parse_url: complete new code for fetching website information

This commit is contained in:
Michael Vogel 2012-07-12 01:17:33 +02:00
parent 47c48aa438
commit 48f67ead9b

View file

@ -1,7 +1,163 @@
<?php
if(!function_exists('deletenode')) {
function deletenode(&$doc, $node)
{
$xpath = new DomXPath($doc);
$list = $xpath->query("//".$node);
foreach ($list as $child)
$child->parentNode->removeChild($child);
}
}
require_once('library/HTML5/Parser.php');
require_once('library/HTMLPurifier.auto.php');
function parseurl_getsiteinfo($url) {
$siteinfo = array();
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $url);
curl_setopt($ch, CURLOPT_HEADER, 1);
curl_setopt($ch, CURLOPT_NOBODY, 0);
curl_setopt($ch, CURLOPT_TIMEOUT, 3);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch,CURLOPT_USERAGENT,'Opera/9.64(Windows NT 5.1; U; de) Presto/2.1.1');
$header = curl_exec($ch);
curl_close($ch);
if (preg_match('/charset=(.*?)\n/', $header, $matches))
$charset = trim(array_pop($matches));
else
$charset = "utf-8";
$pos = strpos($header, "\r\n\r\n");
if ($pos)
$body = trim(substr($header, $pos));
else
$body = $header;
$body = mb_convert_encoding($body, "UTF-8", $charset);
$body = mb_convert_encoding($body, 'HTML-ENTITIES', "UTF-8");
$doc = new DOMDocument();
@$doc->loadHTML($body);
deletenode($doc, 'style');
deletenode($doc, 'script');
deletenode($doc, 'option');
deletenode($doc, 'h1');
deletenode($doc, 'h2');
deletenode($doc, 'h3');
deletenode($doc, 'h4');
deletenode($doc, 'h5');
deletenode($doc, 'h6');
deletenode($doc, 'ol');
deletenode($doc, 'ul');
$xpath = new DomXPath($doc);
$list = $xpath->query("head/title");
foreach ($list as $node)
$siteinfo["title"] = html_entity_decode($node->nodeValue, ENT_QUOTES, "UTF-8");
$list = $xpath->query("head/meta[@name]");
foreach ($list as $node) {
$attr = array();
if ($node->attributes->length)
foreach ($node->attributes as $attribute)
$attr[$attribute->name] = $attribute->value;
$attr["content"] = html_entity_decode($attr["content"], ENT_QUOTES, "UTF-8");
switch (strtolower($attr["name"])) {
case "fulltitle":
$siteinfo["title"] = $attr["content"];
break;
case "description":
$siteinfo["text"] = $attr["content"];
break;
case "dc.title":
$siteinfo["title"] = $attr["content"];
break;
case "dc.description":
$siteinfo["text"] = $attr["content"];
break;
}
}
$list = $xpath->query("head/meta[@property]");
foreach ($list as $node) {
$attr = array();
if ($node->attributes->length)
foreach ($node->attributes as $attribute)
$attr[$attribute->name] = $attribute->value;
$attr["content"] = html_entity_decode($attr["content"], ENT_QUOTES, "UTF-8");
switch (strtolower($attr["property"])) {
case "og:image":
$siteinfo["image"] = $attr["content"];
break;
case "og:title":
$siteinfo["title"] = $attr["content"];
break;
case "og:description":
$siteinfo["text"] = $attr["content"];
break;
}
}
if ($siteinfo["image"] == "") {
require_once('include/Photo.php');
$list = $xpath->query("//img[@src]");
foreach ($list as $node) {
$attr = array();
if ($node->attributes->length)
foreach ($node->attributes as $attribute)
$attr[$attribute->name] = $attribute->value;
// guess mimetype from headers or filename
$type = guess_image_type($attr["src"],true);
$i = fetch_url($attr["src"]);
$ph = new Photo($i, $type);
if(($ph->getWidth() > 200) and ($ph->getHeight() > 200))
$siteinfo["image"] = $attr["src"];
}
}
if ($siteinfo["text"] == "") {
$text = "";
$list = $xpath->query("//div[@class='article']");
foreach ($list as $node)
$text .= " ".trim($node->nodeValue);
if ($text == "") {
$list = $xpath->query("//div[@class='content']");
foreach ($list as $node)
$text .= " ".trim($node->nodeValue);
}
// If none text was found then take the paragraph content
if ($text == "") {
$list = $xpath->query("//p");
foreach ($list as $node)
$text .= " ".trim($node->nodeValue);
}
if ($text != "") {
$text = trim(str_replace(array("\n", "\r"), array(" ", " "), $text));
while (strpos($text, " "))
$text = trim(str_replace(" ", " ", $text));
$siteinfo["text"] = html_entity_decode(substr($text,0,350), ENT_QUOTES, "UTF-8").'...';
}
}
return($siteinfo);
}
function arr_add_hashes(&$item,$k) {
$item = '#' . $item;
@ -40,13 +196,11 @@ function parse_url_content(&$a) {
logger('parse_url: ' . $url);
if($textmode)
$template = $br . '[bookmark=%s]%s[/bookmark]%s' . $br;
else
$template = "<br /><a class=\"bookmark\" href=\"%s\" >%s</a>%s<br />";
$arr = array('url' => $url, 'text' => '');
call_hooks('parse_link', $arr);
