xlt-core 0.0.1

A lightweight Extract-Load tool for building simple, reliable batch data transfer pipelines

XLT — The YAML-first Data Move Tool

Pronounced: ex-ell-tee

XLT is a lightweight Extract–Load tool for building simple, reliable batch data movement pipelines.
Write a single YAML file that defines an extract and a load step, and run it anywhere with a clean, fast CLI.

XLT gives data engineers a predictable, extensible, Git-native way to move data between databases, APIs, files, and cloud warehouses — without operating a control plane or writing yet another custom Python script.

---

🚀 Features

• YAML-first pipelines — define your extract and load declaratively
• CLI-first design — run pipelines from the command line or any orchestrator
• Extensible — create custom extractors/loaders with minimal Python
• Portable — works anywhere Python runs (local, cron, Airflow, Prefect, GitHub Actions, on-prem)
• Incremental loads — simple watermark & key-based incremental patterns
• Load strategies — append, truncate-insert, merge (planned)
• Zero platform overhead — no UI, no scheduler, no metadata DB
• Engineer-friendly — CLI-first, Git-native, orchestration-agnostic

If you’ve ever written a one-off Python script to copy data from A to B, XLT replaces that with a clean YAML pipeline and a battle-tested execution engine.

---

Design Philosophy

XLT aims to occupy the gap between:

“I’ll write another custom Python script…” and “Let’s deploy Airbyte / a managed ELT platform.”

Principles:

• Simplicity over magic
• Declarative over imperative
• Small surface area
• No servers
• Do one thing well: move data

If dbt is the “data build tool”, XLT is the data move tool.

📦 Installation

pip install xlt

🤝 Contributing

XLT is designed to be open, simple, and community-friendly.

Ways to contribute:

• Create a new adapter (database, file, API, cloud service)
• Improve docs and examples
• Add validator rules for YAML schema
• Build testing fixtures
• Submit ideas for features

PRs are welcome!

📄 License

MIT License.