import argparse
import boto3
import botocore
import os


s3 = boto3.resource('s3')


def file_exists_s3(bucket, s3_path):
    try:
        bucket.Object(s3_path).load()
    except botocore.exceptions.ClientError as e:
        if e.response['Error']['Code'] == '404':
            return False
    return True


def export_file(bucket, s3_path, local_file_path, skip_existing=True,
                dryrun=True):
    if skip_existing:
        exists = file_exists_s3(bucket, s3_path)
    else:
        exists = False
    if exists:
        print("%s exists. Skipping" % s3_path)
    else:
        if dryrun:
            print("%s doesn't exist. Would have imported." % s3_path)
        else:
            print("%s doesn't exist. Importing." % s3_path)
            bucket.Object(s3_path).put(Body=open(local_file_path))


def main():
    parser = argparse.ArgumentParser(
        description='Import papertrail logs to s3')
    parser.add_argument('source_dir', type=str, help="The source directory")
    parser.add_argument('bucket', type=str, help="The s3 bucket")
    parser.add_argument('--bucket_folder', type=str, default='papertrail/logs',
                        help="The root folder on s3")
    parser.add_argument('--dryrun', default=False, action='store_true',
                        help="Do a dry run: don't write any data")
    parser.add_argument('--skip_existing', default=False, action='store_true',
                        help="Skip files that already exist in s3")
    args = parser.parse_args()
    bucket = s3.Bucket(args.bucket)
    for root, dirs, files in os.walk(args.source_dir):
        for file in files:
            parts = file.split('.')
            date = parts[0]
            s3_folder = "dt={}".format(date)
            local_file_path = os.path.join(root, file)
            s3_path = os.path.join(args.bucket_folder, s3_folder, file)
            export_file(bucket, s3_path, local_file_path,
                        skip_existing=args.skip_existing, dryrun=args.dryrun)


if __name__ == '__main__':
    main()