|
| 1 | +import argparse |
| 2 | +import os |
| 3 | +import requests |
| 4 | +import boto3 |
| 5 | + |
| 6 | +from concurrent.futures import ThreadPoolExecutor, as_completed |
| 7 | +from urllib.parse import urlparse |
| 8 | + |
| 9 | +def upload_url_to_s3(s3_client, bucket_name, url, object_key): |
| 10 | + response = requests.get(url, stream=True, verify=False, headers={"Cookie": "_c_t_c=1"}) |
| 11 | + response.raise_for_status() |
| 12 | + s3_client.upload_fileobj(response.raw, bucket_name, object_key) |
| 13 | + return url, object_key |
| 14 | + |
| 15 | +def derive_key_from_url(url, prefix=None): |
| 16 | + parsed = urlparse(url) |
| 17 | + filename = os.path.basename(parsed.path) |
| 18 | + if prefix: |
| 19 | + return f"{prefix}/{filename}" |
| 20 | + else: |
| 21 | + return filename |
| 22 | + |
| 23 | +def main(): |
| 24 | + parser = argparse.ArgumentParser(description="Upload multiple URLs from versions.txt to an S3 bucket concurrently.") |
| 25 | + parser.add_argument('--bucket', required=True, help="Name of the S3 bucket.") |
| 26 | + parser.add_argument('--concurrency', required=True, type=int, help="Number of concurrent uploads.") |
| 27 | + parser.add_argument('--versions-file', default='versions.txt', help="File containing one URL per line.") |
| 28 | + parser.add_argument('--access-key', required=True, help="AWS Access Key ID.") |
| 29 | + parser.add_argument('--secret-key', required=True, help="AWS Secret Access Key.") |
| 30 | + parser.add_argument('--endpoint-url', required=True, help="Custom endpoint URL for S3 or S3-compatible storage.") |
| 31 | + args = parser.parse_args() |
| 32 | + |
| 33 | + bucket_name = args.bucket |
| 34 | + concurrency = args.concurrency |
| 35 | + versions_file = args.versions_file |
| 36 | + access_key = args.access_key |
| 37 | + secret_key = args.secret_key |
| 38 | + endpoint_url = args.endpoint_url |
| 39 | + |
| 40 | + with open(versions_file, 'r') as f: |
| 41 | + urls = [line.strip() for line in f if line.strip()] |
| 42 | + |
| 43 | + print(f"Uploading {len(urls)} files...") |
| 44 | + s3_params = {} |
| 45 | + if access_key and secret_key: |
| 46 | + s3_params['aws_access_key_id'] = access_key |
| 47 | + s3_params['aws_secret_access_key'] = secret_key |
| 48 | + if endpoint_url: |
| 49 | + s3_params['endpoint_url'] = endpoint_url |
| 50 | + |
| 51 | + s3 = boto3.client('s3', **s3_params) |
| 52 | + |
| 53 | + futures = [] |
| 54 | + with ThreadPoolExecutor(max_workers=concurrency) as executor: |
| 55 | + for url in urls: |
| 56 | + object_key = derive_key_from_url(url) |
| 57 | + futures.append(executor.submit(upload_url_to_s3, s3, bucket_name, url, object_key)) |
| 58 | + for future in as_completed(futures): |
| 59 | + try: |
| 60 | + uploaded_url, uploaded_key = future.result() |
| 61 | + print(f"Uploaded: {uploaded_url} -> s3://{bucket_name}/{uploaded_key}") |
| 62 | + except Exception as e: |
| 63 | + print(f"Error uploading: {e}") |
| 64 | + |
| 65 | +if __name__ == "__main__": |
| 66 | + main() |
0 commit comments