curlnbalist.php

<?php
/**
 * zhoumengkang < i@zhoumengkang.com >
 * NBA数据采集
 */

set_time_limit(0);

//定义数据库常量(换成你的数据库帐号密码)
define(HOST, 'localhost');
define(USER, 'root');
define(PASSWORD, 'zmkzmk');

$url = 'http://www.stat-nba.com/query_team.php?QueryType=game&order=0&crtcol=tm_out&GameType=season&Team_id=IND&PageNum=30&Season0=2013&Season1=2014';
$baseUrl = 'http://www.stat-nba.com/';

//根据url查看是哪个球队
$teamId = parse_url($url);
parse_str($teamId['query']);
$teamId = $Team_id;//////////////////////////////////////////////////////////////////////////////////teamId
$teamUrl = './team/'.$teamId.'.html';

//获取列表页开始
$res = getContent($url);
preg_match_all('/(game.php\?id=\d{1,})/', $res, $matches);

//连接数据库
$linksArr = $matches[1];
$link = mysql_connect(HOST, USER, PASSWORD);
if (!$link) {
    die('Could not connect: ' . mysql_error());
}
mysql_set_charset('utf8');
mysql_select_db('nba', $link);

include 'phpQuery.php';

foreach ($linksArr as $key => $value) {
  $query = parse_url($linksArr[1]);
  parse_str($query['query']); 
  $matchId = $id;  ////////////////////////////////////////////////////////////////////////////////////matchId
  phpQuery::newDocumentFile($baseUrl.$linksArr[1]);
  //获取比赛时间
  $matchTime = pq('.title:eq(0)')->next('div')->html();
  preg_match_all('/(\d{4}-\d{1,2}-\d{1,2})/', $matchTime, $matches);
  $matchTime = $matches[1][0];///////////////////////////////////////////////////////////////////////matchTime





  $teamTitle = pq('.title a');
  $teamName = $teamTitle->html();

  foreach ($teamTitle as $v) {
      
      //与$teamId核对
      if (pq($v)->attr('href')==$teamUrl) {
          //整队数据表
          $infoTable = pq($v)->parent('.title')->next('div')->next('div')->find('.stat_box')->parent()->html();
          //队员的列表数组
          $playersArr = pq($infoTable)->find(".sort");
          foreach ($playersArr as $kk => $vv) {
              //pq($vv)->html();//注释1
              /*
              <td style="border:0px;"></td>
              <td class="normal player_name_out change_color col0 row0"><a href="./player/1342.html" target="_blank">保罗-乔治</a></td>
              <td class="current gs change_color col1 row0" rank="1">1</td>
              <td class="normal mp change_color col2 row0" rank="39">39</td>
              <td class="normal fgper change_color col3 row0" rank="0.473684210526316">47.4%</td>
              <td class="normal fg change_color col4 row0" rank="9">9</td>
              <td class="normal fga change_color col5 row0" rank="19">19</td>
              <td class="normal threepper change_color col6 row0" rank="0.363636363636364">36.4%</td>
              <td class="normal threep change_color col7 row0" rank="4">4</td>
              <td class="normal threepa change_color col8 row0" rank="11">11</td>
              <td class="normal ftper change_color col9 row0" rank="0.769230769230769">76.9%</td>
              <td class="normal ft change_color col10 row0" rank="10">10</td>
              <td class="normal fta change_color col11 row0" rank="13">13</td>
              <td class="normal trb change_color col12 row0" rank="6">6</td>篮板
              <td class="normal orb change_color col13 row0" rank="0">0</td>
              <td class="normal drb change_color col14 row0" rank="6">6</td>
              <td class="normal ast change_color col15 row0" rank="5">5</td>助攻
              <td class="normal stl change_color col16 row0" rank="2">2</td>抢断
              <td class="normal blk change_color col17 row0" rank="0">0</td>盖帽
              <td class="normal tov change_color col18 row0" rank="2">2</td>失误
              <td class="normal pf change_color col19 row0" rank="2">2</td>犯规
              <td class="normal pts change_color col20 row0" rank="32">32</td>得分          
              */

              $playerId = pq($vv)->find('.player_name_out a')->attr('href');
              $playerId = explode('/', $playerId);
              $playerId = trim($playerId[2],'.html');////////////////////////////////////////////////////////////////////////playerId

