From: Charles Hixson <charleshixsn@...>
Date: 2004-09-20T06:03:53+09:00
Subject: Re: Method improvement request .--

James Edward Gray II wrote:

> On Sep 18, 2004, at 6:14 PM, Charles Hixson wrote:
>
>> James Edward Gray II wrote:
>>
>>> On Sep 18, 2004, at 5:04 PM, James Edward Gray II wrote:
>>>
>>>> What about something like:
>>>>
>>>> def parse1(chunk)
>>>>     if chunk =~ / ^([^-A-Za-z0-9]*) # pre-match
>>>>                (.*) # middle
>>>
>>>
>>>
>>> Sorry, that probably needs to be:
>>>
>>> (.*?)
>>>
>>> My bad.
>>>
>>> James Edward Gray II
>>
>>
>> That would only work if the postmatch pattern were included in the 
>> same pattern as the prematch, thus:
>> if chunk =~ /^([^-A-Za-z0-9.]*)(.*?)[^-A-Za-z0-9.]*)$/
>
>
> I did include all three in the same pattern, I just used the 
> whitespace and comment modifier to pretty it up a bit.
>
> /^([^-A-Za-z0-9.]*)(.*?)([^-A-Za-z0-9.]*)$/
>
> Is the same as:
>
> /
>     ^([^-A-Za-z0-9.]*)    # pre-match
>     (.*?)                # middle
>     ([^-A-Za-z0-9.]*)$    # post-match
> /x
>
> Note the trailing /x.
>
> James Edward Gray II

Sorry.  It took me a bit of digging to find the /.../x documentation 
even after you explicitly pointed it out to me.  (This won't work for 
me, because my actual pre- and post- patterns also exclude spaces, but 
it can certainly clarify the example, if one understands it!)