''. Exception was: {1}",
355                        file.getName(), e.toString(), extension, namepart));
356            }
357        }
358        return res;
359    }
360
361    /**
362     * Clear the cache for the given resource.
363     * This forces a fresh download.
364     * @param name the URL
365     */
366    public static void cleanup(String name) {
367        cleanup(name, null);
368    }
369
370    /**
371     * Clear the cache for the given resource.
372     * This forces a fresh download.
373     * @param name the URL
374     * @param destDir the destination directory (see {@link #setDestDir(java.lang.String)})
375     */
376    public static void cleanup(String name, String destDir) {
377        URL url;
378        try {
379            url = new URL(name);
380            if (!"file".equals(url.getProtocol())) {
381                String prefKey = getPrefKey(url, destDir);
382                List<String> localPath = new ArrayList<>(Main.pref.getCollection(prefKey));
383                if (localPath.size() == 2) {
384                    File lfile = new File(localPath.get(1));
385                    if (lfile.exists()) {
386                        Utils.deleteFile(lfile);
387                    }
388                }
389                Main.pref.putCollection(prefKey, null);
390            }
391        } catch (MalformedURLException e) {
392            Main.warn(e);
393        }
394    }
395
396    /**
397     * Get preference key to store the location and age of the cached file.
398     * 2 resources that point to the same url, but that are to be stored in different
399     * directories will not share a cache file.
400     * @param url URL
401     * @param destDir destination directory
402     * @return Preference key
403     */
404    private static String getPrefKey(URL url, String destDir) {
405        StringBuilder prefKey = new StringBuilder("mirror.");
406        if (destDir != null) {
407            prefKey.append(destDir).append('.');
408        }
409        prefKey.append(url.toString());
410        return prefKey.toString().replaceAll("=", "_");
411    }
412
413    private File checkLocal(URL url) throws IOException {
414        String prefKey = getPrefKey(url, destDir);
415        String urlStr = url.toExternalForm();
416        long age = 0L;
417        long maxAgeMillis = maxAge;
418        Long ifModifiedSince = null;
419        File localFile = null;
420        List<String> localPathEntry = new ArrayList<>(Main.pref.getCollection(prefKey));
421        boolean offline = false;
422        try {
423            checkOfflineAccess(urlStr);
424        } catch (OfflineAccessException e) {
425            Main.trace(e);
426            offline = true;
427        }
428        if (localPathEntry.size() == 2) {
429            localFile = new File(localPathEntry.get(1));
430            if (!localFile.exists()) {
431                localFile = null;
432            } else {
433                if (maxAge == DEFAULT_MAXTIME
434                        || maxAge <= 0 // arbitrary value <= 0 is deprecated
435                ) {
436                    maxAgeMillis = TimeUnit.SECONDS.toMillis(Main.pref.getLong("mirror.maxtime", TimeUnit.DAYS.toSeconds(7)));
437                }
438                age = System.currentTimeMillis() - Long.parseLong(localPathEntry.get(0));
439                if (offline || age < maxAgeMillis) {
440                    return localFile;
441                }
442                if (cachingStrategy == CachingStrategy.IfModifiedSince) {
443                    ifModifiedSince = Long.valueOf(localPathEntry.get(0));
444                }
445            }
446        }
447        if (destDir == null) {
448            destDir = Main.pref.getCacheDirectory().getPath();
449        }
450
451        File destDirFile = new File(destDir);
452        if (!destDirFile.exists()) {
453            Utils.mkDirs(destDirFile);
454        }
455
456        // No local file + offline => nothing to do
457        if (offline) {
458            return null;
459        }
460
461        String a = urlStr.replaceAll("[^A-Za-z0-9_.-]", "_");
462        String localPath = "mirror_" + a;
463        destDirFile = new File(destDir, localPath + ".tmp");
464        try {
465            activeConnection = HttpClient.create(url)
466                    .setAccept(httpAccept)
467                    .setIfModifiedSince(ifModifiedSince == null ? 0L : ifModifiedSince)
468                    .setHeaders(httpHeaders);
469            if (fastFail) {
470                activeConnection.setReadTimeout(1000);
471            }
472            final HttpClient.Response con = activeConnection.connect();
473            if (ifModifiedSince != null && con.getResponseCode() == HttpURLConnection.HTTP_NOT_MODIFIED) {
474                if (Main.isDebugEnabled()) {
475                    Main.debug("304 Not Modified ("+urlStr+')');
476                }
477                if (localFile == null)
478                    throw new AssertionError();
479                Main.pref.putCollection(prefKey,
480                        Arrays.asList(Long.toString(System.currentTimeMillis()), localPathEntry.get(1)));
481                return localFile;
482            } else if (con.getResponseCode() == HttpURLConnection.HTTP_NOT_FOUND) {
483                throw new IOException(tr("The requested URL {0} was not found", urlStr));
484            }
485            try (InputStream bis = new BufferedInputStream(con.getContent())) {
486                Files.copy(bis, destDirFile.toPath(), StandardCopyOption.REPLACE_EXISTING);
487            }
488            activeConnection = null;
489            localFile = new File(destDir, localPath);
490            if (Main.platform.rename(destDirFile, localFile)) {
491                Main.pref.putCollection(prefKey,
492                        Arrays.asList(Long.toString(System.currentTimeMillis()), localFile.toString()));
493            } else {
494                Main.warn(tr("Failed to rename file {0} to {1}.",
495                destDirFile.getPath(), localFile.getPath()));
496            }
497        } catch (IOException e) {
498            if (age >= maxAgeMillis && age < maxAgeMillis*2) {
499                Main.warn(tr("Failed to load {0}, use cached file and retry next time: {1}", urlStr, e));
500                return localFile;
501            } else {
502                throw e;
503            }
504        }
505
506        return localFile;
507    }
508
509    private static void checkOfflineAccess(String urlString) {
510        OnlineResource.JOSM_WEBSITE.checkOfflineAccess(urlString, Main.getJOSMWebsite());
511        OnlineResource.OSM_API.checkOfflineAccess(urlString, OsmApi.getOsmApi().getServerUrl());
512    }
513
514    /**
515     * Attempts to disconnect an URL connection.
516     * @see HttpClient#disconnect()
517     * @since 9411
518     */
519    @Override
520    public void close() {
521        if (activeConnection != null) {
522            activeConnection.disconnect();
523        }
524    }
525
526    /**
527     * Clears the cached file
528     * @throws IOException if any I/O error occurs
529     * @since 10993
530     */
531    public void clear() throws IOException {
532        URL url;
533        try {
534            url = new URL(name);
535            if ("file".equals(url.getProtocol())) {
536                return; // this is local file - do not delete it
537            }
538        } catch (MalformedURLException e) {
539            return; // if it's not a URL, then it still might be a local file - better not to delete
540        }
541        File f = getFile();
542        if (f != null && f.exists()) {
543            Utils.deleteFile(f);
544        }
545    }
546}