pyinfra/scripts/manage_minio.py

import argparse
import gzip
import os
from pathlib import Path

from tqdm import tqdm

from pyinfra.config import CONFIG
from pyinfra.storage.storages import get_s3_storage


def parse_args():
    parser = argparse.ArgumentParser()

    subparsers = parser.add_subparsers(help="sub-command help", dest="command")

    parser_add = subparsers.add_parser("add", help="Add file(s) to the MinIO store")
    parser_add.add_argument("dossier_id")
    add_group = parser_add.add_mutually_exclusive_group(required=True)
    add_group.add_argument("--file", "-f")
    add_group.add_argument("--directory", "-d")

    subparsers.add_parser("purge", help="Delete all files and buckets in the MinIO store")

    args = parser.parse_args()
    return args


def add_file_compressed(storage, bucket_name, dossier_id, path) -> None:

    path_gz = combine_dossier_id_and_file_id_and_extension(dossier_id, Path(path).stem, ".ORIGIN.pdf.gz")

    with open(path, "rb") as f:
        data = gzip.compress(f.read())
        storage.put_object(bucket_name, path_gz, data)


if __name__ == "__main__":

    storage = get_s3_storage()
    bucket_name = CONFIG.storage.bucket

    if not storage.has_bucket(bucket_name):
        storage.make_bucket(bucket_name)

    args = parse_args()

    if args.command == "add":

        if args.file:
            add_file_compressed(storage, bucket_name, args.dossier_id, args.file)

        elif args.directory:
            for fname in tqdm([*os.listdir(args.directory)], desc="Adding files"):
                path = Path(args.directory) / fname
                add_file_compressed(storage, bucket_name, args.dossier_id, path)

    elif args.command == "purge":
        storage.clear_bucket(bucket_name)


def combine_dossier_id_and_file_id_and_extension(dossier_id, file_id, extension):
    return f"{dossier_id}/{file_id}{extension}"


def upload_compressed_response(storage, bucket_name, dossier_id, file_id, result) -> None:
    data = gzip.compress(result.encode())
    path_gz = combine_dossier_id_and_file_id_and_extension(dossier_id, file_id, CONFIG.service.response.extension)
    storage.put_object(bucket_name, path_gz, data)