elaunira-r2index

Python library for uploading and downloading files to/from Cloudflare R2 with the r2index API.

Installation

pip install elaunira-r2index

Usage

Sync Client

from elaunira.r2index import R2IndexClient

client = R2IndexClient(
    index_api_url="https://r2index.example.com",
    index_api_token="your-bearer-token",
    r2_access_key_id="your-r2-access-key-id",
    r2_secret_access_key="your-r2-secret-access-key",
    r2_endpoint_url="https://your-account-id.r2.cloudflarestorage.com",
)

# Upload and register a file
# extension is auto-detected from destination_filename
# media_type is auto-detected from file content
record = client.upload(
    bucket="my-bucket",
    source="./myfile.zip",
    category="software",
    entity="myapp",
    destination_path="/releases/myapp",
    destination_filename="myapp.zip",
    destination_version="v1",
    tags=["release", "stable"],
    create_checksum_files=True,  # Creates .md5, .sha1, .sha256, .sha512 files
)

# Download a file and record the download
# IP address is auto-detected, user agent defaults to "elaunira-r2index/<version>"
path, record = client.download(
    bucket="my-bucket",
    source_path="/releases/myapp",
    source_filename="myapp.zip",
    source_version="v1",
    destination="./downloads/myfile.zip",
    verify_checksum=True,  # Verify SHA-256 checksum after download
)

Async Client

from elaunira.r2index import AsyncR2IndexClient

async with AsyncR2IndexClient(
    index_api_url="https://r2index.example.com",
    index_api_token="your-bearer-token",
    r2_access_key_id="your-r2-access-key-id",
    r2_secret_access_key="your-r2-secret-access-key",
    r2_endpoint_url="https://your-account-id.r2.cloudflarestorage.com",
) as client:
    # Upload (extension and media_type are auto-detected)
    record = await client.upload(
        bucket="my-bucket",
        source="./myfile.zip",
        category="software",
        entity="myapp",
        destination_path="/releases/myapp",
        destination_filename="myapp.zip",
        destination_version="v1",
        tags=["release", "stable"],
    )

    # Download
    path, record = await client.download(
        bucket="my-bucket",
        source_path="/releases/myapp",
        source_filename="myapp.zip",
        source_version="v1",
        destination="./downloads/myfile.zip",
    )

Transfer Configuration

Control multipart transfer settings with R2TransferConfig for both uploads and downloads:

from elaunira.r2index import R2IndexClient, R2TransferConfig

client = R2IndexClient(
    index_api_url="https://r2index.example.com",
    index_api_token="your-bearer-token",
    r2_access_key_id="your-r2-access-key-id",
    r2_secret_access_key="your-r2-secret-access-key",
    r2_endpoint_url="https://your-account-id.r2.cloudflarestorage.com",
)

# Custom transfer settings
transfer_config = R2TransferConfig(
    multipart_threshold=100 * 1024 * 1024,  # 100MB (default)
    multipart_chunksize=32 * 1024 * 1024,   # 32MB chunks
    max_concurrency=64,                      # 64 parallel threads
    use_threads=True,                        # Enable threading (default)
)

# Use with upload
record = client.upload(
    bucket="my-bucket",
    source="./largefile.zip",
    category="data",
    entity="archive",
    destination_path="/data/files",
    destination_filename="largefile.zip",
    destination_version="v1",
    transfer_config=transfer_config,
)

# Use with download
path, record = client.download(
    bucket="my-bucket",
    source_path="/data/files",
    source_filename="largefile.zip",
    source_version="v1",
    destination="./downloads/largefile.zip",
    transfer_config=transfer_config,
)

Default max_concurrency is 2x the number of CPU cores (minimum 4).

Progress Tracking

Uploads and downloads log progress to the elaunira.r2index.storage (or elaunira.r2index.async_storage) logger every 10 seconds by default, showing bytes transferred, percentage, and speed:

Uploading: 150.00 MB / 500.00 MB (30.0%) — 25.00 MB/s
Downloading: 1.20 GB / 3.50 GB (34.3%) — 45.00 MB/s

Control the interval or disable it:

# Log every 30 seconds
path, record = client.download(
    ...,
    progress_interval=30.0,
)

# Disable progress logging
path, record = client.download(
    ...,
    progress_interval=None,
)

You can also provide a custom callback for programmatic progress tracking:

def on_progress(bytes_transferred: int) -> None:
    print(f"Downloaded: {bytes_transferred / 1024 / 1024:.1f} MB")

path, record = client.download(
    bucket="my-bucket",
    source_path="/releases/myapp",
    source_filename="myapp.zip",
    source_version="v1",
    destination="./downloads/myfile.zip",
    progress_callback=on_progress,
)

Skip Existing Downloads

Skip downloading if the destination file already exists:

path, record = client.download(
    bucket="my-bucket",
    source_path="/releases/myapp",
    source_filename="myapp.zip",
    source_version="v1",
    destination="./downloads/myfile.zip",
    overwrite=False,
)

Deleting Files

# Delete from R2 storage (including checksum sidecar files if they exist)
client.delete_from_r2(
    bucket="my-bucket",
    path="/releases/myapp",
    filename="myapp.zip",
    version="v1",
    delete_checksum_files=True,  # Also delete .md5, .sha1, .sha256, .sha512 files
)

# Delete from index (metadata only)
client.delete(file_id)

# Or delete by remote tuple
client.delete_by_tuple(RemoteTuple(
    bucket="my-bucket",
    remote_path="/releases/myapp",
    remote_filename="myapp.zip",
    remote_version="v1",
))

License

MIT