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.Arrays;
34 import java.util.Date;
35 import java.util.Enumeration;
36 import java.util.logging.Level;
37 import java.util.logging.Logger;
38 import java.util.regex.Matcher;
39 import java.util.regex.Pattern;
40 import javax.mail.Header;
41 import javax.mail.MessagingException;
42 import javax.mail.Multipart;
43 import javax.mail.Session;
44 import javax.mail.internet.InternetAddress;
45 import javax.mail.internet.MimeBodyPart;
46 import javax.mail.internet.MimeMessage;
47 import javax.mail.internet.MimeMultipart;
48 import javax.xml.parsers.DocumentBuilder;
49 import javax.xml.parsers.DocumentBuilderFactory;
50 import javax.xml.parsers.ParserConfigurationException;
51 import javax.xml.transform.Transformer;
52 import javax.xml.transform.TransformerException;
53 import javax.xml.transform.TransformerFactory;
54 import javax.xml.transform.dom.DOMSource;
55 import javax.xml.transform.stream.StreamResult;
56 import javax.xml.transform.stream.StreamSource;
57 import org.sonews.daemon.NNTPConnection;
58 import org.sonews.util.io.Resource;
59 import org.w3c.dom.Document;
60 import org.xml.sax.SAXException;
63 * This is MimeMessage which enables custom Message-ID header
64 * (this header will not be overwritten by the default one like in MimeMessage).
66 * Also add header and body separate serialization.
68 * And can be deserialized from SQL ResultSet
70 * @author František Kučera (frantovo.cz)
72 public class DrupalMessage extends MimeMessage {
75 * If body of message posted by user through NNTP starts with this text,
76 * it will be treated as formated text in Markdown syntax.
78 private static final String MARKDOWN_HEADER = "#!markdown\r\n";
79 private static final Logger log = Logger.getLogger(DrupalMessage.class.getName());
80 private static final String MESSAGE_ID_HEADER = "Message-ID";
81 private static final String CRLF = "\r\n";
82 public static final String CHARSET = "UTF-8";
83 private static final String XHTML_CONTENT_TYPE = "text/html; charset=" + CHARSET;
84 private static final String ZNAKČKA_KONCE_ŘÁDKU = "◆";
85 private String messageID;
86 private Long parentID;
88 private TransformerFactory transformerFactory;
89 private DocumentBuilderFactory documentBuilderFactory;
92 * Initializes XML factories (Transformer, DocumentBuilder).
94 private void initFactories() {
95 transformerFactory = TransformerFactory.newInstance();
96 documentBuilderFactory = DocumentBuilderFactory.newInstance();
98 * Komentáře nás nepotřebujeme
99 * (a museli bychom je brát v úvahu při dělení odstavců:
100 * v současné verzi XSLT odstavcovače by nám případný komentář
101 * rozdělil text na dva odstavce, přestože to má být odstavec jede).
103 documentBuilderFactory.setIgnoringComments(true);
107 * Constructs MIME message from SQL result.
108 * @param rs ResultSet containing message data. No {@link ResultSet#next()} will be called, just values from current row will be read.
109 * @param constructBody true if whole message should be constructed | false if we need only message headers (body will be dummy).
111 public DrupalMessage(ResultSet rs, String myDomain, boolean constructBody) throws SQLException, UnsupportedEncodingException, MessagingException, TransformerException, IOException, ParserConfigurationException, SAXException {
112 super(Session.getDefaultInstance(System.getProperties()));
115 groupID = rs.getLong("group_id");
116 addHeader("Message-id", constructMessageId(rs.getInt("id"), groupID, rs.getString("group_name"), myDomain));
117 addHeader("Newsgroups", rs.getString("group_name"));
118 setFrom(new InternetAddress(rs.getString("sender_email"), rs.getString("sender_name")));
119 setSubject(rs.getString("subject"));
120 setSentDate(new Date(rs.getLong("created")));
122 parentID = rs.getLong("parent_id");
124 String parentMessageID = constructMessageId(parentID, rs.getInt("group_id"), rs.getString("group_name"), myDomain);
125 addHeader("In-Reply-To", parentMessageID);
126 addHeader("References", parentMessageID);
130 Multipart multipart = new MimeMultipart("alternative");
131 setContent(multipart);
134 MimeBodyPart htmlPart = new MimeBodyPart();
135 String xhtmlText = readXhtmlText(
136 rs.getString("text"),
137 rs.getString("subject"),
138 rs.getInt("parent_id"),
139 rs.getString("urlBase"),
140 rs.getString("wwwRead"),
141 rs.getString("wwwPost"));
142 htmlPart.setContent(xhtmlText, XHTML_CONTENT_TYPE);
144 /** Plain text part */
145 MimeBodyPart textPart = new MimeBodyPart();
146 String plainText = formatedToPlainText(xhtmlText);
147 textPart.setText(plainText);
148 //addHeader("Lines", String.valueOf(plainText.split("\n").length));
151 * Thunderbirdu záleží, v jakém pořadí části jsou
152 * (když je prostý text druhý, html se nezobrazí),
153 * KNode zobrazuje HTML správně, i když je na prvním místě.
