Node.js+jade抓取博客所有文章生成静态html文件的实例

function crawlerArc( url ){
  var html = '';
  var str = '';
  var arcDetail = {};
  http.get(url, function (res) {
    res.on('data', function (chunk) {
      html += chunk;
    });
    res.on('end', function () {
      arcDetail = filterArticle( html );
      str = jade.renderFile('./views/layout.jade', arcDetail );
      fs.writeFile( './html/' + arcDetail['id'] + '.html', str, function( err ){
        if( err ) {
          console.log( err );
        }
        console.log( 'success:' + url );
        if ( aUrl.length ) crawlerArc( aUrl.shift() );
      } );
    });
  });
}

var fs = require( 'fs' );
var http = require( 'http' );
var cheerio = require( 'cheerio' );
var jade = require( 'jade' );

var aList = [];
var aUrl = [];

function filterArticle(html) {
  var $ = cheerio.load( html );
  var arcDetail = {};
  var title = $( "#cb_post_title_url" ).text();
  var href = $( "#cb_post_title_url" ).attr( "href" );
  var re = /\/(\d+)\.html/;
  var id = href.match( re )[1];
  var body = $( "#cnblogs_post_body" ).html();
  return {
    id : id,
    title : title,
    href : href,
    body : body
  };
}

function crawlerArc( url ){
  var html = '';
  var str = '';
  var arcDetail = {};
  http.get(url, function (res) {
    res.on('data', function (chunk) {
      html += chunk;
    });
    res.on('end', function () {
      arcDetail = filterArticle( html );
      str = jade.renderFile('./views/layout.jade', arcDetail );
      fs.writeFile( './html/' + arcDetail['id'] + '.html', str, function( err ){
        if( err ) {
          console.log( err );
        }
        console.log( 'success:' + url );
        if ( aUrl.length ) crawlerArc( aUrl.shift() );
      } );
    });
  });
}

function filterHtml(html) {
  var $ = cheerio.load(html);
  var arcList = [];
  var aPost = $("#content").find(".post-list-item");
  aPost.each(function () {
    var ele = $(this);
    var title = ele.find("h2 a").text();
    var url = ele.find("h2 a").attr("href");
    ele.find(".c_b_p_desc a").remove();
    var entry = ele.find(".c_b_p_desc").text();
    ele.find("small a").remove();
    var listTime = ele.find("small").text();
    var re = /\d{4}-\d{2}-\d{2}\s*\d{2}[:]\d{2}/;
    listTime = listTime.match(re)[0];

    arcList.push({
      title: title,
      url: url,
      entry: entry,
      listTime: listTime
    });
  });
  return arcList;
}

function nextPage( html ){
  var $ = cheerio.load(html);
  var nextUrl = $("#pager a:last-child").attr('href');
  if ( !nextUrl ) return getArcUrl( aList );
  var curPage = $("#pager .current").text();
  if( !curPage ) curPage = 1;
  var nextPage = nextUrl.substring( nextUrl.indexOf( '=' ) + 1 );
  if ( curPage < nextPage ) crawler( nextUrl );
}

function crawler(url) {
  http.get(url, function (res) {
    var html = '';
    res.on('data', function (chunk) {
      html += chunk;
    });
    res.on('end', function () {
      aList.push( filterHtml(html) );
      nextPage( html );
    });
  });
}

function getArcUrl( arcList ){
  for( var key in arcList ){
    for( var k in arcList[key] ){
      aUrl.push( arcList[key][k]['url'] );
    }
  }
  crawlerArc( aUrl.shift() );
}

var url = 'http://www.cnblogs.com/ghostwu/';
crawler( url );

doctype html
html
  head
    meta(charset='utf-8')
    title jade+node.js express
    link(rel="stylesheet", href='./css/bower_components/bootstrap/dist/css/bootstrap.min.css')
  body
    block header
      div.container
        div.well.well-lg
          h3 ghostwu的博客
          p js高手之路
    block container
      div.container
        h3
          a(href="#{href}" rel="external nofollow" ) !{title}
        p !{body}
    block footer
      div.container
        footer 版权所有 - by ghostwu