Drupal: ignorování XML komentářů (nebudou dělit odstavce).
3 * see AUTHORS for the list of contributors
5 * This program is free software: you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation, either version 3 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License
16 * along with this program. If not, see <http://www.gnu.org/licenses/>.
18 package org.sonews.storage;
20 import java.io.BufferedReader;
21 import java.io.ByteArrayInputStream;
22 import java.io.ByteArrayOutputStream;
23 import java.io.IOException;
24 import java.io.InputStream;
25 import java.io.InputStreamReader;
26 import java.io.PrintStream;
27 import java.io.StringReader;
28 import java.io.StringWriter;
29 import java.io.UnsupportedEncodingException;
30 import java.sql.ResultSet;
31 import java.sql.SQLException;
32 import java.util.ArrayList;
33 import java.util.Date;
34 import java.util.Enumeration;
35 import java.util.logging.Level;
36 import java.util.logging.Logger;
37 import javax.mail.Header;
38 import javax.mail.MessagingException;
39 import javax.mail.Multipart;
40 import javax.mail.Session;
41 import javax.mail.internet.InternetAddress;
42 import javax.mail.internet.MimeBodyPart;
43 import javax.mail.internet.MimeMessage;
44 import javax.mail.internet.MimeMultipart;
45 import javax.xml.parsers.DocumentBuilder;
46 import javax.xml.parsers.DocumentBuilderFactory;
47 import javax.xml.parsers.ParserConfigurationException;
48 import javax.xml.transform.Transformer;
49 import javax.xml.transform.TransformerException;
50 import javax.xml.transform.TransformerFactory;
51 import javax.xml.transform.dom.DOMSource;
52 import javax.xml.transform.stream.StreamResult;
53 import javax.xml.transform.stream.StreamSource;
54 import org.sonews.daemon.NNTPConnection;
55 import org.sonews.util.io.Resource;
56 import org.w3c.dom.Document;
57 import org.xml.sax.SAXException;
60 * This is MimeMessage which enables custom Message-ID header
61 * (this header will not be overwritten by the default one like in MimeMessage).
63 * Also add header and body separate serialization.
65 * And can be deserialized from SQL ResultSet
67 * @author František Kučera (frantovo.cz)
69 public class DrupalMessage extends MimeMessage {
71 private static final Logger log = Logger.getLogger(DrupalMessage.class.getName());
72 private static final String MESSAGE_ID_HEADER = "Message-ID";
73 private static final String CRLF = "\r\n";
74 public static final String CHARSET = "UTF-8";
75 private static final String XHTML_CONTENT_TYPE = "text/html; charset=" + CHARSET;
76 private static final String ZNAKČKA_KONCE_ŘÁDKU = "◆";
77 private String messageID;
78 private Long parentID;
80 private TransformerFactory transformerFactory;
81 private DocumentBuilderFactory documentBuilderFactory;
84 * Initializes XML factories (Transformer, DocumentBuilder).
86 private void initFactories() {
87 transformerFactory = TransformerFactory.newInstance();
88 documentBuilderFactory = DocumentBuilderFactory.newInstance();
90 * Komentáře nás nepotřebujeme
91 * (a museli bychom je brát v úvahu při dělení odstavců:
92 * v současné verzi XSLT odstavcovače by nám případný komentář
93 * rozdělil text na dva odstavce, přestože to má být odstavec jede).
95 documentBuilderFactory.setIgnoringComments(true);
99 * Constructs MIME message from SQL result.
100 * @param rs ResultSet containing message data. No {@link ResultSet#next()} will be called, just values from current row will be read.
101 * @param constructBody true if whole message should be constructed | false if we need only message headers (body will be dummy).
