java/w9/command/CrawlCommand.java

package com.example.datacollect.w9.command;

import com.example.datacollect.w9.model.Article;
import com.example.datacollect.w9.view.ConsoleView;
import java.util.List;
import java.util.regex.Pattern;

public class CrawlCommand implements Command {
    private static final Pattern URL_PATTERN =
            Pattern.compile("^(https?://)?([\\w-]+\\.)+[\\w-]+(/[\\w-./?%&=]*)*$");

    private final ConsoleView view;
    private final List<Article> articles;

    public CrawlCommand(ConsoleView view, List<Article> articles) {
        this.view = view;
        this.articles = articles;
    }

    @Override
    public String getName() {
        return "crawl";
    }

    @Override
    public void execute(String[] args, List<Article> articles) {
        if (args.length < 2) {
            view.printError("用法：crawl <url>");
            return;
        }
        String url = args[1];
        if (!isValidUrl(url)) {
            view.printError("URL 格式不正确");
            return;
        }
        // 模拟爬取
        Article art = new Article(
                "模拟标题-" + (articles.size() + 1),
                url,
                "模拟正文内容",
                "模拟作者",
                "2026-05-31"
        );
        articles.add(art);
        view.printSuccess("爬取成功：" + art.getTitle());
    }

    private boolean isValidUrl(String url) {
        return url != null && URL_PATTERN.matcher(url).matches();
    }
}