frontierschedulerforpconlinemobile.java

来自「一个搜索引擎,希望对大家有用」· Java 代码 · 共 40 行

package my.postprocessor;

import java.util.logging.Logger;

import org.archive.crawler.datamodel.CandidateURI;
import org.archive.crawler.postprocessor.FrontierScheduler;

public class FrontierSchedulerForPconlineMobile extends FrontierScheduler {

	private static Logger LOGGER = Logger
			.getLogger(FrontierSchedulerForPconlineMobile.class.getName());

	// 构造函数
	public FrontierSchedulerForPconlineMobile(String name) {
		super(name);
	}

	protected void schedule(CandidateURI caUri) {

		// 取得URL的字符串
		String url = caUri.toString();

		try {
			// URL选择策略
			if (url.indexOf("product.pconline.com.cn/product/") != -1
					|| url.indexOf("img.pconline.com.cn") != -1
					|| url.indexOf("robots.txt") != -1
					|| url.indexOf("dns:") != -1) {

				getController().getFrontier().schedule(caUri);
			} else {
				return;
			}
		} catch (Exception e) {
			e.printStackTrace();
		} finally {
		}
	}
}