2 * Sone - FreenetLinkParser.java - Copyright © 2010 David Roden
4 * This program is free software: you can redistribute it and/or modify
5 * it under the terms of the GNU General Public License as published by
6 * the Free Software Foundation, either version 3 of the License, or
7 * (at your option) any later version.
9 * This program is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 * GNU General Public License for more details.
14 * You should have received a copy of the GNU General Public License
15 * along with this program. If not, see <http://www.gnu.org/licenses/>.
18 package net.pterodactylus.sone.text;
20 import java.io.BufferedReader;
21 import java.io.IOException;
22 import java.io.Reader;
23 import java.net.MalformedURLException;
24 import java.util.logging.Level;
25 import java.util.logging.Logger;
26 import java.util.regex.Matcher;
27 import java.util.regex.Pattern;
29 import net.pterodactylus.sone.core.Core;
30 import net.pterodactylus.sone.data.Post;
31 import net.pterodactylus.sone.data.Sone;
32 import net.pterodactylus.util.logging.Logging;
33 import freenet.keys.FreenetURI;
36 * {@link Parser} implementation that can recognize Freenet URIs.
38 * @author <a href="mailto:bombe@pterodactylus.net">David ‘Bombe’ Roden</a>
40 public class SoneTextParser implements Parser<SoneTextParserContext> {
43 private static final Logger logger = Logging.getLogger(SoneTextParser.class);
45 /** Pattern to detect whitespace. */
46 private static final Pattern whitespacePattern = Pattern.compile("[\\u000a\u0020\u00a0\u1680\u180e\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a\u200b\u200c\u200d\u202f\u205f\u2060\u2800\u3000]");
49 * Enumeration for all recognized link types.
51 * @author <a href="mailto:bombe@pterodactylus.net">David ‘Bombe’ Roden</a>
53 private enum LinkType {
61 /** Link is an SSK. */
73 /** Link is a Sone. */
76 /** Link is a post. */
82 private final Core core;
85 * Creates a new freenet link parser.
90 public SoneTextParser(Core core) {
102 public Iterable<Part> parse(SoneTextParserContext context, Reader source) throws IOException {
103 PartContainer parts = new PartContainer();
104 BufferedReader bufferedReader = (source instanceof BufferedReader) ? (BufferedReader) source : new BufferedReader(source);
106 boolean lastLineEmpty = true;
108 while ((line = bufferedReader.readLine()) != null) {
109 if (line.trim().length() == 0) {
113 parts.add(new PlainTextPart("\n"));
115 lastLineEmpty = emptyLines == 2;
119 boolean lineComplete = true;
120 while (line.length() > 0) {
121 int nextKsk = line.indexOf("KSK@");
122 int nextChk = line.indexOf("CHK@");
123 int nextSsk = line.indexOf("SSK@");
124 int nextUsk = line.indexOf("USK@");
125 int nextHttp = line.indexOf("http://");
126 int nextHttps = line.indexOf("https://");
127 int nextSone = line.indexOf("sone://");
128 int nextPost = line.indexOf("post://");
129 if ((nextKsk == -1) && (nextChk == -1) && (nextSsk == -1) && (nextUsk == -1) && (nextHttp == -1) && (nextHttps == -1) && (nextSone == -1) && (nextPost == -1)) {
130 if (lineComplete && !lastLineEmpty) {
131 parts.add(new PlainTextPart("\n" + line));
133 parts.add(new PlainTextPart(line));
137 int next = Integer.MAX_VALUE;
138 LinkType linkType = null;
139 if ((nextKsk > -1) && (nextKsk < next)) {
141 linkType = LinkType.KSK;
143 if ((nextChk > -1) && (nextChk < next)) {
145 linkType = LinkType.CHK;
147 if ((nextSsk > -1) && (nextSsk < next)) {
149 linkType = LinkType.SSK;
151 if ((nextUsk > -1) && (nextUsk < next)) {
153 linkType = LinkType.USK;
155 if ((nextHttp > -1) && (nextHttp < next)) {
157 linkType = LinkType.HTTP;
159 if ((nextHttps > -1) && (nextHttps < next)) {
161 linkType = LinkType.HTTPS;
163 if ((nextSone > -1) && (nextSone < next)) {
165 linkType = LinkType.SONE;
167 if ((nextPost > -1) && (nextPost < next)) {
169 linkType = LinkType.POST;
171 if (linkType == LinkType.SONE) {
173 parts.add(new PlainTextPart(line.substring(0, next)));
175 if (line.length() >= (next + 7 + 43)) {