103 public DrupalMessage(ResultSet rs, String myDomain, boolean constructBody) throws SQLException, UnsupportedEncodingException, MessagingException, TransformerException, IOException, ParserConfigurationException, SAXException {
104 super(Session.getDefaultInstance(System.getProperties()));
107 groupID = rs.getLong("group_id");
108 addHeader("Message-id", constructMessageId(rs.getInt("id"), groupID, rs.getString("group_name"), myDomain));
109 addHeader("Newsgroups", rs.getString("group_name"));
110 setFrom(new InternetAddress(rs.getString("sender_email"), rs.getString("sender_name")));
111 setSubject(rs.getString("subject"));
112 setSentDate(new Date(rs.getLong("created")));
114 parentID = rs.getLong("parent_id");
116 String parentMessageID = constructMessageId(parentID, rs.getInt("group_id"), rs.getString("group_name"), myDomain);
117 addHeader("In-Reply-To", parentMessageID);
118 addHeader("References", parentMessageID);
122 Multipart multipart = new MimeMultipart("alternative");
123 setContent(multipart);
126 MimeBodyPart htmlPart = new MimeBodyPart();
127 String xhtmlText = readXhtmlText(
128 rs.getString("text"),
129 rs.getString("subject"),
130 rs.getInt("parent_id"),
131 rs.getString("urlBase"),
132 rs.getString("wwwRead"),
133 rs.getString("wwwPost"));
134 htmlPart.setContent(xhtmlText, XHTML_CONTENT_TYPE);
136 /** Plain text part */
137 MimeBodyPart textPart = new MimeBodyPart();
138 String plainText = readPlainText(rs, xhtmlText);
139 textPart.setText(plainText);
140 //addHeader("Lines", String.valueOf(plainText.split("\n").length));
143 * Thunderbirdu záleží, v jakém pořadí části jsou
144 * (když je prostý text druhý, html se nezobrazí),
145 * KNode zobrazuje HTML správně, i když je na prvním místě.
147 multipart.addBodyPart(textPart);
148 multipart.addBodyPart(htmlPart);
150 /** empty body, just headers */
156 * Constructs MIME message from article posted by user.
157 * @param article article that came through NNTP.
158 * @throws MessagingException
160 public DrupalMessage(Article article) throws MessagingException {
161 super(Session.getDefaultInstance(System.getProperties()), serializeArticle(article));
164 String[] parentHeaders = getHeader("In-Reply-To");
165 if (parentHeaders != null && parentHeaders.length == 1) {
166 String parentMessageID = parentHeaders[0];
167 parentID = parseArticleID(parentMessageID);
168 groupID = parseGroupID(parentMessageID);
170 throw new MessagingException("Message posted by user must have exactly one In-Reply-To header.");
174 private static InputStream serializeArticle(Article a) {
175 byte articleHeaders[] = a.getHeaderSource().getBytes();
176 byte delimiter[] = (NNTPConnection.NEWLINE + NNTPConnection.NEWLINE).getBytes();
177 byte body[] = a.getBody();
179 byte message[] = new byte[articleHeaders.length + delimiter.length + body.length];
181 System.arraycopy(articleHeaders, 0, message, 0, articleHeaders.length);
182 System.arraycopy(delimiter, 0, message, articleHeaders.length, delimiter.length);
183 System.arraycopy(body, 0, message, articleHeaders.length + delimiter.length, body.length);
185 return new ByteArrayInputStream(message);
188 private String readPlainText(ResultSet rs, String xhtmlText) {
190 Transformer textTransformer = transformerFactory.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeTextPart.xsl")));
192 StringReader input = new StringReader(xhtmlText);
193 StringWriter output = new StringWriter(xhtmlText.length());
194 textTransformer.transform(new StreamSource(input), new StreamResult(output));
196 return output.toString();
197 } catch (Exception e) {
199 * TODO: lepší ošetření chyby
201 log.log(Level.WARNING, "Error while transforming article to plain text", e);
202 return makeSimpleXHTML("Při transformaci příspěvku bohužel došlo k chybě.");
206 private DOMSource readDOM(String xml) throws ParserConfigurationException, SAXException, IOException {
207 DocumentBuilder db = documentBuilderFactory.newDocumentBuilder();
208 Document d = db.parse(new ByteArrayInputStream(xml.getBytes("UTF-8")));
209 return new DOMSource(d);
212 private String readXhtmlText(String text, String subject, long parentId, String urlBase, String wwwRead, String wwwPost) throws TransformerException, IOException, ParserConfigurationException, SAXException {