              $playerName = pq($vv)->find('.player_name_out a')->html();///////////////////////////////////////////////////playerName
              //投篮命中率
              $fgper = pq($vv)->find('.fgper')->html();/////////////////////////////////////////////////////////////////////////fgper
              //命中次数
              $fg = pq($vv)->find('.fg')->html();//////////////////////////////////////////////////////////////////////////////////fg
              //出手次数
              $fga = pq($vv)->find('.fga')->html();///////////////////////////////////////////////////////////////////////////////fga
              //三分命中率
              $threepper = pq($vv)->find('.threepper')->html();/////////////////////////////////////////////////////////////threepper
              //三分命中数
              $threep = pq($vv)->find('.threep')->html();//////////////////////////////////////////////////////////////////////threep
              //三分出手次数
              $threepa = pq($vv)->find('.threepa')->html();///////////////////////////////////////////////////////////////////threepa
              //罚球命中率
              $ftper = pq($vv)->find('.ftper')->html();/////////////////////////////////////////////////////////////////////////ftper
              //罚球命中数
              $ft = pq($vv)->find('.ft')->html();//////////////////////////////////////////////////////////////////////////////////ft
              //罚球数
              $fta = pq($vv)->find('.fta')->html();///////////////////////////////////////////////////////////////////////////////fta
              //篮板
              $trb = pq($vv)->find('.trb')->html();///////////////////////////////////////////////////////////////////////////////trb
              //助攻
              $ast = pq($vv)->find('.ast')->html();///////////////////////////////////////////////////////////////////////////////ast
              //抢断
              $stl = pq($vv)->find('.stl')->html();///////////////////////////////////////////////////////////////////////////////stl
              //盖帽
              $blk = pq($vv)->find('.blk')->html();///////////////////////////////////////////////////////////////////////////////blk
              //失误
              $tov = pq($vv)->find('.tov')->html();///////////////////////////////////////////////////////////////////////////////tov
              //犯规
              $pf = pq($vv)->find('.pf')->html();//////////////////////////////////////////////////////////////////////////////////pf
              //得分
              $pts = pq($vv)->find('.pts')->html();///////////////////////////////////////////////////////////////////////////////pts
              //数据库插入操作略了......
          }

      }
  }
}



function getContent($url ){
    if(!$url){
        die('未给定URL参数.');
    }
    $ch = curl_init($url);
    curl_setopt($ch, CURLOPT_USERAGENT, 'Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Win64; x64; Trident/5.0; KB974487)');
    curl_setopt($ch, CURLOPT_HEADER,0);
    curl_setopt($ch, CURLOPT_AUTOREFERER,true);
    curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, 0);
    curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 0);
    curl_setopt($ch, CURLOPT_ENCODING, ' gzip, deflate');
    curl_setopt($ch, CURLOPT_RETURNTRANSFER,1);
    curl_setopt($ch, CURLOPT_FOLLOWLOCATION,1);
    $cons = curl_exec($ch);
    curl_close($ch);
    return $cons;
}

mysql_close($link);

//数据表的创建
/*********************************************************************************************************************************
SET SQL_MODE = "NO_AUTO_VALUE_ON_ZERO";
SET time_zone = "+00:00";
CREATE DATABASE IF NOT EXISTS `nba` DEFAULT CHARACTER SET utf8 COLLATE utf8_general_ci;
USE `nba`;

DROP TABLE IF EXISTS `match`;
CREATE TABLE IF NOT EXISTS `match` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `match_id` int(11) NOT NULL,
  `time` int(11) NOT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 AUTO_INCREMENT=1 ;

DROP TABLE IF EXISTS `players`;
CREATE TABLE IF NOT EXISTS `players` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `players_id` int(11) NOT NULL,
  `name` varchar(200) NOT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 AUTO_INCREMENT=1 ;

DROP TABLE IF EXISTS `record`;
CREATE TABLE IF NOT EXISTS `record` (
  `id` int(11) NOT NULL,
  `match_id` int(11) NOT NULL,
  `player_id` int(11) NOT NULL,
  `fgper` decimal(10,2) NOT NULL,
  `fg` int(11) NOT NULL,
  `fga` int(11) NOT NULL,
  `threepper` decimal(10,2) NOT NULL,
  `threep` int(11) NOT NULL,
  `threepa` int(11) NOT NULL,
  `ftper` decimal(10,2) NOT NULL,
  `ft` int(11) NOT NULL,
  `fta` int(11) NOT NULL,
  `trb` int(11) NOT NULL,
  `ast` int(11) NOT NULL,
  `stl` int(11) NOT NULL,
  `blk` int(11) NOT NULL,
  `tov` int(11) NOT NULL,
  `pf` int(11) NOT NULL,
  `pts` int(11) NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

DROP TABLE IF EXISTS `team`;
CREATE TABLE IF NOT EXISTS `team` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `team_id` varchar(100) NOT NULL,
  `name` varchar(200) NOT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 AUTO_INCREMENT=1 ;

********************************************************************************************************************************/