gbif-cruft/src/parquet/__init__.py

25 lines
617 B
Python

# SPDX-FileCopyrightText: 2023-present Jeff Moe <moe@spacecruft.org>
#
# SPDX-License-Identifier: AGPL-3.0-or-later
import click
from gbif_cruft.__about__ import __version__
from gbif_cruft.lazy_group import LazyGroup
import pyarrow.parquet as pq
@click.command()
@click.argument(
'filename',
nargs=1,
default="/srv/gbif/gbif_2023-10-01/occurrence.parquet/000000",
type=click.File('rb'),
)
def cli(filename):
print("Reading", filename.name)
muh = pq.read_table(filename)
# print(muh)
parquet_file = pq.ParquetFile(filename)
muhmeta = parquet_file.metadata
print(muhmeta)