X-Git-Url: http://wamblee.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=crawler%2Fkiss%2Fsrc%2Forg%2Fwamblee%2Fcrawler%2Fkiss%2FKissCrawler.java;h=d96ef3e581605eec5d99e6656a9e96d2c10ad5ff;hb=3e7fabe4d28276a0c1c87069e6df43f1790b4fb8;hp=7a0d421d294e50a777c9fe02d6186924960ffbe5;hpb=57d50b5b5f5f61e327018824903d3c78a694d7ff;p=utils diff --git a/crawler/kiss/src/org/wamblee/crawler/kiss/KissCrawler.java b/crawler/kiss/src/org/wamblee/crawler/kiss/KissCrawler.java index 7a0d421d..d96ef3e5 100644 --- a/crawler/kiss/src/org/wamblee/crawler/kiss/KissCrawler.java +++ b/crawler/kiss/src/org/wamblee/crawler/kiss/KissCrawler.java @@ -16,6 +16,7 @@ package org.wamblee.crawler.kiss; +import java.io.ByteArrayOutputStream; import java.io.File; import java.io.FileInputStream; import java.io.FileNotFoundException; @@ -24,24 +25,26 @@ import java.io.IOException; import java.io.InputStream; import java.io.PrintStream; import java.util.ArrayList; +import java.util.Arrays; import java.util.Date; import java.util.List; import java.util.Properties; -import java.util.Set; -import java.util.TreeSet; import java.util.regex.Matcher; import java.util.regex.Pattern; -import javax.mail.Message; import javax.mail.MessagingException; import javax.mail.Session; -import javax.mail.Transport; import javax.mail.internet.InternetAddress; -import javax.mail.internet.MimeMessage; +import javax.xml.transform.TransformerException; import org.apache.commons.httpclient.HttpClient; import org.apache.commons.logging.Log; import org.apache.commons.logging.LogFactory; +import org.apache.commons.mail.EmailException; +import org.apache.commons.mail.HtmlEmail; +import org.apache.xml.serialize.OutputFormat; +import org.apache.xml.serialize.XMLSerializer; +import org.w3c.dom.Document; import org.wamblee.crawler.Action; import org.wamblee.crawler.Configuration; import org.wamblee.crawler.Crawler; @@ -49,6 +52,8 @@ import org.wamblee.crawler.Page; import org.wamblee.crawler.PageException; import org.wamblee.crawler.impl.ConfigurationParser; import org.wamblee.crawler.impl.CrawlerImpl; +import org.wamblee.io.FileResource; +import org.wamblee.xml.XSLT; /** * The KiSS crawler for automatic recording of interesting TV shows. @@ -127,19 +132,20 @@ public class KissCrawler { try { HttpClient client = new HttpClient(); - // client.getHostConfiguration().setProxy("127.0.0.1", 3128); + //client.getHostConfiguration().setProxy("127.0.0.1", 3128); Crawler crawler = createCrawler(aCrawlerConfig, os, client); InputStream programConfigFile = new FileInputStream(new File( aProgramConfig)); - List programFilters = new ProgramConfigurationParser() - .parse(programConfigFile); - + ProgramConfigurationParser parser = new ProgramConfigurationParser(); + parser.parse(programConfigFile); + List programFilters = parser.getFilters(); + Page page = getStartPage(aStartUrl, crawler); TVGuide guide = createGuide(page); PrintVisitor printer = new PrintVisitor(System.out); guide.accept(printer); - processResults(programFilters, guide); + processResults(programFilters, guide, parser.getNotifier()); } finally { os.flush(); os.close(); @@ -158,19 +164,22 @@ public class KissCrawler { * In case of problems sending a summary mail. */ private void processResults(List aProgramCondition, - TVGuide aGuide) throws MessagingException { + TVGuide aGuide, Notifier aNotifier) throws MessagingException { ProgramActionExecutor executor = new ProgramActionExecutor(); for (ProgramFilter filter : aProgramCondition) { List programs = filter.apply(aGuide); - ProgramAction action = filter.getAction(); - for (Program program: programs) { + ProgramAction action = filter.getAction(); + for (Program program : programs) { action.execute(program, executor); } } executor.commit(); - String msg = executor.getReport(); - System.out.println(msg); - sendMail(msg); + try { + aNotifier.send(executor.getXmlReport()); + } catch (NotificationException e) { + throw new RuntimeException(e); + } + sendMail(executor); } /** @@ -276,8 +285,8 @@ public class KissCrawler { keywords = programInfo.getContent().element("keywords") .getText().trim(); } catch (PageException e) { - LOG - .warn("Program details could not be determined for '" + LOG.warn( + "Program details could not be determined for '" + action.getName() + "'", e); } } @@ -299,22 +308,42 @@ public class KissCrawler { * @throws MessagingException * In case of problems sending mail. */ - private void sendMail(String aText) throws MessagingException { + private void sendMail(ProgramActionExecutor aExecutor) throws MessagingException { + String textReport = aExecutor.getReport(); + System.out.println("Text report: \n" + textReport); + System.out.println("XML report:\n" + aExecutor.getXmlReport().asXML()); + + Properties props = new Properties(); props.put("mail.transport.protocol", "smtp"); props.put("mail.smtp.host", "falcon"); props.put("mail.smtp.port", "25"); Session mailSession = Session.getInstance(props); - Message message = new MimeMessage(mailSession); - - message.setFrom(new InternetAddress("erik@brakkee.org")); - message.setRecipient(Message.RecipientType.TO, new InternetAddress( - "erik@brakkee.org")); - message.setSentDate(new Date()); - message.setSubject("KiSS crawler update"); - message.setText(aText); - Transport.send(message); + InternetAddress from = new InternetAddress("erik@brakkee.org"); + + HtmlEmail mail = new HtmlEmail(); + mail.setMailSession(mailSession); + try { + mail.setFrom("erik@brakkee.org"); + mail.setTo(Arrays.asList(new InternetAddress[] { from })); + mail.setSentDate(new Date()); + mail.setSubject("KiSS Crawler Update"); + String html = aExecutor.getXmlReport().asXML(); + Document document = new XSLT().transform(html.getBytes(), new FileResource(new File("reportToHtml.xsl"))); + ByteArrayOutputStream xhtml = new ByteArrayOutputStream(); + XMLSerializer serializer = new XMLSerializer(xhtml, new OutputFormat()); + serializer.serialize(document); + mail.setHtmlMsg(xhtml.toString()); + mail.setTextMsg(textReport); + mail.send(); + } catch (EmailException e) { + throw new RuntimeException(e); + } catch (TransformerException e) { + throw new RuntimeException(e); + } catch (IOException e) { + throw new RuntimeException(e); + } } }