node.js实现爬虫教程

var http=require('http');
var cheerio=require('cheerio');
 
var url='http://www.imooc.com/learn/348';
 
/****************************
打印得到的数据结构
[{
 chapterTitle:'',
 videos:[{
 title:'',
 id:''
 }]
}]
********************************/
function printCourseInfo(courseData){
 courseData.forEach(function(item){
 var chapterTitle=item.chapterTitle;
 console.log(chapterTitle+'\n');
 item.videos.forEach(function(video){
 console.log(' 【'+video.id+'】'+video.title+'\n');
 })
 });
}
 
 
/*************
分析从网页里抓取到的数据
**************/
function filterChapter(html){
 var courseData=[];
 
 var $=cheerio.load(html);
 var chapters=$('.chapter');
 chapters.each(function(item){
 var chapter=$(this);
 var chapterTitle=chapter.find('strong').text(); //找到章节标题
 var videos=chapter.find('.video').children('li');
 
 var chapterData={
 chapterTitle:chapterTitle,
 videos:[]
 };
 
 videos.each(function(item){
 var video=$(this).find('.studyvideo');
 var title=video.text();
 var id=video.attr('href').split('/video')[1];
 
 chapterData.videos.push({
 title:title,
 id:id
 })
 })
 
 courseData.push(chapterData);
 });
 
 return courseData;
}
 
http.get(url,function(res){
 var html='';
 
 res.on('data',function(data){
 html+=data;
 })
 
 res.on('end',function(){
 var courseData=filterChapter(html);
 printCourseInfo(courseData);
 })
}).on('error',function(){
 console.log('获取课程数据出错');
})