215 * - znovupoužívat XSL transformér (nejen v instanci)
216 * - používat cache, ukládat si vygenerované články
218 String inputText = makeSimpleXHTML(text);
220 Transformer paragraphTransformer = transformerFactory.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart-make-paragraphs.xsl")));
222 String paragraphedText;
223 boolean tidyWasUsed = false;
225 StringWriter output = new StringWriter(2 * inputText.length());
226 paragraphTransformer.transform(readDOM(inputText), new StreamResult(output));
227 paragraphedText = output.toString();
228 } catch (Exception e) {
229 log.log(Level.FINER, "HTML input was shitty – Tidy had to be called.", e);
230 StringWriter output = new StringWriter(2 * inputText.length());
231 paragraphTransformer.transform(readDOM(tidyXhtml(inputText)), new StreamResult(output));
232 paragraphedText = output.toString();
236 Transformer xhtmlTransformer = transformerFactory.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart.xsl")));
237 xhtmlTransformer.setParameter("isRoot", (parentId == 0));
238 xhtmlTransformer.setParameter("title", subject);
239 xhtmlTransformer.setParameter("urlBase", urlBase);
240 xhtmlTransformer.setParameter("wwwRead", wwwRead);
241 xhtmlTransformer.setParameter("wwwPost", wwwPost);
242 xhtmlTransformer.setParameter("headComment", String.format("Drupal-NNTP bridge. Transformed: %1$tc. Tidy had to be used: %2$b", new Date(), tidyWasUsed));
243 StringReader input = new StringReader(paragraphedText);
244 StringWriter output = new StringWriter(2 * paragraphedText.length());
245 xhtmlTransformer.transform(new StreamSource(input), new StreamResult(output));
247 return output.toString();
251 * Does not parse XML works just with text.
252 * @param body XHTML fragment that should be put between <body> and </body>
253 * @return simple XHTML document (body wrapped in html and body tags)
255 private static String makeSimpleXHTML(String body) {
256 return "<html xmlns=\"http://www.w3.org/1999/xhtml\"><body>" + body + "</body></html>";
260 * Does not parse XML works just with text.
261 * @param xhtml whole XHTML page
262 * @return content between <body> and </body> tags.
264 private static String makeFragmentXHTML(String xhtml) {
265 final String startTag = "<body>";
266 final String endTag = "</body>";
268 int start = xhtml.indexOf(startTag) + startTag.length();
269 int end = xhtml.lastIndexOf(endTag);
271 return xhtml.substring(start, end);
275 * TODO: refaktorovat, přesunout
277 private static String tidyXhtml(String inputText) throws IOException {
279 * Viz https://sourceforge.net/tracker/index.php?func=detail&aid=3424437&group_id=27659&atid=390966
282 * - použít delší zástupný řetězec, ne jen jeden znak
283 * - umísťovat ho jen tam, kde už nějaký text je (ne mezi >\s*<)
285 inputText = označKonceŘádků(inputText);
287 Runtime r = Runtime.getRuntime();
288 Process p = r.exec(new String[]{"tidy", // http://tidy.sourceforge.net
289 "-asxml", // well formed XHTML
290 "-numeric", // číselné entity
292 "--show-warnings", "false", // žádná varování nás nezajímají
293 "--show-errors", "0", // ani chyby
294 "--doctype", "omit", // doctype nepotřebujeme (doplníme si případně vlastní v XSLT)
295 "--logical-emphasis", "true", // em a strong místo i a b
296 "--literal-attributes", "true", // zachovat mezery a konce řádků v atributech
297 "--force-output", "true" // neznámé značky zahodíme, vložíme jen jejich obsah
300 PrintStream vstupProcesu = new PrintStream(p.getOutputStream());
301 vstupProcesu.print(inputText);
302 vstupProcesu.close();
304 String outputText = streamToString(p.getInputStream());
306 outputText = vraťKonceŘádků(outputText);
311 private static String označKonceŘádků(String text) {
312 text = text.replaceAll(">\\s+<", "> <");
313 text = text.replaceAll("\\n", ZNAKČKA_KONCE_ŘÁDKU + "\n");
317 private static String vraťKonceŘádků(String text) {
318 text = text.replaceAll(ZNAKČKA_KONCE_ŘÁDKU + "\\n", "\n");
319 text = text.replaceAll(ZNAKČKA_KONCE_ŘÁDKU, "\n");
324 * TODO: refaktorovat, přesunout
326 private static String streamToString(InputStream proud) throws IOException {