155 multipart.addBodyPart(textPart);
156 multipart.addBodyPart(htmlPart);
158 /** empty body, just headers */
164 * Constructs MIME message from article posted by user.
165 * @param article article that came through NNTP.
166 * @throws MessagingException
168 public DrupalMessage(Article article) throws MessagingException {
169 super(Session.getDefaultInstance(System.getProperties()), serializeArticle(article));
172 String[] replyToHeaders = getHeader("In-Reply-To");
173 String[] referencesHeaders = getHeader("References");
174 String parentMessageID;
175 if (replyToHeaders != null && replyToHeaders.length == 1) {
176 parentMessageID = replyToHeaders[0];
177 } else if (referencesHeaders != null && referencesHeaders.length == 1) {
178 Pattern p = Pattern.compile("(\\s*<.*>)*\\s*(<.*>)");
179 Matcher m = p.matcher(referencesHeaders[0]);
182 parentMessageID = m.group(2);
184 throw new MessagingException("Message posted by user had invalid References header: " + referencesHeaders[0]);
187 throw new MessagingException("Message posted by user must have exactly one In-Reply-To header. Reply-To headers: " + Arrays.toString(replyToHeaders) + " Referemces headers: " + Arrays.toString(referencesHeaders));
190 parentID = parseArticleID(parentMessageID);
191 groupID = parseGroupID(parentMessageID);
194 private static InputStream serializeArticle(Article a) {
195 byte articleHeaders[] = a.getHeaderSource().getBytes();
196 byte delimiter[] = (NNTPConnection.NEWLINE + NNTPConnection.NEWLINE).getBytes();
197 byte body[] = a.getBody();
199 byte message[] = new byte[articleHeaders.length + delimiter.length + body.length];
201 System.arraycopy(articleHeaders, 0, message, 0, articleHeaders.length);
202 System.arraycopy(delimiter, 0, message, articleHeaders.length, delimiter.length);
203 System.arraycopy(body, 0, message, articleHeaders.length + delimiter.length, body.length);
205 return new ByteArrayInputStream(message);
209 * @param xhtmlText well-formed XHTML
210 * @return plain text representation of this formated text
212 private String formatedToPlainText(String xhtmlText) {
214 Transformer textTransformer = transformerFactory.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeTextPart.xsl")));
216 StringReader input = new StringReader(xhtmlText);
217 StringWriter output = new StringWriter(xhtmlText.length());
218 textTransformer.transform(new StreamSource(input), new StreamResult(output));
220 return output.toString();
221 } catch (Exception e) {
223 * TODO: lepší ošetření chyby
225 log.log(Level.WARNING, "Error while transforming article to plain text", e);
226 return "Při transformaci příspěvku bohužel došlo k chybě.";
230 private DOMSource readDOM(String xml) throws ParserConfigurationException, SAXException, IOException {
231 DocumentBuilder db = documentBuilderFactory.newDocumentBuilder();
232 Document d = db.parse(new ByteArrayInputStream(xml.getBytes("UTF-8")));
233 return new DOMSource(d);
236 private String readXhtmlText(String sourceText, String subject, long parentId, String urlBase, String wwwRead, String wwwPost) throws TransformerException, IOException, ParserConfigurationException, SAXException {
239 * - znovupoužívat XSL transformér (nejen v instanci)
240 * - používat cache, ukládat si vygenerované články
242 String wrappedText = makeSimpleXHTML(sourceText);
244 Transformer paragraphTransformer = transformerFactory.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart-make-paragraphs.xsl")));
245 String paragraphedText;
246 boolean tidyWasUsed = false;
248 StringWriter output = new StringWriter(2 * wrappedText.length());
249 paragraphTransformer.transform(readDOM(wrappedText), new StreamResult(output));
250 paragraphedText = output.toString();
251 } catch (Exception e) {
252 log.log(Level.FINER, "HTML input was shitty – Tidy had to be called.", e);
253 StringWriter output = new StringWriter(2 * wrappedText.length());
254 paragraphTransformer.transform(readDOM(tidyXhtml(wrappedText)), new StreamResult(output));
255 paragraphedText = output.toString();
259 Transformer xhtmlTransformer = transformerFactory.newTransformer(new StreamSource(Resource.getAsStream("helpers/mimeXhtmlPart.xsl")));
260 xhtmlTransformer.setParameter("isRoot", (parentId == 0));
261 xhtmlTransformer.setParameter("title", subject);
262 xhtmlTransformer.setParameter("urlBase", urlBase);
263 xhtmlTransformer.setParameter("wwwRead", wwwRead);
264 xhtmlTransformer.setParameter("wwwPost", wwwPost);
265 xhtmlTransformer.setParameter("headComment", String.format("Drupal-NNTP bridge. Transformed: %1$tc. Tidy had to be used: %2$b", new Date(), tidyWasUsed));
266 StringReader paragraphedReader = new StringReader(paragraphedText);
267 StringWriter xhtmlWriter = new StringWriter(2 * paragraphedText.length());
268 xhtmlTransformer.transform(new StreamSource(paragraphedReader), new StreamResult(xhtmlWriter));
270 return xhtmlWriter.toString();
274 * Converts markdown to XHTML.
