From: Robert Klemme <bob.news@...>
Date: 2003-09-18T17:53:26+09:00
Subject: Re: Why IO#readlines does'nt accept a Regexp?


"gabriele renzi" <surrender_it@remove.yahoo.it> schrieb im Newsbeitrag
news:uroimv475apt4pn9bqqtp2aa1s7ul4ngui@4ax.com...
> as in the subject, I just noticed that readlines just accepts a string
> as line Separator, and I wonder why it works this way.
> Some explanations?

Just a guess: normally it's not necessary and another reason might be
performance, since the overhead of a regexp might be significant for large
files.

However, you can simulate it if you read a complete file into a string and
then split with a regexp.

> BTW, if I want to read a file in a array of 'words' I have to do :
>
>
> File.new('myfile').gets(nil).split
>
> no better way ?

For large Files this is more efficient:

words=[]
IO.foreach("myfile") do |line|
  words.push( *line.scan( /\w+/oi ) )
end

If you have many repeating words you can save even more mem:

cache = Hash.new {|h,k| h[k]=k}
words = []

IO.foreach("myfile") do |line|
  words.push( *( line.scan( /\w+/oi ).map {|w| cache[w]} ) )
end

> on a sidenote, what are the efficiency issue related to the use of
> IO#each vs IO#foreach(anIO) vs a simple 'while line=gets..'  ?

Try ruby -profile with each method and see what happens.  I'd guess that
there is not much difference.

Regards

    robert