def test_datasetdict_from_json_splits(split, jsonl_path, tmp_path):
if split:
path = {split: jsonl_path}
else:
split = "train"
path = {"train": jsonl_path, "test": jsonl_path}
cache_dir = tmp_path / "cache"
expected_features = {
"col_1": "string",
"col_2": "int64",
"col_3": "float64"
}
dataset = JsonDatasetReader(path, cache_dir=cache_dir).read()
_check_json_datasetdict(dataset,
expected_features,
splits=list(path.keys()))
assert all(dataset[split].split == split for split in path.keys())
def test_dataset_from_json_features(features, jsonl_path, tmp_path):
cache_dir = tmp_path / "cache"
default_expected_features = {
"col_1": "string",
"col_2": "int64",
"col_3": "float64"
}
expected_features = features.copy(
) if features else default_expected_features
features = (Features({
feature: Value(dtype)
for feature, dtype in features.items()
}) if features is not None else None)
dataset = JsonDatasetReader(jsonl_path,
features=features,
cache_dir=cache_dir).read()
_check_json_dataset(dataset, expected_features)
def test_dataset_from_json_with_mismatched_features(jsonl_312_path, tmp_path):
# jsonl_312_path features are {"col_3": "float64", "col_1": "string", "col_2": "int64"}
features = {"col_2": "int64", "col_3": "float64", "col_1": "string"}
expected_features = features.copy()
features = (Features({
feature: Value(dtype)
for feature, dtype in features.items()
}) if features is not None else None)
cache_dir = tmp_path / "cache"
dataset = JsonDatasetReader(jsonl_312_path,
features=features,
cache_dir=cache_dir).read()
assert isinstance(dataset, Dataset)
assert dataset.num_rows == 2
assert dataset.num_columns == 3
assert dataset.column_names == ["col_2", "col_3", "col_1"]
for feature, expected_dtype in expected_features.items():
assert dataset.features[feature].dtype == expected_dtype
def test_json_dataset_reader(
path_type,
split,
features,
keep_in_memory,
jsonl_path,
tmp_path,
):
file_path = jsonl_path
field = None
if issubclass(path_type, str):
path = file_path
elif issubclass(path_type, list):
path = [file_path]
cache_dir = tmp_path / "cache"
expected_split = str(split) if split else "train"
default_expected_features = {
"col_1": "string",
"col_2": "int64",
"col_3": "float64"
}
expected_features = features.copy(
) if features else default_expected_features
features = Features(
{feature: Value(dtype)
for feature, dtype in features.items()}) if features else None
with assert_arrow_memory_increases(
) if keep_in_memory else assert_arrow_memory_doesnt_increase():
dataset = JsonDatasetReader(path,
split=split,
features=features,
cache_dir=cache_dir,
keep_in_memory=keep_in_memory,
field=field).read()
assert isinstance(dataset, Dataset)
assert dataset.num_rows == 4
assert dataset.num_columns == 3
assert dataset.column_names == ["col_1", "col_2", "col_3"]
assert dataset.split == expected_split
for feature, expected_dtype in expected_features.items():
assert dataset.features[feature].dtype == expected_dtype
def test_datasetdict_from_json_features(features, jsonl_path, tmp_path):
cache_dir = tmp_path / "cache"
# CSV file loses col_1 string dtype information: default now is "int64" instead of "string"
default_expected_features = {
"col_1": "string",
"col_2": "int64",
"col_3": "float64"
}
expected_features = features.copy(
) if features else default_expected_features
features = (Features({
feature: Value(dtype)
for feature, dtype in features.items()
}) if features is not None else None)
dataset = JsonDatasetReader({
"train": jsonl_path
},
features=features,
cache_dir=cache_dir).read()
_check_json_datasetdict(dataset, expected_features)
def test_json_datasetdict_reader(
split,
features,
keep_in_memory,
jsonl_path,
tmp_path,
):
file_path = jsonl_path
field = None
if split:
path = {split: file_path}
else:
split = "train"
path = {"train": file_path, "test": file_path}
cache_dir = tmp_path / "cache"
default_expected_features = {
"col_1": "string",
"col_2": "int64",
"col_3": "float64"
}
expected_features = features.copy(
) if features else default_expected_features
features = Features(
{feature: Value(dtype)
for feature, dtype in features.items()}) if features else None
with assert_arrow_memory_increases(
) if keep_in_memory else assert_arrow_memory_doesnt_increase():
dataset = JsonDatasetReader(path,
features=features,
cache_dir=cache_dir,
keep_in_memory=keep_in_memory,
field=field).read()
assert isinstance(dataset, DatasetDict)
dataset = dataset[split]
assert dataset.num_rows == 4
assert dataset.num_columns == 3
assert dataset.column_names == ["col_1", "col_2", "col_3"]
assert dataset.split == split
for feature, expected_dtype in expected_features.items():
assert dataset.features[feature].dtype == expected_dtype
def test_dataset_from_json_split(split, jsonl_path, tmp_path):
cache_dir = tmp_path / "cache"
expected_features = {"col_1": "string", "col_2": "int64", "col_3": "float64"}
dataset = JsonDatasetReader(jsonl_path, cache_dir=cache_dir, split=split).read()
_check_json_dataset(dataset, expected_features)
assert dataset.split == str(split) if split else "train"
def test_dataset_from_json_keep_in_memory(keep_in_memory, jsonl_path, tmp_path):
cache_dir = tmp_path / "cache"
expected_features = {"col_1": "string", "col_2": "int64", "col_3": "float64"}
with assert_arrow_memory_increases() if keep_in_memory else assert_arrow_memory_doesnt_increase():
dataset = JsonDatasetReader(jsonl_path, cache_dir=cache_dir, keep_in_memory=keep_in_memory).read()
_check_json_dataset(dataset, expected_features)