--- android/TrainInfoService/src/dk/thoerup/traininfoservice/banedk/TimetableFetcher.java 2009/10/02 15:06:08 387 +++ android/TrainInfoService/src/dk/thoerup/traininfoservice/banedk/TimetableFetcher.java 2010/07/14 08:05:31 992 @@ -1,101 +1,169 @@ package dk.thoerup.traininfoservice.banedk; -import java.io.IOException; + import java.net.URL; +import java.sql.SQLException; import java.util.ArrayList; import java.util.List; +import java.util.Map; +import java.util.logging.Level; import java.util.logging.Logger; -import com.gargoylesoftware.htmlunit.Page; -import com.gargoylesoftware.htmlunit.RefreshHandler; -import com.gargoylesoftware.htmlunit.WebClient; -import com.gargoylesoftware.htmlunit.html.DomNodeList; -import com.gargoylesoftware.htmlunit.html.HtmlElement; -import com.gargoylesoftware.htmlunit.html.HtmlPage; +import org.jsoup.nodes.Document; +import org.jsoup.nodes.Element; +import org.jsoup.select.Elements; + +import dk.thoerup.circuitbreaker.CircuitBreaker; +import dk.thoerup.circuitbreaker.CircuitBreakerManager; +import dk.thoerup.traininfoservice.StationDAO; +import dk.thoerup.traininfoservice.Statistics; public class TimetableFetcher { - - class NullRefreshHandler implements RefreshHandler { - public void handleRefresh(Page arg0, URL arg1, int arg2) throws IOException { - } - - } + - TimeoutCache> cache = new TimeoutCache>(120 * 1000); + Map> cache; + Map stationCache; + + StationDAO stationDao = new StationDAO(); Logger logger = Logger.getLogger(TimetableFetcher.class.getName()); + private boolean useTempSite; + + public TimetableFetcher(boolean tmpSite, int cacheTimeout) { + useTempSite = tmpSite; + + cache = new TimeoutMap>(cacheTimeout); + stationCache = new TimeoutMap( 3*60*60*1000 ); + } + List cachedLookupTimetable(String trainID, String type) throws Exception { String key = trainID+type; List list = cache.get(key); if (list == null) { - logger.warning("Timetable: Cache miss " + trainID); //remove before production list = lookupTimetable(trainID,type); cache.put(key, list); } else { - logger.warning("Timetable: Cache hit " + trainID); //remove before production + Statistics.getInstance().incrementTimetableCacheHits(); + logger.info("Timetable: Cache hit " + trainID); } return list; } + + List lookupTimetable(String trainID, String type) throws Exception { + if (useTempSite == false ){ + return lookupTimetableRealSite(trainID, type); + } else { + return new ArrayList(); // no timetable data on temp site + } + } + + int getStationId(String name) { + Integer id = stationCache.get(name); + + if (id == null) { + try { + id = stationDao.getIdByName(name); + stationCache.put(name, id); + } catch (SQLException e) { + logger.log(Level.SEVERE, "getStationId failed", e); + id = -1; + } + } + + return id; + } - List lookupTimetable(String trainID, String type) throws Exception { + List lookupTimetableRealSite(String trainID, String type) throws Exception { List timetableList = new ArrayList(); - String url = "http://www.bane.dk/visRute.asp?W=" + type + "&TogNr=" + trainID + "&artikelId=4276"; - + //String url = "http://www.bane.dk/visRute.asp?W=" + type + "&TogNr=" + trainID + "&artikelId=4276"; + String url = "http://trafikinfo.bane.dk/TrafikInformation/Ruteplan/" + trainID; - final WebClient webClient = new WebClient(); - webClient.setTimeout(2500); - webClient.setJavaScriptEnabled(false); - webClient.setRefreshHandler( new NullRefreshHandler() ); - webClient.setCssEnabled(false); + JsoupInvocation wrapper = new JsoupInvocation( new URL(url) , 2500); + CircuitBreaker breaker = CircuitBreakerManager.getManager().getCircuitBreaker("banedk"); - final HtmlPage page = webClient.getPage(url); - + Document doc = (Document) breaker.invoke(wrapper); + boolean currentStation = false; boolean currentStationSaved = false; - List tables = page.getDocumentElement().getElementsByAttribute("table", "class", "Rute"); + Elements tables = doc.getElementsByClass("Rute"); + if (tables.size() == 1) { - HtmlElement timetable = tables.get(0); - DomNodeList rows = timetable.getElementsByTagName("tr"); + Element timetable = tables.get(0); + Elements rows = timetable.getElementsByTag("tr"); for (int i=0; i fields = row.getElementsByTagName("td"); + Element row = rows.get(i); + Elements fields = row.getElementsByTag("td"); + - if (currentStationSaved == false && fields.get(0).getAttribute("class").equalsIgnoreCase("Tidsstreg")) { + if (currentStationSaved == false && fields.get(0).attr("class").equalsIgnoreCase("Tidsstreg")) { currentStation = true; continue; } TimetableBean bean = new TimetableBean(); - bean.setStation( fields.get(0).asText() ); - bean.setArrival( fields.get(1).asText() ); - bean.setDeparture( fields.get(2).asText() ); + + String station = fields.get(0).text() ; + if (station.equals("København")) + station = "København H"; //correct inconsistency in naming + + bean.setStation( station ); + bean.setArrival( fields.get(1).text() ); + bean.setDeparture( fields.get(2).text() ); + + boolean cancelled = fields.get(3).text().equalsIgnoreCase("aflyst"); + bean.setCancelled(cancelled); if (currentStation == true && currentStationSaved == false ) { bean.setCurrent(currentStation); currentStationSaved = true; } + bean.setStationId( getStationId( station )); + timetableList.add(bean); - + } + + //TODO: There is an off-by-one error in this cancelled parser thingie + final String cancelledString = "Aflyst"; + for (int i=0;i0 && i