Regex PHP, pattern matching

Question

Hi i would like to have a regex in php which matches a word in a string but if the word is a link. The problem is that I replace words with links for example:

"text" => < a href = "mylink">text< /a>.

But sometimes I have the problem that it is replaced twice. So I want to avoid this problem. My pattern now is /text/i.

Eg. This is my text. This is my text

I would like to match and replace the first "text" but not the "text" in second string.

edited: I use this for building a glossary.

Anybody Any idea?

Answer 1

but sometimes I have the problem that it is replaced twice. So I want to avoid this problem.

Specify 1 as the fourth argument to your preg_replace call and it will only perform the replacement once.

For example...

preg_replace
    ( '/test/i'
    , '<a class="dfn" href="glossary.php?term=\0">\0</a>'
    , $input
    , 1
    );

More details on the preg_replace man page.

In response to:

Consider this one:

This is my text. In this part also the text should be replaced. However in this part this <a href="#text">text</a> should not be replaced.

In this text two matches from the text should be replaced the third match no because it is an anchor.

This is not a problem for regex, but one that requires parsing the HTML and traversing the DOM in order to run your replace on text nodes not inside a link.

For example, here is some pseudo-code:

MyHtml = parseHtmlDom( Input ); // turn HTML text into a HTML DOM.
SpecialWords = 'comma,delimited,text';

linkify(MyHtml.ChildNodes); // kick off the recursive function.

function linkify(HtmlDomNodes,SpecialWords)
{
    for (CurNode in HtmlDomNodes) // Loop through each node.
    {
     if (CurNode.is('a')) continue; // if hyperlink, skip to next node in loop.

     CurNode.Text = addLinks( CurNode.Text , SpecialWords ); // perform the regex replacements to add the hyperlinks.

     linkify(CurNode.ChildNodes); // recursively call function until all text is replaced.
    }
}

function addLinks(Text,SpecialWords)
{
    for (CurWord in SpecialWords.split(','))
    {
        Text = preg_replace
            ( '/'&CurWord&'/i'
            , '<a class="dfn" href="glossary.php?term=\0">\0</a>'
            , Text
            );
    }
    return Text;
}

It's not PHP because I can't be bothered reading the documentation for a PHP DOM parser to work out the correct commands - you can do that bit. :)

As for which parser to use PHP Simple HTML DOM Parser is one option, but there are others if you don't get on with that.

Answer 2

preg_replace ( mixed $pattern , mixed $replacement , mixed $subject [, int $limit = -1 [, int &$count ]] )

Note that "Limit" is optional.

preg_replace("/text/i', '', $input, 1);

Or something like that ;)