275 * @param markdown text in Markdown syntax
276 * @return XHTML document (with html/body elements)
277 * @throws StorageBackendException when markdown proces returned any errors
278 * (other exceptions are thrown when afterwards XHTML validation fails).
280 private String readXhtmlTextMarkdown(String markdown) throws TransformerException, IOException, ParserConfigurationException, SAXException, StorageBackendException {
281 Runtime r = Runtime.getRuntime();
282 Process p = r.exec(new String[]{"sudo", "-u", "markdown", "/usr/bin/markdown"});
284 PrintStream processInput = new PrintStream(p.getOutputStream());
285 processInput.print(markdown);
286 processInput.close();
288 String errors = streamToString(p.getErrorStream());
289 String htmlFragment = streamToString(p.getInputStream());
291 if (errors.length() == 0) {
292 String htmlDocument = makeSimpleXHTML(htmlFragment);
293 String xhtmlDocument = readXhtmlText(htmlDocument, null, -1, null, null, null);
294 return xhtmlDocument;
296 throw new StorageBackendException("Error while transforming Markdown to XHTML: " + errors);
301 * Does not parse XML works just with text.
302 * @param body XHTML fragment that should be put between <body> and </body>
303 * @return simple XHTML document (body wrapped in html and body tags)
305 private static String makeSimpleXHTML(String body) {
306 return "<html xmlns=\"http://www.w3.org/1999/xhtml\"><body>" + body + "</body></html>";
310 * Does not parse XML works just with text.
311 * @param xhtml whole XHTML page
312 * @return content between <body> and </body> tags.
314 private static String makeFragmentXHTML(String xhtml) {
315 final String startTag = "<body>";
316 final String endTag = "</body>";
318 int start = xhtml.indexOf(startTag) + startTag.length();
319 int end = xhtml.lastIndexOf(endTag);
321 return xhtml.substring(start, end);
325 * TODO: refaktorovat, přesunout
327 private static String tidyXhtml(String inputText) throws IOException {
329 * Viz https://sourceforge.net/tracker/index.php?func=detail&aid=3424437&group_id=27659&atid=390966
332 * - použít delší zástupný řetězec, ne jen jeden znak
333 * - umísťovat ho jen tam, kde už nějaký text je (ne mezi >\s*<)
335 inputText = označKonceŘádků(inputText);
337 Runtime r = Runtime.getRuntime();
338 Process p = r.exec(new String[]{"tidy", // http://tidy.sourceforge.net
339 "-asxml", // well formed XHTML
340 "-numeric", // číselné entity
342 "--show-warnings", "false", // žádná varování nás nezajímají
343 "--show-errors", "0", // ani chyby
344 "--doctype", "omit", // doctype nepotřebujeme (doplníme si případně vlastní v XSLT)
345 "--logical-emphasis", "true", // em a strong místo i a b
346 "--literal-attributes", "true", // zachovat mezery a konce řádků v atributech
347 "--force-output", "true" // neznámé značky zahodíme, vložíme jen jejich obsah
350 PrintStream vstupProcesu = new PrintStream(p.getOutputStream());
351 vstupProcesu.print(inputText);
352 vstupProcesu.close();
354 String outputText = streamToString(p.getInputStream());
356 outputText = vraťKonceŘádků(outputText);
361 private static String označKonceŘádků(String text) {
362 text = text.replaceAll(">\\s+<", "> <");
363 text = text.replaceAll("\\n", ZNAKČKA_KONCE_ŘÁDKU + "\n");
367 private static String vraťKonceŘádků(String text) {
368 text = text.replaceAll(ZNAKČKA_KONCE_ŘÁDKU + "\\n", "\n");
369 text = text.replaceAll(ZNAKČKA_KONCE_ŘÁDKU, "\n");
374 * TODO: refaktorovat, přesunout
376 private static String streamToString(InputStream proud) throws IOException {
377 StringBuilder výsledek = new StringBuilder();
378 BufferedReader buf = new BufferedReader(new InputStreamReader(proud));
380 String radek = buf.readLine();
384 výsledek.append(radek);
385 výsledek.append("\n");
388 return výsledek.toString();
391 public static String constructMessageId(long articleID, long groupID, String groupName, String domainName) {
392 StringBuilder sb = new StringBuilder();
394 sb.append(articleID);
