def test_arrow_file_does_not_exist(self):
with tempfile_context() as path:
path.unlink()
with self.assertRaisesRegex(
InvalidArrowFile,
"arrow-validate: .*No such file or directory"):
validate_arrow_file(path)
def test_arrow_file_does_not_validate(self):
array = pyarrow.StringArray.from_buffers(
1,
# value_offsets: first item spans buffer offsets 0 to 1
pyarrow.py_buffer(struct.pack("II", 0, 1)),
# data: a not-UTF8-safe character
pyarrow.py_buffer(b"\xc9"),
)
with arrow_file({"A": array}) as path:
with self.assertRaisesRegex(
InvalidArrowFile,
"arrow-validate: --check-utf8 failed on column A"):
validate_arrow_file(path)
def test_arrow_file_does_not_validate(self):
array = pa.StringArray.from_buffers(
1,
# value_offsets: first item spans buffer offsets 0 to 1
pa.py_buffer(struct.pack("II", 0, 1)),
# data: a not-UTF8-safe character
pa.py_buffer(b"\xc9"),
)
table = pa.table({"A": array})
with tempfile_context() as path:
with pa.ipc.RecordBatchFileWriter(path, table.schema) as writer:
writer.write_table(table)
with self.assertRaisesRegex(InvalidArrowFile,
"arrow-validate: --check-safe failed"):
validate_arrow_file(path)
def test_arrow_file_does_not_open(self):
with tempfile_context() as path:
path.write_bytes(b"this is not an Arrow file")
with self.assertRaisesRegex(InvalidArrowFile,
"arrow-validate: .*Not an Arrow file"):
validate_arrow_file(path)
def test_happy_path(self):
with arrow_table_context(make_column("A", ["x"])) as (path, _):
validate_arrow_file(path) # do not raise
def test_happy_path(self):
with arrow_file({"A": [1, 2, 3]}) as path:
validate_arrow_file(path) # do not raise