veenm/jobfindr
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

RELIABILITY FIX
All checks were successful
Build and Push Quarkus Image / build-and-push (push) Successful in 4m59s
Details

Browse Source
This commit is contained in:
Mel van Veen
2026-03-02 21:27:45 +01:00
parent 5b0219404a
commit 2676a73f29
1 changed files with 19 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
src/main/java/nl/veenm/jobfindr/scrapers/ScraperService.java
View File

@@ -1,7 +1,9 @@
package nl.veenm.jobfindr.scrapers; package nl.veenm.jobfindr.scrapers;
import jakarta.enterprise.context.ApplicationScoped; import jakarta.enterprise.context.ApplicationScoped;
import jakarta.inject.Inject;
import nl.veenm.jobfindr.domain.Vacature; import nl.veenm.jobfindr.domain.Vacature;
import nl.veenm.jobfindr.util.EventService;
import org.jboss.logging.Logger; import org.jboss.logging.Logger;
import org.jsoup.Jsoup; import org.jsoup.Jsoup;
import org.jsoup.nodes.Document; import org.jsoup.nodes.Document;
@@ -13,15 +15,20 @@ import java.time.LocalDate;
import java.time.format.DateTimeFormatter; import java.time.format.DateTimeFormatter;
import java.util.ArrayList; import java.util.ArrayList;
import java.util.List; import java.util.List;
import java.util.stream.Collectors;
@ApplicationScoped @ApplicationScoped
public class ScraperService { public class ScraperService {
private static final Logger log = Logger.getLogger(ScraperService.class); private static final Logger log = Logger.getLogger(ScraperService.class);
private final EventService eventService;
DateTimeFormatter formatter = DateTimeFormatter.ofPattern("dd-MM-yyyy"); DateTimeFormatter formatter = DateTimeFormatter.ofPattern("dd-MM-yyyy");
@Inject
public ScraperService(EventService eventService) {
this.eventService = eventService;
}
public List<Vacature> scrapeVacatures(String url) throws IOException { public List<Vacature> scrapeVacatures(String url) throws IOException {
// Maak verbinding met de website // Maak verbinding met de website
Document doc = Jsoup.connect(url).get(); Document doc = Jsoup.connect(url).get();
@@ -42,9 +49,15 @@ public class ScraperService {
link = "https://www.meesterbaan.nl" + link; link = "https://www.meesterbaan.nl" + link;
} }
assert vacatureElement.parent() != null; String location = "Onbekende locatie";
Element locationElement = vacatureElement.parent().selectFirst("div.location.mt-1"); Element parent = vacatureElement.parent();
String location = locationElement != null ? locationElement.text() : "Onbekende locatie";
if (parent != null) {
Element locationElement = parent.selectFirst("div.location.mt-1");
if (locationElement != null) {
location = locationElement.text();
}
}
// Maak een Vacature-object en voeg het toe aan de lijst // Maak een Vacature-object en voeg het toe aan de lijst
vacatures.add(new Vacature(title, link, location)); vacatures.add(new Vacature(title, link, location));
@@ -54,11 +67,11 @@ public class ScraperService {
try { try {
getVacatureDetails(vacature); getVacatureDetails(vacature);
} catch (IOException e) { } catch (IOException e) {
throw new RuntimeException(e); eventService.logError(getClass().getName(), "scrapeVacatures", "error during scraping", e);
} }
}); });
return vacatures.stream().filter(a -> !a.getLocatie().equals("Onbekende locatie")).collect(Collectors.toList()); return vacatures.stream().filter(a -> !a.getLocatie().equals("Onbekende locatie")).toList();
} }
public void getVacatureDetails(Vacature vacature) throws IOException { public void getVacatureDetails(Vacature vacature) throws IOException {