From: Robert Klemme <shortcutter@...>
Date: 2009-11-22T23:52:26+09:00
Subject: Re: How to match and count

On 22.11.2009 10:12, Ruby Newbee wrote:
> 2009/11/22 Phrogz <phrogz@mac.com>:
> 
>> # Don't read the whole file into memory, but do it one line at a time
>> i = 0
>> file = File.open( "foo.txt" )
>> file.each_line do |line|
>>  pieces = line.split( /\s+/ )
>>  i += 1 if pieces[ 3 ] =~ /something/
>> end

> I like that, thank you!

The code above does not close the file handle properly.  Also if can be 
done shorter:

File.foreach "file.txt" do |line|
   ...
end

You can even use Ruby like awk which seems to be rarely done - but it's 
possible.

> awk '{if ($4~/something/) {i+=1}} END {print i}'  file.txt

Can be done like

ruby -nae 'BEGIN {$i=0}; $i+=1 if /something/ =~ $F[3]; END {puts $i}' 
file.txt
ruby -nae 'BEGIN {$i=0}; /something/ =~ $F[3] and $i+=1; END {puts $i}' 
file.txt

For a script, I'd probably do something similar to what Phrogz suggested 
but with the difference that I'd use ARGF.  That way you fetch file 
names from the command line and do not need to change the script if the 
file name changes:

i = 0

ARGF.each do |line|
   bit = line.split(/\s+/)[3]
   i += 1 if /something/ =~ bit
end

puts i

Or, do the matching in one step which seems more efficient

i = 0

ARGF.each do |line|
   i += 1 if /^\s*(?:\S+\s+){3}something/ =~ line
end

puts i

There are about 2,843 million other ways to do it in Ruby.

Kind regards

	robert


-- 
remember.guy do |as, often| as.you_can - without end
http://blog.rubybestpractices.com/