From: Przemyslaw Daniel <przemyslaw_daniel@...>
Date: 2007-07-17T00:21:25+09:00
Subject: Re: awk regexp search

small update to make it easier

# AWK implementation

module AWK
  class ClassAwk
      def initialize(filename = "")
        @NR = 0;            # record number
        @NF = 0;            # field number
        @FS = /\s+/;        # field separator
        @f = [];            # fields of macthed line, f[0] - line
        @trace = [];        # regexp trace

        #input file
        @file = filename == "" ? ARGF: File.open(filename, "r");
      end

      #NOTE: every rule has to be in separate line
      #to get unique rule id
      def rule(regexp1, regexp2 = nil)
        msg = "regexp parameter must be Regexp";
        raise ArgumentError, msg unless regexp1.kind_of?(Regexp);

        if regexp2 == nil
          yield if @f[0] =~ regexp1;
        else
          raise ArgumentError, msg unless regexp2.kind_of?(Regexp);
          rule_id = /.+:([0-9]+)/.match(caller.first.to_s)[1].to_i;

          @trace[rule_id] = true if @f[0] =~ regexp1;
          yield if @trace[rule_id]
          @trace[rule_id] = false if @f[0] =~ regexp2;
        end
      end

      def analyze()
        @NR = 0;
        @file.each { |line|
          @NR += 1;
          @f = line.split(@FS)
          @NF = @f.length
          @f.unshift(line.chop);
          yield
        }
      end

      attr_reader :NR, :NF, :f;
   end
end

example:

require "awk.rb"

awk = AWK::ClassAwk.new();

awk.analyze() {
  awk.rule(/start1/, /stop1/) {
    print "1, NR:#{awk.NR}, NF:#{awk.NF}, #{awk.f[0]}\n";
  };
  awk.rule(/start2/, /stop2/) {
    print "2, NR:#{awk.NR}, NF:#{awk.NF}, #{awk.f[0]}\n";
  };
  awk.rule(/start1/) {
    print "3, NR:#{awk.NR}, NF:#{awk.NF}, #{awk.f[0]}\n";
  };
}



-- 
Posted via http://www.ruby-forum.com/.