From: Robert Klemme <shortcutter@...>
Date: 2009-04-24T01:45:03+09:00
Subject: Re: Write output to multiple files

On 23.04.2009 16:20, Milo Thurston wrote:
> I have encountered something I thought would be trivial but I can't
> quite get it to work properly. I have a load of data that is stored in
> various locations and need to create several files, one for each
> location, each containing the data in that location. Getting an array of
> objects containing the file name and location is easy, so I was thinking
> along the lines of this pseudocode:
> 
> data.each do |d|
>   output_filename = "#{d.location}.txt"
>   if !File.exists?(output_filename)
>     # create it and write d.file_name to it
>   else
>     # simply write d.file_name to it
>   end
> end
> # close all open files
> 
> The locations of the data are numerical ids so I could expect filenames
> like 1.txt, 2.txt and so on, and I have also sorted them before
> processing so that they could be opened and filled sequentially. But, it
> is not clear to me how to properly manage the file handles so any
> suggestions would be welcome.

If I understood you properly:

# untested
last_name = nil
io = nil
begin
   data.sort_by {|d| d.location}.each do |d|
     output_filename = "#{d.location}.txt"

     unless output_filename == last_name
       io.close rescue nil
       io = File.open(output_filename, "a")
     end

     io.puts d.file_name
ensure
   io.close rescue nil
end


If you want to keep 'data' unsorted you might want to store IO objects 
in a Hash, e.g.

files = Hash.new {|h,k| h[k] = File.open(k, "a")}
begin
   data.each do |d|
     output_filename = "#{d.location}.txt"
     files[output_filename].puts d.file_name
   end
ensure
   files.each do |k,io|
     io.close rescue nil
   end
end

Kind regards

	robert


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