解析php利用正则表达式解决采集内容排版的问题

  做采集经常遇到的问题是内容排版问题,用了一些时间写了个用正则替换html标签和样式的函数,共享下。

  

复制代码 代码如下:

  /**

  * 格式化内容

  * @param string $content 内容最好统一用utf-8编码

  * @return string

  * !本函数需要开启tidy扩展

  */

  function removeFormat($content) {

  $replaces = array (

  "/<font.*?>/i" => '',

  "/<\/font>/i" => '',

  "/<strong>/i" => '',

  "/<\/strong>/i" => '',

  "/<span.*?>/i" => '',

  "/<\/span>/i" => '',

  "/<div.*?>/i" => "<p>",

  "/<\/div>/i" => "</p>",

  "/<!--<.*?>*-->/i"=>'',

  /* "/<table.*?>/i" => '',//遇到有表格的内容就不要启用

  "/<\/table>/i" => '',

  "/<tbody.*?>/i" => '',

  "/<\/tbody>/i" => '',

  "/<tr.*?>/i" => '<p>',

  "/<\/tr>/i" => '</p>',

  "/<td.*?>/i" => '', */

  "/style=.+?['|\"]/i" => '',

  "/class=.+?['|\"]/i" => '',

  "/id=.+?['|\"]/i"=>'',

  "/lang=.+?['|\"]/i"=>'',

  //"/width=.+?['|\"]/i"=>'',//不好控制注释掉

  //"/height=.+?['|\"]/i"=>'',

  "/border=.+?['|\"]/i"=>'',

  "/face=.+?['|\"]/i"=>'',

  "/<br.*?>[ ]*/i" => "</p><p>",

  "/<iframe.*?>.*<\/iframe>/i" => '',

  "/ /i" => ' ',//空格替换掉

  "/<p.*?>[ |\x{3000}|\r\n]*/ui" => '<p>    ',//替换半角、全角空格,换行符,用 排除写入数据库时产生的编码问题

  );

  $config = array(

  //'indent' => TRUE, //是否缩进

  'output-html' => TRUE,//是否是输出xhtml

  'show-body-only'=>TRUE,//是否只获得到body

  'wrap' => 0

  );

  $content = tidy_repair_string($content, $config, 'utf8');//先利用php自带的tidy类库修复html标签,不然替换的时候容易出现各种诡异的情况

  $content = trim($content);

  foreach ( $replaces as $k => $v ) {

  $content = preg_replace ( $k, $v, $content );

  }

  if(strpos($content,'<p>')>6)//部分内容开头可能缺失<p>标签

  $content = '<p>    '.$content;

  $content = tidy_repair_string($content, $config, 'utf8');//再修复一次,可以去除html空标签

  $content = trim($content);

  return $content;

  }