require 'test/unit'
require 'spec_id'


class SpecIDTest < Test::Unit::TestCase

  def initialize(arg)
    super(arg)
    @tfiles = File.dirname(__FILE__) + '/tfiles/'
    @tfiles_l = File.dirname(__FILE__) + '/tfiles_large/'
    @bw = @tfiles + "bioworks_small.xml"
    @old_prot_proph = @tfiles + 'yeast_gly_small-prot.xml'
    @prot_proph = @tfiles + 'opd1/000_020_3prots-prot.xml'
    @srf = @tfiles_l + '7MIX_STD_110802_1.srf'
  end

  def test_spec_id_creation
    sp = SpecID.new(@bw)
    assert_equal(106, sp.prots.size)
  end

  def test_classify_by_false_flag
    file = @tfiles + "bioworks_with_INV_small.xml"
    sp = SpecID.new(file)
    assert_equal(19, sp.prots.size)
    (tp, fp) = sp.classify_by_false_flag(:prots, "INV_", true, true)
    assert_equal(4, fp.size, "num false pos")
    assert_equal(15, tp.size, "num true pos")
  end

  
  def test_precision
    require 'roc'
    file = @tfiles + "bioworks_with_INV_small.xml"
    # 4 INV and 15 non-inv for 19 total prots
    answ = %w( t t t t t t t t t t F t t t t F t F F )
    index = 0
    answ.collect! do |bool|
      bo = false
      if bool == 't'; bo = true end
      index += 1 
      write_index = index
      ## in the bioworks_with_INV_small.xml, protein 8 and 9 have the same
      ## probability as protein 7
      if write_index == 8 || write_index == 9
        write_index = 7
      end
      [write_index, bo]
    end
    roc = ROC.new
    tp, fp = ROC.new.prep_list(answ)
    (exp_tp, exp_fp) = roc.tps_and_ppv(tp, fp)

    sp = SpecID.new(file)
    assert_equal(19, sp.prots.size)
    tp, fp = sp.rank_and_classify(:prots, proc {|prt| prt.probability }, proc {|prt| if prt.reference =~ /^INV_/ ; false; else; true; end })
    (tps, ys) = roc.tps_and_ppv(tp, fp)
    assert_equal(exp_tp, tps)
    assert_equal(exp_fp, ys)
    (num_hits, prec) = sp.num_hits_and_ppv_for_prob("INV_", true)
    # @TODO: assert these guys for consistencies sake:
    assert_in_delta_arrays([1, 2, 3, 4, 5, 6, 9, 10, 11, 12, 13, 14, 15], tps, 0.0000001)
    # Consistency check only:
    assert_in_delta_arrays([1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.909090909090909, 0.916666666666667, 0.923076923076923, 0.928571428571429, 0.866666666666667], prec, 0.0000001)
  end

  def assert_in_delta_arrays(one, two, delta, message=nil)
    one.each_with_index do |v,i|
      assert_in_delta(v, two[i], delta, message)
    end
  end

  def test_file_type
    assert_equal('bioworks', SpecID.file_type(@bw))
    assert_equal('protproph', SpecID.file_type(@prot_proph))
    assert_equal('srg', SpecID.file_type('whatever.srg'))
    ## WOULD BE NICE TO GET THIS WORKING, TOO
    # assert_equal('protproph', SpecID.file_type(@old_prot_proph))
    if File.exist? @tfiles_l
      assert File.exist?(@srf), "file #{@srf} is there"
      assert_equal('srf', SpecID.file_type(@srf))
    else
      assert_nil( puts("\n--SKIPPING TEST-- (missing dir: #{@tfiles_l})") )
    end
  end

  def test_non_standard_aa_removal
    hash = {"K.PEPTIDE.Z" => "K.PEPTIDE.Z", "K.*M" => "K.M", "aI" => 'I', "YI.&" => "YI.", "EI.!@#\$%^&*(){}[]|\\;:'\"<>,?/EI" => 'EI.EI'}
    cl = proc {|v| SpecID::Pep.remove_non_amino_acids(v) }
    hash.each do |k,v|
      assert_equal(v, cl.call(k))
    end
  end



end

class MyProt ; include SpecID::Prot ; end
class MyPep ; include SpecID::Pep ; attr_accessor :xcorr end


class TestOccamsRazor < Test::Unit::TestCase

  def test_small
    
    prots = (0..6).to_a.map do |n|
      prot = MyProt.new
      prot.reference = "ref_#{n}"
      prot
    end

    peps = (0..12).to_a.map {|v| MyPep.new }

    #           0   1   2   3   4   5   6   7   8   9   10  11    12 
    aaseqs = %w(AAA BBB CCC ABC AAA BBB CCC ABC DDD EEE FFF EEEEE DDD)
    xcorrs = [1.0, 2.0, 3.0, 4.0, 1.0, 2.0, 3.0, 4.0, 0.5, 0.6, 0.7, 0.8, 0.5]

    peps.zip(aaseqs, xcorrs) do |pep,aaseq,xcorr|
      pep.aaseq = aaseq
      pep.xcorr = xcorr
    end

    prots[0].peps = peps[0,4]
    prots[1].peps = [peps[2]]  ## should be missing

    test_prots = prots[0,2]
    answ = SpecID.occams_razor(test_prots)
    answ.each do |an|
      assert( an[0].is_a?(SpecID::Prot), "prots are there")
    end
    first = answ.first
    assert_equal( prots[0], first[0])
    assert_equal_array_content( prots[0].peps, first[1])

    require 'pp'
    #pp answ


    prots[0].peps = peps[0,4]
    prots[1].peps = [peps[2]]  ## should be missing
    prots[2].peps = []  ## should be missing

    answ = SpecID.occams_razor(test_prots, true)
    #pp answ


    #prots[2].peps = [peps[2]]
    #prots[2].peps.push( peps[3] ) ## should be there since it has 2
    #prots[3].peps = [peps[3]] ## should be missing
  end

  def assert_equal_array_content(exp1, ans, message='')
    exp1.each do |item|
      assert(ans.include?(item), "finding #{item}: #{message}")
    end
  end
end


require 'fasta'

class TestProteinGroups < Test::Unit::TestCase

  def test_small
    prots = []

    aaseq = ('A'..'Z').to_a.join('')
    header = "prot1"
    prots << Fasta::Prot.new(header, aaseq)

    aaseq = ('A'..'Z').to_a.reverse.join('')
    header = "prot1_reverse"
    prots << Fasta::Prot.new(header, aaseq)

    aaseq = ('A'..'Z').to_a.join('')
    header = "prot1_identical"
    prots << Fasta::Prot.new(header, aaseq)

    aaseq = ('A'..'E').to_a.join('')
    header = "prot1_short"
    prots << Fasta::Prot.new(header, aaseq)

    aaseq = ('A'..'E').to_a.reverse.join('')
    header = "prot1_reverse_short"
    prots << Fasta::Prot.new(header, aaseq)

    fasta = Fasta.new(prots)

    pep_seqs = %w(ABCD DEFG ABCD DEFG EDCB FEDCB XYZ RANDOM AEABA) 

    arr = SpecID::Pep.protein_groups_by_sequence(pep_seqs, fasta)
    
    exp = [[prots[0], prots[2], prots[3]], [prots[0], prots[2]], [prots[0], prots[2], prots[3]], [prots[0],prots[2]], [prots[1], prots[4]], [prots[1]], [prots[0], prots[2]], [], []]
    
    assert_equal(exp, arr)
  end

end