可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效，请关闭广告屏蔽插件后再试):

问题:

I dont parse this url: http://foldmunka.net

$ch = curl_init("http://foldmunka.net");

//curl_setopt($ch, CURLOPT_NOBODY, true);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
//curl_setopt($ch, CURLOPT_HEADER, true);
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true); //not necessary unless the file redirects (like the PHP example we're using here)
$data = curl_exec($ch);
$info = curl_getinfo($ch);
curl_close($ch);
clearstatcache();
if ($data === false) {
  echo 'cURL failed';
  exit;
}
$dom = new DOMDocument();
$data = mb_convert_encoding($data, 'HTML-ENTITIES', "utf-8");
$data = preg_replace('/<\!\-\-\[if(.*)\]>/', '', $data);
$data = str_replace('<![endif]-->', '', $data);
$data = str_replace('<!--', '', $data);
$data = str_replace('-->', '', $data);
$data = preg_replace('@<script[^>]*?>.*?</script>@si', '', $data);
$data = preg_replace('@<style[^>]*?>.*?</style>@si', '', $data);

$data = mb_convert_encoding($data, 'HTML-ENTITIES', "utf-8");
@$dom->loadHTML($data);

$els = $dom->getElementsByTagName('*');
foreach($els as $el){
  print $el->nodeName." | ".$el->getAttribute('content')."<hr />";
  if($el->getAttribute('title'))$el->nodeValue = $el->getAttribute('title')." ".$el->nodeValue;
  if($el->getAttribute('alt'))$el->nodeValue = $el->getAttribute('alt')." ".$el->nodeValue;
  print $el->nodeName." | ".$el->nodeValue."<hr />";
}

I need sequentially the alt, title attributes and the simple text, but this page i cannot access the nodes within the body tag.

回答1:

Here is a solution with DomDocument and DOMXPath. It is much shorter and runs much faster (~100ms against ~2300ms) than the other solution with Simple HTML DOM Parser.

<?php

function makePlainText($source)
{
    $dom = new DOMDocument();
    $dom->loadHtmlFile($source);

    // use this instead of loadHtmlFile() to load from string:
    //$dom->loadHtml('<html><title>Hello</title><body>Hello this site<img src="asdasd.jpg" alt="alt attr" title="title attr"><a href="open.php" alt="alt attr" title="title attr">click</a> Some text.</body></html>');

    $xpath = new DOMXPath($dom);

    $plain = '';

    foreach ($xpath->query('//text()|//a|//img') as $node)
    {
        if ($node->nodeName == '#cdata-section')
            continue;

        if ($node instanceof DOMElement)
        {
            if ($node->hasAttribute('alt'))
                $plain .= $node->getAttribute('alt') . ' ';
            if ($node->hasAttribute('title'))
                $plain .= $node->getAttribute('title') . ' ';
        }
        if ($node instanceof DOMText)
            $plain .= $node->textContent . ' ';
    }

    return $plain;
}

echo makePlainText('http://foldmunka.net');

回答2:

I'm not sure I'm getting what this script does - the replace operations look like an attempt at sanitation but I'm not sure what for, if you're just extracting some parts of the code - but have you tried the Simple HTML DOM Browser? It may be able to handle the parsing part more easily. Check out the examples.

回答3:

Here is a Simple Html DOM Parser solution just for comparison. It's output is similar for the DomDocument solution's, but this one is more complicated and runs much slower (~2300ms against DomDocument's ~100ms), so I don't recommend to use it:

Updated to work with <img> elements inside <a> elements.

<?php
require_once('simple_html_dom.php');
// we are needing this because Simple Html DOM Parser's callback handler
// doesn't handle arguments
static $processed_plain_text = '';

define('LOAD_FROM_URL', 'loadfromurl');
define('LOAD_FROM_STRING', 'loadfromstring');

function callback_cleanNestedAnchorContent($element)
{
    if ($element->tag == 'a')
        $element->innertext = makePlainText($element->innertext, LOAD_FROM_STRING);
}

function callback_buildPlainText($element)
{
    global $processed_plain_text;

    $excluded_tags = array('script', 'style');

    switch ($element->tag)
    {
        case 'text':
            // filter when 'text' is descendant of 'a', because we are
            // processing the anchor tags with the required attributes
            // separately at the 'a' tag,
            // and also filter out other unneccessary tags
            if (($element->parent->tag != 'a') && !in_array($element->parent->tag, $excluded_tags))
                $processed_plain_text .= $element->innertext . ' ';
            break;
        case 'img':
            $processed_plain_text .= $element->alt . ' ';
            $processed_plain_text .= $element->title . ' ';
            break;
        case 'a':
            $processed_plain_text .= $element->alt . ' ';
            $processed_plain_text .= $element->title . ' ';
            $processed_plain_text .= $element->innertext . ' ';
            break;
    }
}

function makePlainText($source, $mode = LOAD_FROM_URL)
{
    global $processed_plain_text;

    if ($mode == LOAD_FROM_URL)
        $html = file_get_html($source);
    elseif ($mode == LOAD_FROM_STRING)
        $html = str_get_dom ($source);
    else
        return 'Wrong mode defined in makePlainText: ' . $mode;

    $html->set_callback('callback_cleanNestedAnchorContent');

    // processing with the first callback to clean up the anchor tags
    $html = str_get_html($html->save());
    $html->set_callback('callback_buildPlainText');

    // processing with the second callback to build the full plain text with
    // the required attributes of the 'img' and 'a' tags, and excluding the
    // unneccessary ones like script and style tags
    $html->save();

    $return = $processed_plain_text;

    // cleaning the global variable
    $processed_plain_text = '';

    return $return;
}

//$html = '<html><title>Hello</title><body>Hello <span>this</span> site<img src="asdasd.jpg" alt="alt attr" title="title attr"><a href="open.php" alt="alt attr" title="title attr">click <span><strong>HERE</strong></span><img src="image.jpg" title="IMAGE TITLE INSIDE ANCHOR" alt="ALTINACNHOR"></a> Some text.</body></html>';

echo makePlainText('http://foldmunka.net');
//echo makePlainText($html, LOAD_FROM_STRING);