正则表达式排除被特殊字符包围的单词

Question

I've been having issues with finding a solution to a regex conundrum I'm having.

Recently, I worked on a project where we needed to replace a list of words in a given text with a list of anchor tags.

For example, given a string

This is a test string

I may want to replace the word “test” with

<a target="_blank"  href="https://website.com/string-random“>test</a>. 

The resulting string should look like this

This is a <a target="_blank" href="https://website.com/string-random“>test</a> string

The replacement of the words is done in a loop

foreach ($documents as $document)
 
    foreach ($links as $link)
        replace keywords

What ends up happening in some scenarios is some of the urls in the anchor tags contain words that could potentially be replaced

For example, given this list of words to replace

[
    {
        'keyword': 'test',
        'link': 'https://website.com/string-random'
    },
    {
        'keyword': 'string',
        'link': 'https://random.com/string'
    }
]

After all the replacements are done, the sample string I gave above would look like this

This is a <a target="_blank" href="https://website.com/<a target="_blank"  href="https://random.com/string“>string</a>-random“>test</a> <a target="_blank" href="https://random.com/string“>string</a>

Instead of

This is a <a target="_blank" href="https://website.com/string-random“>test</a> <a target="_blank" href="https://random.com/string“>string</a>

Currently, I am looking for a regular expression that would not match on any words surrounded by special characters as I think this would solve my problem.

Also very open to any other ideas on how to tackle this problem

Answer 1

This is not just about the previous replacements: any word that occurs within tag attributes / names / values is an issue.

In other words, you want to replace strings that are followed some chars where next < occurs before next > (strings between tags and not within tags)

Hence try this one : (string-to-match)(?=[^>]*?<)

(replace string-to-match, obviously)

The other block is a lookahead : it ensures that you can read any char but >, as many times as needed, then a <

Answer 2

Try :

foreach ($wordlist as $word){
     $document = preg_replace("~(?! )($word[keyword])(?! )~i","<a href='$word[link]'>$1<")
}

Answer 3

I found a pattern that works pretty well for me hear $pattern = '/(?<!(>|\\/|-))\\b' . preg_quote($stringToReplace, '/') . '\\b(?!(<|\\/|-))/i'; $pattern = '/(?<!(>|\\/|-))\\b' . preg_quote($stringToReplace, '/') . '\\b(?!(<|\\/|-))/i';