@ -60,9 +214,9 @@ function parse_url_content(&$a) {
if($url && $title && $text) {
if($textmode)
$text = $br . $br . '[quote]' . $text . '[/quote]' . $br;
$text = $br . $br . '[quote]' . trim($text) . '[/quote]' . $br;
else
$text = '<br /><br /><blockquote>' . $text . '</blockquote><br />';
$text = '<br /><br /><blockquote>' . trim($text) . '</blockquote><br />';
$title = str_replace(array("\r","\n"),array('',''),$title);
@ -74,145 +228,18 @@ function parse_url_content(&$a) {
killme();
}
$siteinfo = parseurl_getsiteinfo($url);
if($url) {
$s = fetch_url($url);
if($siteinfo["title"] == "") {
echo sprintf($template,$url,$url,'') . $str_tags;
killme();
} else {
echo '';
killme();
$image = $siteinfo["image"];
$text = $siteinfo["text"];
$title = $siteinfo["title"];
}
// logger('parse_url: data: ' . $s, LOGGER_DATA);
if(! $s) {
echo sprintf($template,$url,$url,'') . $str_tags;
killme();
}
$matches = '';
$c = preg_match('/\<head(.*?)\>(.*?)\<\/head\>/ism',$s,$matches);
if($c) {
// logger('parse_url: header: ' . $matches[2], LOGGER_DATA);
try {
$domhead = HTML5_Parser::parse($matches[2]);
} catch (DOMException $e) {
logger('scrape_dfrn: parse error: ' . $e);
}
if($domhead)
logger('parsed header');
}
if(! $title) {
if(strpos($s,'<title>')) {
$title = substr($s,strpos($s,'<title>')+7,64);
if(strpos($title,'<') !== false)
$title = strip_tags(substr($title,0,strpos($title,'<')));
}
}
$config = HTMLPurifier_Config::createDefault();
$config->set('Cache.DefinitionImpl', null);
$purifier = new HTMLPurifier($config);
$s = $purifier->purify($s);
// logger('purify_output: ' . $s);
try {
$dom = HTML5_Parser::parse($s);
} catch (DOMException $e) {
logger('scrape_dfrn: parse error: ' . $e);
}
if(! $dom) {
echo sprintf($template,$url,$url,'') . $str_tags;
killme();
}
$items = $dom->getElementsByTagName('title');
if($items) {
foreach($items as $item) {
$title = trim($item->textContent);
break;
}
}
if(! $text) {
$divs = $dom->getElementsByTagName('div');
if($divs) {
foreach($divs as $div) {
$class = $div->getAttribute('class');
if($class && (stristr($class,'article') || stristr($class,'content'))) {
$items = $div->getElementsByTagName('p');
if($items) {
foreach($items as $item) {
$text = $item->textContent;
if(stristr($text,'<script')) {
$text = '';
continue;
}
$text = strip_tags($text);
if(strlen($text) < 100) {
$text = '';
continue;
}
$text = substr($text,0,250) . '...' ;
break;
}
}
}
if($text)
break;
}
}
if(! $text) {
$items = $dom->getElementsByTagName('p');
if($items) {
foreach($items as $item) {
$text = $item->textContent;
if(stristr($text,'<script'))
continue;
$text = strip_tags($text);
if(strlen($text) < 100) {
$text = '';
continue;
}
$text = substr($text,0,250) . '...' ;
break;
}
}
}
}
if(! $text) {
logger('parsing meta');
$items = (isset($domhead) && is_object($domhead) ? $domhead->getElementsByTagName('meta') : null);
if($items) {
foreach($items as $item) {
$property = $item->getAttribute('property');
if($property && (stristr($property,':description'))) {
$text = $item->getAttribute('content');
if(stristr($text,'<script')) {
$text = '';
continue;
}
$text = strip_tags($text);
$text = substr($text,0,250) . '...' ;
}
if($property && (stristr($property,':image'))) {
$image = $item->getAttribute('content');
if(stristr($text,'<script')) {
$image = '';
continue;
}
$image = strip_tags($image);
if ($image != "") {
$i = fetch_url($image);
if($i) {
require_once('include/Photo.php');
@ -229,18 +256,11 @@ function parse_url_content(&$a) {
$image = $br . $br . '[img=' . $new_width . 'x' . $new_height . ']' . $image . '[/img]';
else
$image = '<br /><br /><img height="' . $new_height . '" width="' . $new_width . '" src="' .$image . '" alt="photo" />';
}
else {
} else {
if($textmode)
$image = $br . $br . '[img]' . $image . '[/img]';
$image = $br.$br.'[img]'.$image.'[/img]';
else
$image = '<br /><br /><img src="' . $image . '" alt="photo" />';
}
}
else
$image = '';
}
$image = '<br /><br /><img src="'.$image.'" alt="photo" />';
}
}
}
@ -248,13 +268,13 @@ function parse_url_content(&$a) {
if(strlen($text)) {
if($textmode)
$text = $br .$br . '[quote]' . $text . '[/quote]' . $br ;
$text = $br.$br.'[quote]'.trim($text).'[/quote]'.$br ;
else
$text = '<br /><br /><blockquote>' . $text . '</blockquote><br />';
$text = '<br /><br /><blockquote>'.trim($text).'</blockquote><br />';
}
if($image) {
$text = $image . $br . $text;
$text = $image.$br.$text;
}
$title = str_replace(array("\r","\n"),array('',''),$title);