php 批量导入关键词,利用php抓取批量关键词百度推广广告中网址保存在txt文件中...

结合服务器的定时任务可以定时查找关键词广告的竞争程度

$fp = @fopen ( "semallurl.txt", "a+" );

$kws1 = "上海酒店,北京酒店,广州酒店,天津酒店,广州酒店";

$kws = explode ( ",", $kws1 );

foreach ( $kws as $kw ) {

$keywords = $kw;

$enkeywords = urlencode ( $keywords );

$pageURL = "http://www.baidu.com/s?word=$enkeywords";

$contents = fetch ( $pageURL ); /* 抓取页面 */

$contents = preg_replace ( '/

*

* 去掉js代码

*/

$contents_left = "";

$contents_right = "";

$ads_left_green = "";

$ads_left_white = "";

$contentsbytwoside = "";

$ads_right = ""; /* 变量初始化 */

$contentsbytwoside = explode ( '

$contents_left = $contentsbytwoside [2];

$contents_left = '

preg_match_all ( '/(

preg_match_all ( '(

]*class=\\"EC_im EC_fr EC_PP EC_idea1017 \\">.*? )', $contents_right, $ads_right );

echo "------------Keywords ads for" . $kw . "start ------------------------------------
";

fwrite ( $fp, "----------" . $kw . " ads start------------------------- \\r\\n" );

echo "left ads with green background is
";

/* print_r($ads_left_green[0]); */

foreach ( $ads_left_green [0] as $tg1 ) {

preg_match ( '/.*?<\\/span>/', $tg1, $tg11 );

fwrite ( $fp, strip_tags ( $tg11 [0] ) . "\\r\\n" );

echo $tg11 [0] . "
";

}

;

echo "

-------------
";

echo "left ads with white background is
";

/* print_r($ads_left_white[0]); */

foreach ( $ads_left_white [0] as $tg2 ) {

preg_match ( '/.*?<\\/span>/', $tg2, $tg22 );

fwrite ( $fp, strip_tags ( $tg22 [0] ) . "\\r\\n" );

echo $tg22 [0] . "
";

}

;

echo "

-------------
";

echo "right ads with is
";

/* print_r($ads_right[0]); */

foreach ( $ads_right [0] as $tg3 ) {

preg_match ( '/(.*?<\\/font>)/', $tg3, $tg33 );

fwrite ( $fp, strip_tags ( $tg33 [0] ) . "\\r\\n" );

echo $tg33 [0] . "
";

}

;

echo "---------------Keywords ads for" . $kw . "END ------------------------------------
";

fwrite ( $fp, "----------" . $kw . " ads End------------------------- \\r\\n" );

}

;

fwrite ( $fp, date ( "Y-m-d H:i:s" ) . " PHP代码自动运行！\\r\\n" );

fclose ( $fp );

function fetch($Date) {

$ch = curl_init ();

$timeout = 5;

curl_setopt ( $ch, CURLOPT_URL, "$Date" );

curl_setopt ( $ch, CURLOPT_RETURNTRANSFER, 1 );

curl_setopt ( $ch, CURLOPT_USERAGENT, "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)" );

curl_setopt ( $ch, CURLOPT_CONNECTTIMEOUT, $timeout );

$contents = curl_exec ( $ch );

curl_close ( $ch );

return $contents;

}

?>