From: Glenn <glenn_ritz@...>
Date: 2008-03-01T20:44:11+09:00
Subject: Re: Question about split method

--0-1441077914-1204371843=:23628
Content-Type: text/plain; charset=us-ascii

Thanks for all of the tips about this question.

I actually shortened the suggestion below to:

"a.b. c! d?? e.".scan(/.*?[.!?]+/)

This returns all of the sentences as elements of the array and includes the punctuation.  It works even if there are no spaces between the punctuation and the next sentence, and it works if there are 1 or more '.', '?', or '!':

["a.", "b.", " c!", " d??", " e."]

Thanks again,

Glenn

----- Original Message ----
From: Robert Klemme <shortcutter@googlemail.com>
To: ruby-talk ML <ruby-talk@ruby-lang.org>
Sent: Monday, February 25, 2008 7:43:45 AM
Subject: Re: Question about split method

2008/2/25, Glenn <glenn_ritz@yahoo.com>:
> Hello,
>
>  I'm wondering if there is a method for the String class that splits a string on some characters and keeps the split characters in the elements of the resulting array?
>
>  The split method returns an array in this example:
>
>  p "This is a sentence. This is a sentence! This is a sentence?".strip.split(/\.|\?|\!/)
>
>
>  ["This is a sentence", " This is a sentence", " This is a sentence"]
>
>  The three sentences in the above string have very different meanings, but loose those meanings without the punctuation, so I'd like to keep the punctuation.  I'd like a method that keeps the split characters, and returns this array:
>
>  ["This is a sentence.", " This is a sentence!", " This is a sentence?"]
>
>  Does such an array exist?  If not, would it be possible to modify the split  method to  produce that  result?
>
>  I'm running Ruby 1.8.6 on Windows.

Hm, you could do it with lookbehind on 1.9.  On 1.8 you only have
lookforward which gives you this:

irb(main):002:0> "a. b.".split /(?=\.\s+)/
=> ["a", ". b."]

Not quite what you wanted. :-)

But here's an alternative approach which works with 1.8:

irb(main):005:0> "a. b. c! d? e.".scan /.*?[.!?](?:\s|$)/
=> ["a. ", "b. ", "c! ", "d? ", "e."]

Kind regards

robert

-- 
use.inject do |as, often| as.you_can - without end





--0-1441077914-1204371843=:23628--