java/实验三/MovieCrawler.java

import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
import java.util.ArrayList;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class MovieCrawler extends BaseMovieCrawler {

    private static final String URL = "https://movie.douban.com/top250?start=0";

    @Override
    public List<Movie> crawl(int limit) {
        List<Movie> movies = new ArrayList<>();
        try {
            Document doc = Jsoup.connect(URL)
                    .userAgent("Mozilla/5.0")
                    .timeout(10000)
                    .get();

            Elements items = doc.select(".item");
            int count = 0;

            for (Element item : items) {
                if (count >= limit) break;

                Movie movie = parseMovie(item);
                if (movie != null) {
                    movies.add(movie);
                    count++;
                }
            }
        } catch (Exception e) {
            System.out.println("模拟电影数据（因网络403）");
            for (int i = 0; i < limit; i++) {
                if (i % 2 == 0) {
                    movies.add(new TheatreMovie("肖申克的救赎 " + i, 1994, 9.7, "剧情", 49.9));
                } else {
                    movies.add(new AnimatedMovie("哪吒之魔童降世 " + i, 2019, 8.4, "动画", "可可豆"));
                }
            }
        }
        return movies;
    }

    @Override
    protected Movie parseMovie(Element element) {
        try {
            String title = element.select(".hd .title").first().text();
            String info = element.select(".bd p").first().text();
            Matcher m = Pattern.compile("(\\d{4})").matcher(info);
            int year = m.find() ? Integer.parseInt(m.group(1)) : 2023;
            double rating = Double.parseDouble(element.select(".rating_num").first().text());
            String genre = info.contains("/") ? info.split("/")[2].trim() : "未知";

            return new TheatreMovie(title, year, rating, genre, 59.9);
        } catch (Exception e) {
            return null;
        }
    }
}