sample: 'hello world' parser (Ruby/Racc)

hello_parser.ry

# -*- coding : UTF-8 -*-

# 
# to generate 'hello_parser.rb'
# 
#     $ racc -o hello_parser.rb hello_parser.ry
# 
# to execute
# 
#     $ ruby hello_parser.rb <<EOT
#     hello world                     #=> OK(1)
#     EOT
# 
#     $ ruby hello_parser.rb <<EOT
#     Hello, World                    #=> OK(2)
#     EOT
# 
#     $ ruby hello_parser.rb <<EOT
#     Hello: World                    # Racc::ParseError
#     EOT
# 

class HelloParser
rule
  message : HELLO WORLD EOL           { puts "OK(1)" ; result = val.join }
          | HELLO ',' WORLD EOL       { puts "OK(2)" ; result = val.join }
end

---- header
require 'strscan'

---- inner
  attr_accessor :yydebug

  def parse(text)
    s = StringScanner.new text

    tokens = []
    case
      when s.scan(/hello\b/i) ; tokens << [:HELLO,    s.matched]
      when s.scan(/world\b/i) ; tokens << [:WORLD,    s.matched]
      when s.scan(/\n/)       ; tokens << [:EOL,      s.matched]
      when s.scan(/\s/)       ; # ignore white space
      when s.scan(/./m)       ; tokens << [s.matched, s.matched]
    end until s.eos?

    define_singleton_method(:next_token) { tokens.shift }

    do_parse
  end

---- footer

if __FILE__ == $0
  require 'optparse'

  parser = HelloParser.new

  ARGV.options do |opt|
    opt.on('-d', '--debug') { parser.yydebug = true }
    opt.parse!
  end

  parser.parse ARGF.read
end

# vi:set ts=2 sw=2 et:

hello_parser2.ry

# -*- coding : UTF-8 -*-

#
# to generate 'hello_parser2.rb'
#
# $ racc -o hello_parser2.rb hello_parser2.ry
#
# to execute
#
# $ ruby hello_parser2.rb <<EOT
# hello world #=> OK(1)
# EOT
#
# $ ruby hello_parser2.rb <<EOT
# Hello, World #=> OK(2)
# EOT
#
# $ ruby hello_parser2.rb <<EOT
# Hello: World # Racc::ParseError
# EOT
#

class HelloParser
rule
  message     :
              | message helloworld      { result = val.join }

  helloworld  : HELLO WORLD EOL         { puts "OK(1)" ; result = val.join }
              | HELLO ',' WORLD EOL     { puts "OK(2)" ; result = val.join }
end

---- header
require 'strscan'

---- inner
  attr_accessor :yydebug

  def parse(text="", &source)
    @source = source || proc {}
    @s = StringScanner.new text

    do_parse
  end

  class SkipToken < RuntimeError ; end

  def next_token
    @s << (@source.call or raise EOFError) if @s.eos?

    case
      when @s.scan(/hello\b/i)  ; [:HELLO,     @s.matched]
      when @s.scan(/world\b/i)  ; [:WORLD,     @s.matched]
      when @s.scan(/\n/)        ; [:EOL,       @s.matched]
      when @s.scan(/\s/)        ; raise SkipToken
      when @s.scan(/./m)        ; [@s.matched, @s.matched]
    end
  rescue SkipToken
    retry
  rescue EOFError
    nil
  end

---- footer

if __FILE__ == $0
  require 'optparse'

  parser = HelloParser.new

  ARGV.options do |opt|
    opt.on('-d', '--debug') { parser.yydebug = true }
    opt.parse!
  end

  parser.parse { ARGF.gets }
end

# vi:set ts=2 sw=2 et:

Makefile

#!/usr/bin/make -f
# -*- coding: UTF-8 -*-
#
# to generate 'foo.rb' from 'foo.ry'
#
#    ex.1) $ make foo.rb
#
#    ex.2) $ make FLAGS=-C foo.rb       # syntax check only
#

                                        # see also 'racc --help'
RACCFLAGS   += --debug                  # -t
#RACCFLAGS  += --verbose                # -v
#RACCFLAGS  += --log-file=filename      # -Ofilename
#RACCFLAGS  += --execute=ruby           # -erubypath
#RACCFLAGS  += --embedded               # -E
#RACCFLAGS  += --line-convert-all
#RACCFLAGS  += --no-line-convert        # -l
#RACCFLAGS  += --no-omit-actions        # -a
#RACCFLAGS  += --superclass=CLASSNAME
#RACCFLAGS  += --runtime=FEATURE
#RACCFLAGS  += --check-only             # -C
#RACCFLAGS  += --output-status          # -S
#RACCFLAGS  += --output-status          # -S
#RACCFLAGS  += -P

RACCFLAGS   += ${FLAGS}

%.rb: %.ry
    racc ${RACCFLAGS} -o $@ $<

.PHONY : all

#
# to generate 'hello_parser.rb'
#
#    ex.1) $ make
#
#    ex.2) $ make FLAGS=-C      # syntax check only
#

all : hello_parser.rb

# vi:set ts=4 sw=4: