java/w10/CrawlCommand.java

package command;

import model.Paper;
import strategy.CrawlerStrategy;
import strategy.StrategyFactory;
import repository.PaperRepository;
import view.ConsoleView;
import java.util.List;

public class CrawlCommand implements Command {
    private ConsoleView view;
    private StrategyFactory strategyFactory;
    
    public CrawlCommand(ConsoleView view, StrategyFactory strategyFactory) {
        this.view = view;
        this.strategyFactory = strategyFactory;
    }
    
    @Override
    public void execute(String[] args, PaperRepository repository) {
        if (args.length < 2) {
            view.showError("请提供要爬取的论文网站URL，格式: crawl <URL>");
            return;
        }
        
        String url = args[1];
        
        if (!isValidUrl(url)) {
            view.showError("无效的URL格式，请提供有效的论文网站URL");
            return;
        }
        
        try {
            view.showInfo("开始爬取论文...");
            view.showInfo("目标URL: " + url);
            
            CrawlerStrategy crawler = strategyFactory.createCrawlerByUrl(url);
            if (crawler == null) {
                view.showError("不支持的网站，请提供支持的论文网站URL");
                return;
            }
            
            String platformName = crawler.getPlatformName();
            repository.init(platformName);
            
            List<Paper> papers = crawler.crawl(url, 10);
            
            if (papers.isEmpty()) {
                view.showInfo("未获取到论文");
            } else {
                List<Paper> uniquePapers = repository.removeDuplicates(papers);
                repository.savePapers(uniquePapers);
                view.showSuccess("成功爬取 " + uniquePapers.size() + " 篇论文");
            }
        } catch (Exception e) {
            view.showError("爬取失败: " + e.getMessage());
        }
    }
    
    private boolean isValidUrl(String url) {
        return url != null && (url.startsWith("http://") || url.startsWith("https://"));
    }

    @Override
    public String getDescription() {
        return "爬取指定URL的论文，格式: crawl <论文网站URL>";
    }
    
    @Override
    public String getName() {
        return "crawl";
    }
}