From: Brian Candler <b.candler@...>
Date: 2008-09-18T23:11:50+09:00
Subject: Re: Suggestion for string parsing

> what I would like to avoid is to write something like this 
> (that works)
> 
> line =~ /(\w*)=([0-9A-Za-z -.:]*),([0-9A-Za-z -.:]*),([0-9A-Za-z 
> -.:]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*),([0-9]*)/
> 
> I wanted to this in one line but in a more mainable way, otherwise I 
> could always pars the string char by char.

If you don't actually need to match the data against a pattern, then 
just use

  line.split(',')

If you only want to proceed if the line is "valid", then write a 
suitable regexp pattern to validate it. There are plenty of shortcuts. 
For example, \d is the same as [0-9]. {n} means repeat the preceeding 
element exactly n times. So:

  case line
  when /^(\w*)=([^,]*),(\d+(,\d+){9})$/
    key1 = $1
    key2 = $2
    numbers = $3.split(/,/).collect { |n| n.to_i }
    # or: numbers = $3.scanf("%d %d %d %d %d %d %d %d %d %d") if you 
prefer
  else
    puts "Invalid line!"
  end

That matches word=string,n,n,n,n,n,n,n,n,n,n

Furthermore you can substitute patterns you use repeatedly:

  WORD = "[0-9A-Za-z -.:]*"
  ...
  when /^(#{WORD})=(#{WORD}),(#{WORD}),(\d+(,\d+){9})$/o

(//o means that the regexp is built only once, the substitutions aren't 
done every time round)

You can also use extended syntax to make the RE more maintainable:

  VALID_LINE = %r{ ^
    (\w*) =         # key ($1)
    (#{WORD}),      # format ($2)
    (\d+),          # size ($3)
    (\d+)           # sample rate ($4)
  $ }x

  if VALID_LINE =~ line
    ..
  end

You can also do groupings which *don't* capture data using (?:  ..  )

Compact enough?
-- 
Posted via http://www.ruby-forum.com/.