c4f3799d5713d467ca577a6d0593f242076c9540,tensorflow_datasets/core/test_utils.py,FeatureExpectationsTestCase,test_encode_decode,#FeatureExpectationsTestCase#,65

Before Change

    fdict = features.FeaturesDict(
        {exp.name: exp.feature for exp in expectations})

    decoded_sample = features_encode_decode(
        fdict, dict([(exp.name, exp.value) for exp in expectations]))

    for exp in expectations:
      self.assertAllEqual(decoded_sample[exp.name], exp.expected)
      // TODO(rsepassi): test shape and dtype against exp.feature

After Change

  def test_encode_decode(self):
    // Maybe should try to use metaclass instead and dynamically generate one
    // method per feature expectation.
    for exp in self.expectations:
      tf.logging.info("Testing feature %s", exp.name)

      // Check the shape/dtype
      self.assertEqual(exp.feature.shape, exp.shape)
      self.assertEqual(exp.feature.dtype, exp.dtype)

      // Check the serialized features
      if exp.serialized_features is not None:
        self.assertEqual(
            exp.serialized_features,
            exp.feature.get_serialized_features(),
        )

      // Create the feature dict
      fdict = features.FeaturesDict({exp.name: exp.feature})
      for test in exp.tests:
        input_value = {exp.name: test.value}

        if test.raise_cls is not None:
          if not test.raise_msg:
            raise ValueError(
                "test.raise_msg should be set with {}for test {}".format(
                    test.raise_cls, exp.name))
          with self.assertRaisesWithPredicateMatch(
              test.raise_cls, test.raise_msg):
            features_encode_decode(fdict, input_value)
        else:
          // Test the serialization only
          if test.expected_serialized is not None:
            self.assertEqual(
                test.expected_serialized,
                exp.feature.encode_sample(test.value),
            )

          // Test serialization + decoding from disk
          decoded_samples = features_encode_decode(fdict, input_value)
          self.assertAllEqual(test.expected, decoded_samples[exp.name])
          // TODO(rsepassi): test shape and dtype against exp.feature


def features_encode_decode(features_dict, sample):
  Runs the full pipeline: encode > write > tmp files > read > decode.
  // Encode sample
  encoded_sample = features_dict.encode_sample(sample)

Instances