Resource Bundler

A system for managing resource bundles consisting collections of files and directories associated with a build. The system consists of:

• A server that stores files locally, deduplicated by content.
• A CLI that allows users to create, list, and download bundles.

The implementation prioritizes clarity, correctness, and extensibility over complexity.

Features:

• Create bundles from files or directories.
• Deduplicate files across bundles using SHA-256.
• Persist bundle metadata locally.
• Download bundles as tar.gz archives.
• Minimal dependencies and configuration.
• Project Structure:

src/main/java/com/pcisha/resourcebundler
├── server   # Spring Boot server
└── cli      # Command-line client

• Runtime data (created automatically):

data/
├── files/        # Deduplicated file blobs (by hash)
└── bundles.json  # Bundle metadata

• Build: Requires Java 21.

./mvnw clean package

• Run Server:

java -jar target/resourcebundler-1.0.0.jar server

• The server listens on:

http://localhost:8080

CLI Usage:

• Create a bundle:

java -jar target/resourcebundler-1.0.0.jar bundle create <path_to_files_or_directories>

• <path_to_files_or_directories> may be a file or directory.
• Directories are traversed recursively.
• Files already uploaded in previous bundles are not re-stored.

Example:

java -jar target/resourcebundler-1.0.0.jar bundle create src/main/resources/small

java -jar target/resourcebundler-1.0.0.jar bundle create src/main/resources/nested

java -jar target/resourcebundler-1.0.0.jar bundle create src/main/resources/duplicate

• List bundles:

java -jar target/resourcebundler-1.0.0.jar bundle list

Outputs bundle metadata including: bundle ID, number of files, total size, creation timestamp.

• Download a bundle:

java -jar target/resourcebundler-1.0.0.jar bundle download <bundle_id>

Example:

java -jar target/resourcebundler-1.0.0.jar bundle download 1404819b-2fd

Downloads a tar.gz archive containing the bundled files.

• Un-archive to confirm files or directories were bundled correctly:

tar -tzf output/1404819b-2fd.tar.gz

Design:

• Deduplication is implemented via SHA-256 content hashing.
• Multipart uploads are streamed to avoid loading files into memory.
• Archives are generated on demand without temporary files.
• Persistence uses local disk storage for simplicity.
• No unit tests are included per the challenge instructions.
• Error and exception handling is implemented.

Tradeoffs

• Multipart requests are constructed manually to minimize dependencies.
• Archives are buffered in memory for simplicity.
• Concurrency control and authentication are out of scope.

Extensibility:

The codebase is structured to support easy additions such as:

• Bundle deletion.
• Metadata enrichment.
• Pagination.
• Batching and Streaming downloads.
• Alternative storage backends: Database or caching.

Author: Prachi Shah @ https://pcisha.my.canva.site/

Date: December 24, 2025

P.S. The default copyright laws apply.