regex help?

kbc1

Hi all

I think I need help with replacing string content that meets certain criteria.

Basically, I allow users to enter their own content via a WYSIWYG editor in a CMS. This content is then displayed on the site however when users create internal hyperlinks to other pages, the source code is as follows:

<a href="page-one/page-two">my link</a>

This creates massive problems as the user might be viewing the following page for example:

www.mysite.com/mypage/test

When the user clicks the above example link, the URL is redirected to:

www.mysite.com/mypage/test/page-one/page-two

instead of

www.mysite.com/page-one/page-two

I need to make sure any hyperlinks added to the source code are prefixed with a forward slash so the page is referenced from the web root.

Would a function that wraps around the output of my content achieve this and if so, how!?

Thanks for reading
kbc

anoopmail

johanafm

You could of course do it when performing output, but that means doing identical work on every single page request. You should in other words do it when they save their content since this means doing it once.

Personally I'd go with DOM. Wrap their content in '<html><head><title>temp</title><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body id="body_id">' and '</body></html>' to get a valid html document. The meta element is (or at least used to be) necessary for DOM to use utf-8 encoding for html documents (which I'm assuming you use yourself).

After this, you just create a DOM doc, load the html code, use getElementsByTagName to get all anchors, iterate over them and inspect if they need to have a / prepended with getAttribute, use setAttribute to update those elements that need updating. Then you grab the body element with getElementById('body_id'), recreate a string containing html code with saveHTML($body), supplying the body node to get just part of the html code, and finally substr() away the openinig and closing <body> and </body> tags.