327 StringBuilder výsledek = new StringBuilder();
328 BufferedReader buf = new BufferedReader(new InputStreamReader(proud));
330 String radek = buf.readLine();
334 výsledek.append(radek);
335 výsledek.append("\n");
338 return výsledek.toString();
341 public static String constructMessageId(long articleID, long groupID, String groupName, String domainName) {
342 StringBuilder sb = new StringBuilder();
344 sb.append(articleID);
348 sb.append(groupName);
350 sb.append(domainName);
352 return sb.toString();
356 * @return article ID of parent of this message | or null, if this is root article and not reply to another one
358 public Long getParentID() {
363 * @return group ID of this message | or null, if this message is not reply to any other one – which is wrong because we have to know the group
365 public Long getGroupID() {
371 * @param messageID <{0}-{1}-{2}@domain.tld> where {0} is nntp_id and {1} is group_id and {2} is group_name
372 * @return array where [0] = nntp_id and [1] = group_id and [2] = group_name or returns null if messageID is invalid
374 private static String[] parseMessageID(String messageID) {
375 if (messageID.matches("<[0-9]+\\-[0-9]+\\-[a-z0-9\\.]+@.+>")) {
376 return messageID.substring(1).split("@")[0].split("\\-");
382 public static Long parseArticleID(String messageID) {
383 String[] localPart = parseMessageID(messageID);
384 if (localPart == null) {
387 return Long.parseLong(localPart[0]);
391 public static Long parseGroupID(String messageID) {
392 String[] localPart = parseMessageID(messageID);
393 if (localPart == null) {
396 return Long.parseLong(localPart[1]);
398 // parseGroupName() will be same as this method, just with:
399 // return localPart[2];
404 public void setHeader(String name, String value) throws MessagingException {
405 super.setHeader(name, value);
407 if (MESSAGE_ID_HEADER.equalsIgnoreCase(name)) {
413 public final void addHeader(String name, String value) throws MessagingException {
414 super.addHeader(name, value);
416 if (MESSAGE_ID_HEADER.equalsIgnoreCase(name)) {
422 public void removeHeader(String name) throws MessagingException {
423 super.removeHeader(name);
425 if (MESSAGE_ID_HEADER.equalsIgnoreCase(name)) {
430 public void setMessageID(String messageID) {
431 this.messageID = messageID;
435 protected void updateMessageID() throws MessagingException {
436 if (messageID == null) {
437 super.updateMessageID();
439 setHeader(MESSAGE_ID_HEADER, messageID);
444 * Call {@link #saveChanges()} before this method, if you want all headers including such ones like:
446 * <pre>MIME-Version: 1.0
447 *Content-Type: multipart/alternative;</pre>
449 * @return serialized headers
450 * @throws MessagingException if getAllHeaders() fails
452 public String getHeaders() throws MessagingException {
453 StringBuilder sb = new StringBuilder();
454 for (Enumeration eh = getAllHeaderLines(); eh.hasMoreElements();) {
455 sb.append(eh.nextElement());
458 return sb.toString();
461 public byte[] getBody() throws IOException, MessagingException {
464 ArrayList<String> skipHeaders = new ArrayList<String>();
465 for (Enumeration eh = getAllHeaders(); eh.hasMoreElements();) {
466 Header h = (Header) eh.nextElement();
467 skipHeaders.add(h.getName());
470 ByteArrayOutputStream baos = new ByteArrayOutputStream(1024);
471 writeTo(baos, skipHeaders.toArray(new String[skipHeaders.size()]));
472 return baos.toByteArray();
476 * Transforms message content to valid XHTML and strips html and body tags.
477 * When receiving message from user through NNTP
478 * this method is used to get text that should be saved into databse.
479 * @return XHTML fragment – content between <body> and </body> tags.
481 public String getBodyXhtmlFragment() throws StorageBackendException {
483 * TODO: podporovat i zprávy přímo v HTML a multipart.
486 Object c = getContent();
487 if (isMimeType("text/plain") && c instanceof String) {
488 String xhtml = readXhtmlText(
495 return makeFragmentXHTML(xhtml);
497 throw new StorageBackendException("Only text/plain messages are supported for now – post it as plain text please.");
499 } catch (Exception e) {
500 throw new StorageBackendException(e);