test_phashion.rb

require 'helper'
require 'sqlite3'
require 'tempfile'

class TestPhashion < Minitest::Test
  def test_text_hash
    matches = Tempfile.open('foo') do |f|
      100.times do |i|
        f.write "hello world #{i}"
      end
      f.close
      a = Phashion.texthash_for f.path
      b = Phashion.texthash_for f.path

      assert_operator a.length, :>, 0
      assert_operator b.length, :>, 0
      a.each { |hash| assert_instance_of Phashion::TextHashPoint, hash }
      b.each { |hash| assert_instance_of Phashion::TextHashPoint, hash }

      Phashion.textmatches_for(a, b)
    end
    assert_operator matches.length, :>, 0
    matches.each { |match| assert_instance_of Phashion::TextMatch, match }
  end

  def split(hash)
    r = hash & 0xFFFFFFFF
    l = (hash >> 32) & 0xFFFFFFFF
    [l, r]
  end

  def test_db_bad_arg
    db = SQLite3::Database.new ':memory:'
    return unless db.respond_to? :enable_load_extension

    db.enable_load_extension true
    db.load_extension Phashion.so_file

    res = db.execute "SELECT hamming_distance('foo', 'bar', 'baz', 'zot')"
    assert_equal [[0]], res
  end

  def test_db_extension
    db = SQLite3::Database.new ':memory:'
    return unless db.respond_to? :enable_load_extension

    db.enable_load_extension true
    db.load_extension Phashion.so_file

    db.execute <<-SQL
      CREATE TABLE "images" (
        "id" INTEGER PRIMARY KEY AUTOINCREMENT NOT NULL,
        "fingerprint_l" integer NOT NULL,
        "fingerprint_r" integer NOT NULL)
    SQL

    jpg = relative_path '/jpg/Broccoli_Super_Food.jpg'
    png = relative_path '/png/Broccoli_Super_Food.png'

    hash1 = Phashion.image_hash_for jpg
    hash2 = Phashion.image_hash_for png

    l, r = split hash1
    db.execute "INSERT INTO images (fingerprint_l, fingerprint_r) VALUES (#{l}, #{r})"

    expected = Phashion.hamming_distance hash1, hash2

    l, r = split hash2
    rows = db.execute "SELECT hamming_distance(fingerprint_l, fingerprint_r, #{l}, #{r}) FROM images"
    assert_equal expected, rows.first.first
  end

  def test_mh_hash_for
    jpg = relative_path '/jpg/Broccoli_Super_Food.jpg'
    png = relative_path '/png/Broccoli_Super_Food.png'

    hash1 = Phashion.mh_hash_for jpg
    hash2 = Phashion.mh_hash_for png

    assert_kind_of Array, hash1
    assert_kind_of Array, hash2
    assert_in_delta 0.100, Phashion.hamming_distance2(hash1, hash2), 0.033
  end

  def test_mh_distance_from
    jpg = relative_path '/jpg/Broccoli_Super_Food.jpg'
    png = relative_path '/png/Broccoli_Super_Food.png'

    assert_in_delta 0.100, Phashion::Image.new(jpg).mh_distance_from(Phashion::Image.new(png)), 0.033
  end

  def test_duplicate_detection
    files = %w[86x86-0a1e.jpeg 86x86-83d6.jpeg 86x86-a855.jpeg]
    images = files.map { |f| Phashion::Image.new("#{relative_path '/../test/jpg/'}#{f}") }
    assert_duplicate images[0], images[1]
    assert_duplicate images[1], images[2]
    assert_duplicate images[0], images[2]
  end

  def test_duplicate_detection_2
    files = %w[b32aade8c590e2d776c24f35868f0c7a588f51e1.jpeg df9cc82f5b32d7463f36620c61854fde9d939f7f.jpeg
               e7397898a7e395c2524978a5e64de0efabf08290.jpeg]
    images = files.map { |f| Phashion::Image.new("#{relative_path '/../test/jpg/'}#{f}") }
    assert_duplicate images[0], images[1]
    assert_duplicate images[1], images[2]
    assert_duplicate images[0], images[2]
  end

  def test_not_duplicate
    files = %w[86x86-0a1e.jpeg 86x86-83d6.jpeg 86x86-a855.jpeg avatar.jpg]
    images = files.map { |f| Phashion::Image.new("#{relative_path '/../test/jpg/'}#{f}") }
    assert_not_duplicate images[0], images[3]
    assert_not_duplicate images[1], images[3]
    assert_not_duplicate images[2], images[3]
  end

