package task;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.PrintWriter;
import java.util.ArrayList;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
public class Crawler {
//1.成员变量
private static String outputfile = "output\\猫眼电影数据.txt";
private static String sourcefile = "resource\\TOP100榜 - 猫眼电影.htm";
//2.解析本地网页,返回Document
public static Document getHtml() throws IOException {
File f=new File(sourcefile); //创建文件对象
//解析本地文件,第一个参数是文件对象
//第二个参数是网页的字符编码格式
Document doc=Jsoup.parse(f,"utf-8");
return doc;
}
//3.提取数据,返回集合
public static ArrayList<EntityBean> getData(Document doc) throws IOException {
Elements elements=doc.select("#app > div > div > div.main > dl > dd > div > div");
ArrayList<EntityBean> list=new ArrayList<EntityBean>();//创建集合对象
for(Element e :elements ) { //10部电影信息
String name=e.select("[class=name]").text();
String actor=e.select("[class=star]").text();
String releasetime=e.select("[class=releasetime]").text();
String score=e.select("[class=score]").text();
//System.out.println(name+" "+actor+" "+releasetime+" "+score);
EntityBean eb=new EntityBean(name,actor,releasetime,score);
list.add(eb);
}
return list;
}
//4.写出到本地文件系统
public static void wirteToFile(ArrayList<EntityBean> list) throws FileNotFoundException {
PrintWriter pw=new PrintWriter(outputfile);//创建字节打印流对象
for(EntityBean eb : list) { //10
pw.println(eb.getName()+"\t"+eb.getActor()+"\t"+eb.getReleasetime()+
"\t"+eb.getScore());
}
pw.close();
}
//5.主方法
public static void main(String[] args) throws IOException {
Document doc = Crawler.getHtml();//解析网页,得到doc对象
System.out.println("网页标题:"+doc.title());
ArrayList<EntityBean> entitys = Crawler.getData(doc); //提取数据
Crawler.wirteToFile(entitys);
}
}
package task;
public class EntityBean {
//成员变量列表
private String name;
private String actor;
private String releasetime;
private String score;
//get()方法和set()方法
public String getName() {
return name;
}
public void setName(String name) {
this.name = name;
}
public String getActor() {
return actor;
}
public void setActor(String actor) {
this.actor = actor;
}
public String getReleasetime() {
return releasetime;
}
public void setReleasetime(String releasetime) {
this.releasetime = releasetime;
}
public String getScore() {
return score;
}
public void setScore(String score) {
this.score = score;
}
//有参构造方法
public EntityBean(String name, String actor, String releasetime, String score) {
super();
this.name = name;
this.actor = actor;
this.releasetime = releasetime;
this.score = score;
}
//无参构造方法
public EntityBean() {
super();
}
//重写toString()方法
@Override
public String toString() {
return "EntityBean [name=" + name + ", actor=" + actor + ", releasetime=" + releasetime + ", score=" + score
+ "]";
}
}