#!/usr/bin/ruby

doc=<<"DOC"

DOC

require '~/Ruby/cmd_line.rb'

BIN='/Applications/tree-tagger-MacOSX-3.1/bin/'
CMD='/Applications/tree-tagger-MacOSX-3.1/tagger-scripts/cmd/'
LIB='/Applications/tree-tagger-MacOSX-3.1/tagger-scripts/lib/'

TOKENIZER=BIN+'separate-punctuation'
TAGGER=BIN+'tree-tagger'
ABBR_LIST=LIB+'german-abbreviations'
PARFILE=LIB+'german.par'
FILTER=CMD+'filter-german-tags'

# The +s switch seems to activate separation of punctuation.
# I'll bet +l activates abreviation processing.
# What does +1 do?

def make_prolog(line)
  word, tag, lemma = line.split
  "word('%s', '%s', '%s')," % [word, tag, lemma]
end

long_switches  = ['pro', 'tok', 'lemmas', 'words']
needs_switches = []
$cl = Command_line.new('h', doc, long_switches, needs_switches)
args=$cl.rest
args.each do |file_name|
  if $cl['tok']
    `#{TOKENIZER} +1 +s +l #{ABBR_LIST} #{file_name}`.each do |ln|
        puts ln
      end
  elsif $cl['pro']
    text=[]
    sentence=[]
    puts "#{TAGGER} #{PARFILE} #{file_name}"
    `#{TOKENIZER} +1 +s +l #{ABBR_LIST} #{file_name} | #{TAGGER} #{PARFILE} -token -lemma`.each do |ln|
      word, tag, lemma = ln.split
      sentence << ("word('%s', '%s', '%s')" % [word, lemma, tag])
      if word == '.'
        text << '['+sentence.join(",\n  ") + ']'
        sentence = []
      end
    end
    puts '[' + text.join(",\n ") +'].'
  else
    puts "#{TAGGER} #{PARFILE} #{file_name}"
    `#{TOKENIZER} +1 +s +l #{ABBR_LIST} #{file_name} | #{TAGGER} #{PARFILE} -token -lemma`.each do |ln|
      if $cl['lemmas']
        puts ln.split[2]
      elsif $cl['words']
        puts ln.split[0]
      else
        puts ln
      end
    end
  end
end