Welcome to the MacNN Forums.

Ghoser777 · Jun 27, 2002, 12:51 AM

So here's my question. I need to take a text file and pare out all the words. Using regular expressions, that's pretty easy:

<blockquote>code:<hr /><pre style="font-size:x-small; font-family: monospace;">"[a-zA-Z-]+"</pre><hr /></blockquote>Now what I want to do is find every pair of words. So for the phrase: "I am a pretty little girl", I need to be able to get 'I am', 'am a', 'a pretty', 'pretty little', little girl'. I tried do this:

<blockquote>code:<hr /><pre style="font-size:x-small; font-family: monospace;">"[a-zA-Z-]+[^a-zA-Z-]+[a-zA-Z-]+"</pre><hr /></blockquote>This sort of works, except that I end up getting 'I am', 'a pretty', 'little girl'. Does anyone know how I can setup the regular expression correctly to get the result I want? It has something to do with rewind marks, but I'm not much of a regular expression buff.

Thanks!
F-bacher

smeger · Jun 27, 2002, 02:42 AM

I didn't think you could rewind regular expressions. BBEdit's got nice documentation for 'em in their help if you've got it. Or you could 'man regexp' in the terminal.

Mactoid · Jun 27, 2002, 03:44 AM

At the risk of sounding terribly ignorant, is this really necessairy?

I mean, once you have a list of the words, it's trivial to generate the word pairs. Maybe for some reason this won't work in your situation, but I'd wager that the code would be more maintenable than some messy regexp if you were able to do it.

That being said, I don't know how that would be done with just a regexp.

sadie · Jun 27, 2002, 04:11 AM

Isn't there an option for getting *everything* that matches the criteria? I forget the syntax, and don't even know what form of regex you're using, but i'm pretty sure Perl could do this.

Ghoser777 · Jun 27, 2002, 06:55 AM

I did do that for awhile (working with the list), but for large texts it can be somewhat expensive. I'll probably end up doing that again.

Thanks,
F-bacher