176 String soneId = line.substring(next + 7, next + 50);
177 Sone sone = core.getSone(soneId, false);
179 parts.add(new SonePart(sone));
181 parts.add(new PlainTextPart(line.substring(next, next + 50)));
183 line = line.substring(next + 50);
185 parts.add(new PlainTextPart(line.substring(next)));
190 if (linkType == LinkType.POST) {
192 parts.add(new PlainTextPart(line.substring(0, next)));
194 if (line.length() >= (next + 7 + 36)) {
195 String postId = line.substring(next + 7, next + 43);
196 Post post = core.getPost(postId, false);
197 if ((post != null) && (post.getSone() != null)) {
198 String postText = post.getText();
199 postText = postText.substring(0, Math.min(postText.length(), 20)) + "…";
200 parts.add(new PostPart(post));
202 parts.add(new PlainTextPart(line.substring(next, next + 43)));
204 line = line.substring(next + 43);
206 parts.add(new PlainTextPart(line.substring(next)));
211 if ((next >= 8) && (line.substring(next - 8, next).equals("freenet:"))) {
213 line = line.substring(0, next) + line.substring(next + 8);
215 Matcher matcher = whitespacePattern.matcher(line);
216 int nextSpace = matcher.find(next) ? matcher.start() : line.length();
217 if (nextSpace > (next + 4)) {
218 if (!lastLineEmpty && lineComplete) {
219 parts.add(new PlainTextPart("\n" + line.substring(0, next)));
221 parts.add(new PlainTextPart(line.substring(0, next)));
223 String link = line.substring(next, nextSpace);
225 logger.log(Level.FINER, "Found link: %s", link);
226 logger.log(Level.FINEST, "Next: %d, CHK: %d, SSK: %d, USK: %d", new Object[] { next, nextChk, nextSsk, nextUsk });
228 if ((linkType == LinkType.KSK) || (linkType == LinkType.CHK) || (linkType == LinkType.SSK) || (linkType == LinkType.USK)) {
230 if (name.indexOf('?') > -1) {
231 name = name.substring(0, name.indexOf('?'));
233 if (name.endsWith("/")) {
234 name = name.substring(0, name.length() - 1);
237 uri = new FreenetURI(name);
238 name = uri.lastMetaString();
240 name = uri.getDocName();
243 name = link.substring(0, Math.min(9, link.length()));
245 boolean fromPostingSone = ((linkType == LinkType.SSK) || (linkType == LinkType.USK)) && (context.getPostingSone() != null) && link.substring(4, Math.min(link.length(), 47)).equals(context.getPostingSone().getId());
246 parts.add(new FreenetLinkPart(link, name, fromPostingSone));
247 } catch (MalformedURLException mue1) {
248 /* not a valid link, insert as plain text. */
249 parts.add(new PlainTextPart(link));
250 } catch (NullPointerException npe1) {
251 /* FreenetURI sometimes throws these, too. */
252 parts.add(new PlainTextPart(link));
253 } catch (ArrayIndexOutOfBoundsException aioobe1) {
254 /* oh, and these, too. */
255 parts.add(new PlainTextPart(link));
257 } else if ((linkType == LinkType.HTTP) || (linkType == LinkType.HTTPS)) {
258 name = link.substring(linkType == LinkType.HTTP ? 7 : 8);
259 int firstSlash = name.indexOf('/');
260 int lastSlash = name.lastIndexOf('/');
261 if ((lastSlash - firstSlash) > 3) {
262 name = name.substring(0, firstSlash + 1) + "…" + name.substring(lastSlash);
264 if (name.endsWith("/")) {
265 name = name.substring(0, name.length() - 1);
267 if (((name.indexOf('/') > -1) && (name.indexOf('.') < name.lastIndexOf('.', name.indexOf('/'))) || ((name.indexOf('/') == -1) && (name.indexOf('.') < name.lastIndexOf('.')))) && name.startsWith("www.")) {
268 name = name.substring(4);
270 if (name.indexOf('?') > -1) {
271 name = name.substring(0, name.indexOf('?'));
273 parts.add(new LinkPart(link, name));
275 line = line.substring(nextSpace);
277 if (!lastLineEmpty && lineComplete) {
278 parts.add(new PlainTextPart("\n" + line.substring(0, next + 4)));
280 parts.add(new PlainTextPart(line.substring(0, next + 4)));
282 line = line.substring(next + 4);
284 lineComplete = false;
286 lastLineEmpty = false;
288 for (int partIndex = parts.size() - 1; partIndex >= 0; --partIndex) {
289 Part part = parts.getPart(partIndex);
290 if (!(part instanceof PlainTextPart) || !"\n".equals(((PlainTextPart) part).getText())) {
293 parts.removePart(partIndex);