如何在PHP中用链接和标签替换纯文本URL

Question

I am trying to convert plain text into links, hashtags and @tags. I have managed to partially do this, but can't find any way of differentiating between a hashtag and a link containing a hash.

I'm new to using regex so it may be a bit messy!

//link
$message = preg_replace('/((http(s)?)(\:\/\/)|(www\.))([a-zA-Z0-9_\-\.\/\&\%\?\=\+\#\:\;\~\[\]\!\,\@\$\'\(\)\*]+)/', '<a href="http$3://$5$6">$0</a>', $message );
//handle
$message = preg_replace('/[@]+([A-Za-z0-9-_]+)/', '<a href="#$1">$1</a>', $message );
//hashtag
$message = preg_replace('/[#]+([A-Za-z0-9-_]+)/', '<a href="#$1">$1</a>', $message );

The plain text converts to a link as desired, and then breaks at the point of the hash.

Desired text:

www.hello.com/about_us/test%20page/test-page.php#header?this=12345&that=YES

Actual text:

header?this=12345&that=YES">www.hello.com/about_us/test%20page/test-page.php#header?this=12345&that=YES

Is there any way of checking if the hash is part of a URL before converting it to a hashtag?

Answer 1

Your regex for hashtag is this:

/[#]+([A-Za-z0-9-_]+)/

Your stated goal is to make sure it's not part of a URL, which you identify by:

/https?\:\/\//

You can try to use a negative look-behind:

/(?<!https?\:\/\/)[^#]*[#]+([A-Za-z0-9-_]+)

This is not enough for all general cases, but it sounds like you're trying to solve a problem within a scope under your control (a text file you own or something) so hopefully this well help you.

Answer 2

一个对我有用的解决方案：

$message = preg_replace('/^(?<!http)^(?<!www\.)[#]+([A-Za-z0-9-_]+)/', '<a href="#$1">$1</a>', $message );//#hashtag