PHP函数替换RSS提要中的链接的时间长于x char

Question

I have an RSS feed which I am parsing and in some pages of the feed there are very long URLs that break my page. For example, some are like this when you see them on the page.

http://example.com/coolthings/893748662/photos/37774656-ID/another_dam_dir/MORESTUFF.php?id=7837839946HS67355

So because they used the actual URL for some of the links in the page instead of linking to some text about it I am left with crazy long URL-based links. I want a way to build some code into the parser that will automatically detect these crazy long links anywhere in the page content and maybe shorten them to something like http://example.com/coolthings/ .

I would like some type of PHP function I can use that will scan through the page and shorten the crazy long URLs if they are found. I have looked all over and even tried to make something that can do this and have failed at the attempts.

I can use preg_match_all to find the URLs but can't seem to find a simple way of replacing them within the page.

The page is parsed via curl and SimpleXML and a few regexs. Ideas how I can do this greatly appreciated.

//Not Working attempt ... This will find the crazy long urls but then they must be replaced in the page with the short version and i cant seem to make that part of my idea work. I also wrote a func cutto_mid to trim at middle. I will use that when i can get this link replace code to work.

function short_link($text){
$regex = "a[\s]+[^>]*?href[\s]?=[\s\"\']+"."(.*?)[\"\']+.*?>"."([^<]+|.*?)?<\/a>";
preg_match_all ("/$regex/i", $text, &$matches);
$matches = $matches[1];
foreach($matches as $vars){     
if (strlen($vars) > '95'){
//return "<BR>".cutto_mid($vars, 20, 20, 50); //testing ...
return preg_replace("#$regex#i",'<a href="'.$vars .'">Replace W Short Link</a>', $text);
}
}
}

Answer 1

You should not be using Regular Expressions to parse HTML, especially not when you're already using an XML parser. Rather use XPath to fetch all anchors with an href, then test the length of their value. Something like this:

$links = $xml->xpath ("//a[@href]");
foreach ($links as &$l) {
    if (strlen ($l[0]) < CUTOFF)
        continue;

    $l[0] = substr ($l[0], 0, CUTOFF-3).'...';
}