Question

I’m wondering if it’s possible to extract a number of variables from a predefined sentence with regex or similar.

e.g.

If this was the pattern…

"How many * awards did * win in *?"

And someone typed…

"How many gold awards did johnny win in 2008?"

How can I somehow return…

["gold","johnny","2008"]

I’d also like to return the fact that it matches the pattern before retrieving the variables as there will be many different patterns. Note: It will also be possible for someone to type multiple words in place of a * e.g. johnny english instead of just johnny

Thanks

Answer 1

Building on Derek’s answer and SimpleCoder’s comment, here would be the full function:

// This function escapes a regex string
// http://simonwillison.net/2006/jan/20/escape/
function escapeRegex(text) {
    return text.replace(/[-[\]{}()*+?.,\\^$|#\s]/g, "\\$&");
}

function match(pattern, text) {
    var regex = '^' + escapeRegex(pattern).replace(/\\\*/g, '(.+)') + '$';
    var query = text.match(new RegExp(regex, 'i'));
    if (!query)
        return false;

    query.shift(); // remove first element
    return query;
}

match("How many * awards did * win in *?", "How many gold awards did johnny win in 2008?");