  def test_multiple_types
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    png = Phashion::Image.new(relative_path('/png/Broccoli_Super_Food.png'))
    gif = Phashion::Image.new(relative_path('/gif/Broccoli_Super_Food.gif'))
    assert_duplicate jpg, png
    assert_duplicate gif, png
    assert_duplicate jpg, gif
  end

  def test_fingerprint_png_is_different
    png1 = Phashion::Image.new(relative_path('/png/Broccoli_Super_Food.png'))
    png2 = Phashion::Image.new(relative_path('/png/linux.png'))
    png3 = Phashion::Image.new(relative_path('/png/grass.png'))
    png4 = Phashion::Image.new(relative_path('/png/Broccoli_Super_Food.png'))

    fingerprints = []
    fingerprints << png1.fingerprint
    fingerprints << png2.fingerprint
    fingerprints << png3.fingerprint
    fingerprints << png4.fingerprint

    assert fingerprints.uniq.size == 3, 'array should contain 3 unique fingerprints'
  end

  def test_duplicate_with_custom_distance_threshold
    # NOTE: this test depends on the smaller_jpg test still asserting a distance of 2
    # note-2: threshold is a :less-than-or-equal-to comparison, which is a change from version 1.0.8
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.100px.jpg'))

    refute(jpg.duplicate?(jpg_x, threshold: 1))
    assert(jpg.duplicate?(jpg_x, threshold: 2))
  end

  def test_duplicate_meta_methods
    # NOTE: this test depends on the smaller_jpg test still asserting a distance of 2
    # note-2: threshold is a :less-than-or-equal-to comparison, which is a change from version 1.0.8
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.100px.jpg'))

    refute(jpg.dupe_at_threshold_1?(jpg_x))
    assert(jpg.dupe_at_threshold_2?(jpg_x))
    assert_raises(NoMethodError) { jpg.dupe_at_threshold_100?(jpg_x) }
  end

  ### distance methods
  def test_distance_from_jpg_to_png_dupe
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    png = Phashion::Image.new(relative_path('/png/Broccoli_Super_Food.png'))

    assert_equal(jpg.distance_from(png), 0)
  end

  def test_distance_from_lossy_jpg
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.lossy.jpg'))

    assert_equal(jpg.distance_from(jpg_x), 0)
  end

  def test_distance_from_smaller_jpg
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.100px.jpg'))

    assert_equal(jpg.distance_from(jpg_x), 2)
  end

  def test_distance_from_color_correction
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.color-corrected.jpg'))

    assert_equal(jpg.distance_from(jpg_x), 2)
  end

  def test_distance_from_black_and_white
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.bw.jpg'))

    assert_equal(jpg.distance_from(jpg_x), 2)
  end

  def test_distance_from_bounding_box
    # Control-image is cropped to remove empty whitespace around image details
    # from 500x349 to 466x312
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.bounding-box.jpg'))

    assert_equal(jpg.distance_from(jpg_x), 12)
  end

  def test_distance_from_rotation_of_5degrees_c2
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.rotate5cw.jpg'))

    assert_equal(jpg.distance_from(jpg_x), 14)
  end

  def test_distance_from_horizontal_flip
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.horizontal-flip.jpg'))

    assert_operator(jpg.distance_from(jpg_x), :>, Phashion::DEFAULT_DUPE_THRESHOLD)
  end

  def test_duplicate_with_module_method
    jpg = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.jpg'))
    jpg_x = Phashion::Image.new(relative_path('/jpg/Broccoli_Super_Food.bw.jpg'))

    assert_duplicate_with_module_method(jpg, jpg_x)
  end

  private

  def relative_path(path)
    "#{File.dirname(__FILE__)}#{path}"
  end

  def assert_duplicate(a, b)
    assert a.duplicate?(b), "#{a.filename} not dupe of #{b.filename}"
  end

  def assert_not_duplicate(a, b)
    assert !a.duplicate?(b), "#{a.filename} dupe of #{b.filename}"
  end

  def assert_duplicate_with_module_method(a, b)
    assert Phashion.duplicate?(a.fingerprint, b.fingerprint), "#{a.filename} not dupe of #{b.filename}"
  end
end