398 sb.append(groupName);
400 sb.append(domainName);
402 return sb.toString();
406 * @return article ID of parent of this message | or null, if this is root article and not reply to another one
408 public Long getParentID() {
413 * @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
415 public Long getGroupID() {
421 * @param messageID <{0}-{1}-{2}@domain.tld> where {0} is nntp_id and {1} is group_id and {2} is group_name
422 * @return array where [0] = nntp_id and [1] = group_id and [2] = group_name or returns null if messageID is invalid
424 private static String[] parseMessageID(String messageID) {
425 if (messageID.matches("<[0-9]+\\-[0-9]+\\-[a-z0-9\\.]+@.+>")) {
426 return messageID.substring(1).split("@")[0].split("\\-");
432 public static Long parseArticleID(String messageID) {
433 String[] localPart = parseMessageID(messageID);
434 if (localPart == null) {
437 return Long.parseLong(localPart[0]);
441 public static Long parseGroupID(String messageID) {
442 String[] localPart = parseMessageID(messageID);
443 if (localPart == null) {
446 return Long.parseLong(localPart[1]);
448 // parseGroupName() will be same as this method, just with:
449 // return localPart[2];
454 public void setHeader(String name, String value) throws MessagingException {
455 super.setHeader(name, value);
457 if (MESSAGE_ID_HEADER.equalsIgnoreCase(name)) {
463 public final void addHeader(String name, String value) throws MessagingException {
464 super.addHeader(name, value);
466 if (MESSAGE_ID_HEADER.equalsIgnoreCase(name)) {
472 public void removeHeader(String name) throws MessagingException {
473 super.removeHeader(name);
475 if (MESSAGE_ID_HEADER.equalsIgnoreCase(name)) {
480 public void setMessageID(String messageID) {
481 this.messageID = messageID;
485 protected void updateMessageID() throws MessagingException {
486 if (messageID == null) {
487 super.updateMessageID();
489 setHeader(MESSAGE_ID_HEADER, messageID);
494 * Call {@link #saveChanges()} before this method, if you want all headers including such ones like:
496 * <pre>MIME-Version: 1.0
497 *Content-Type: multipart/alternative;</pre>
499 * @return serialized headers
500 * @throws MessagingException if getAllHeaders() fails
502 public String getHeaders() throws MessagingException {
503 StringBuilder sb = new StringBuilder();
504 for (Enumeration eh = getAllHeaderLines(); eh.hasMoreElements();) {
505 sb.append(eh.nextElement());
508 return sb.toString();
511 public byte[] getBody() throws IOException, MessagingException {
514 ArrayList<String> skipHeaders = new ArrayList<String>();
515 for (Enumeration eh = getAllHeaders(); eh.hasMoreElements();) {
516 Header h = (Header) eh.nextElement();
517 skipHeaders.add(h.getName());
520 ByteArrayOutputStream baos = new ByteArrayOutputStream(1024);
521 writeTo(baos, skipHeaders.toArray(new String[skipHeaders.size()]));
522 return baos.toByteArray();
526 * Transforms message content to valid XHTML and strips html and body tags.
527 * When receiving message from user through NNTP
528 * this method is used to get text that should be saved into databse.
529 * @return XHTML fragment – content between <body> and </body> tags.
531 public String getBodyXhtmlFragment() throws StorageBackendException {
533 * TODO: podporovat i zprávy přímo v HTML a multipart.
536 Object c = getContent();
537 if (isMimeType("text/plain") && c instanceof String) {
538 String inputText = (String) c;
541 if (inputText.startsWith(MARKDOWN_HEADER)) {
542 xhtml = readXhtmlTextMarkdown(inputText.substring(MARKDOWN_HEADER.length()));
545 xhtml = readXhtmlText(
553 return makeFragmentXHTML(xhtml);
555 throw new StorageBackendException("Only text/plain messages are supported for now – post it as plain text please.");
557 } catch (Exception e) {
558 throw new StorageBackendException(e);
562 public String getBodyPlainText() throws StorageBackendException {
564 * TODO: netransformovat XHTML 2x
566 return formatedToPlainText(makeSimpleXHTML(getBodyXhtmlFragment()));