1 |
package dk.thoerup.traininfoservice.banedk; |
package dk.thoerup.traininfoservice.banedk; |
2 |
|
|
3 |
|
|
4 |
import java.io.IOException; |
|
5 |
import java.net.URL; |
import java.net.URL; |
6 |
import java.sql.SQLException; |
import java.sql.SQLException; |
7 |
import java.util.ArrayList; |
import java.util.ArrayList; |
10 |
import java.util.logging.Level; |
import java.util.logging.Level; |
11 |
import java.util.logging.Logger; |
import java.util.logging.Logger; |
12 |
|
|
13 |
import com.gargoylesoftware.htmlunit.BrowserVersion; |
import org.jsoup.nodes.Document; |
14 |
import com.gargoylesoftware.htmlunit.Page; |
import org.jsoup.nodes.Element; |
15 |
import com.gargoylesoftware.htmlunit.RefreshHandler; |
import org.jsoup.select.Elements; |
|
import com.gargoylesoftware.htmlunit.WebClient; |
|
|
import com.gargoylesoftware.htmlunit.html.DomNodeList; |
|
|
import com.gargoylesoftware.htmlunit.html.HtmlElement; |
|
|
import com.gargoylesoftware.htmlunit.html.HtmlPage; |
|
16 |
|
|
17 |
import dk.thoerup.circuitbreaker.CircuitBreaker; |
import dk.thoerup.circuitbreaker.CircuitBreaker; |
18 |
import dk.thoerup.circuitbreaker.CircuitBreakerManager; |
import dk.thoerup.circuitbreaker.CircuitBreakerManager; |
20 |
import dk.thoerup.traininfoservice.Statistics; |
import dk.thoerup.traininfoservice.Statistics; |
21 |
|
|
22 |
public class TimetableFetcher { |
public class TimetableFetcher { |
23 |
|
|
|
class NullRefreshHandler implements RefreshHandler { |
|
|
public void handleRefresh(Page arg0, URL arg1, int arg2) throws IOException { |
|
|
} |
|
|
|
|
|
} |
|
24 |
|
|
25 |
Map<String, List<TimetableBean>> cache; |
Map<String, List<TimetableBean>> cache; |
26 |
Map<String, Integer> stationCache; |
Map<String, Integer> stationCache; |
84 |
//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"; |
85 |
String url = "http://trafikinfo.bane.dk/TrafikInformation/Ruteplan/" + trainID; |
String url = "http://trafikinfo.bane.dk/TrafikInformation/Ruteplan/" + trainID; |
86 |
|
|
|
final WebClient webClient = new WebClient(BrowserVersion.FIREFOX_3); |
|
|
webClient.setTimeout(2500); |
|
|
webClient.setJavaScriptEnabled(false); |
|
|
webClient.setRefreshHandler( new NullRefreshHandler() ); |
|
|
webClient.setCssEnabled(false); |
|
|
|
|
87 |
|
|
88 |
HtmlunitInvocation wrapper = new HtmlunitInvocation(webClient, url); |
JsoupInvocation wrapper = new JsoupInvocation( new URL(url) , 2500); |
89 |
CircuitBreaker breaker = CircuitBreakerManager.getManager().getCircuitBreaker("banedk"); |
CircuitBreaker breaker = CircuitBreakerManager.getManager().getCircuitBreaker("banedk"); |
90 |
|
|
91 |
HtmlPage page = (HtmlPage) breaker.invoke(wrapper); |
Document doc = (Document) breaker.invoke(wrapper); |
92 |
|
|
93 |
|
|
94 |
boolean currentStation = false; |
boolean currentStation = false; |
95 |
boolean currentStationSaved = false; |
boolean currentStationSaved = false; |
96 |
|
|
97 |
List<HtmlElement> tables = page.getDocumentElement().getElementsByAttribute("table", "class", "Rute"); |
Elements tables = doc.getElementsByClass("Rute"); |
98 |
|
|
99 |
if (tables.size() == 1) { |
if (tables.size() == 1) { |
100 |
HtmlElement timetable = tables.get(0); |
Element timetable = tables.get(0); |
101 |
DomNodeList<HtmlElement> rows = timetable.getElementsByTagName("tr"); |
Elements rows = timetable.getElementsByTag("tr"); |
102 |
|
|
103 |
for (int i=0; i<rows.size(); i++) { |
for (int i=0; i<rows.size(); i++) { |
104 |
if (i==0) //First row is column headers |
if (i==0) //First row is column headers |
105 |
continue; |
continue; |
106 |
|
|
107 |
|
|
108 |
HtmlElement row = rows.get(i); |
Element row = rows.get(i); |
109 |
DomNodeList<HtmlElement> fields = row.getElementsByTagName("td"); |
Elements fields = row.getElementsByTag("td"); |
110 |
|
|
111 |
|
|
112 |
if (currentStationSaved == false && fields.get(0).getAttribute("class").equalsIgnoreCase("Tidsstreg")) { |
if (currentStationSaved == false && fields.get(0).attr("class").equalsIgnoreCase("Tidsstreg")) { |
113 |
currentStation = true; |
currentStation = true; |
114 |
continue; |
continue; |
115 |
} |
} |
116 |
|
|
117 |
TimetableBean bean = new TimetableBean(); |
TimetableBean bean = new TimetableBean(); |
118 |
|
|
119 |
String station = fields.get(0).asText() ; |
String station = fields.get(0).text() ; |
120 |
if (station.equals("København")) |
if (station.equals("København")) |
121 |
station = "København H"; //correct inconsistency in naming |
station = "København H"; //correct inconsistency in naming |
122 |
|
|
123 |
bean.setStation( station ); |
bean.setStation( station ); |
124 |
bean.setArrival( fields.get(1).asText() ); |
bean.setArrival( fields.get(1).text() ); |
125 |
bean.setDeparture( fields.get(2).asText() ); |
bean.setDeparture( fields.get(2).text() ); |
126 |
|
|
127 |
boolean cancelled = fields.get(3).asText().equalsIgnoreCase("aflyst"); |
boolean cancelled = fields.get(3).text().equalsIgnoreCase("aflyst"); |
128 |
bean.setCancelled(cancelled); |
bean.setCancelled(cancelled); |
129 |
|
|
130 |
if (currentStation == true && currentStationSaved == false ) { |
if (currentStation == true && currentStationSaved == false ) { |
163 |
} else { |
} else { |
164 |
logger.warning("No time table found, trainID=" + trainID + " type=" + type); |
logger.warning("No time table found, trainID=" + trainID + " type=" + type); |
165 |
} |
} |
166 |
webClient.closeAllWindows(); |
|
167 |
|
|
168 |
return timetableList; |
return timetableList; |
169 |
} |
} |