Bridge: Harbour ⇄ Python AI & Data Ready Virtual Environments

XDPYHARBOUR – Python meets Harbour

Keep your proven Harbour/Clipper code – and plug straight into the modern Python world: AI frameworks, data tools, HTTP clients, virtual environments and more. XDPYHARBOUR lets your PRGs talk Python, and Python talk back to Harbour.

Join Discord & Get the Toolkit Back to Clipper2Web Overview

Why Python is central in 2025

Python has become the default language for AI, data science and “glue code”. Most cutting-edge examples, tutorials and client libraries appear in Python first. Connecting Harbour to Python means your legacy-strong business logic can benefit from that entire ecosystem without a rewrite.

AI & ML ecosystems

Frameworks like PyTorch, TensorFlow, scikit-learn, LangChain or LlamaIndex are all Python-first. With a bridge, Harbour can orchestrate these tools for text generation, recommendations, embeddings and more.

The “glue” for modern stacks

Python libraries cover HTTP APIs, cloud services, PDF/Excel I/O, crypto, automation and DevOps tooling. Instead of re-implementing clients in Harbour, you call well-tested Python packages directly.

Fast prototyping → production

Jupyter notebooks and scripts make experiments fast. Once a prototype works, XDPYHARBOUR lets you call it from your existing Harbour workflows – no need to move your DBF logic elsewhere.

Huge community & resources

Blogs, StackOverflow answers, example repos – almost everything has a Python version. Harbour developers can now reuse these building blocks instead of starting from zero.

What XDPYHARBOUR brings to the table

XDPYHARBOUR is not “just” a foreign-function wrapper. It is designed as a controlled bridge: Harbour remains the stable core, Python acts as a powerful co-processor for AI, data and integrations.

  • Two-way execution: run Python code from Harbour and call Harbour routines from Python, so both runtimes can reuse each other’s strengths.
  • Virtual environments from code: create, list and remove Python virtual environments programmatically, directly from your PRGs.
  • Package management API: install and uninstall packages (like requests, pandas, PDF tools, etc.) without touching the system-wide Python installation.
  • Safe sandboxes for experiments: test new libraries and AI workflows in isolated environments that can be deleted afterwards – your production stack stays clean.
  • Console-friendly UX: the toolkit is designed with classic Harbour/FiveWin workflows in mind: clear messages, simple prompts, no hidden magic.

Typical scenarios for Harbour developers

You do not have to change your database format or your UI toolkit. Think of XDPYHARBOUR as a plug-in slot that adds “Python superpowers” to your existing systems.

AI helpers for hotel & business apps

Generate reply drafts for customer emails, travel suggestions, or marketing texts based on DBF data. Harbour handles the records, Python talks to the AI models.

Document & data processing

Use Python libraries to read PDFs, Excel sheets or CSV files, run analysis with pandas, then hand the results back into your Harbour workflows.

Modern API integrations

Connect your Harbour apps to REST/GraphQL APIs, payment providers, or cloud dashboards using mature Python clients – no need to re-implement complex protocols in C.

Getting started in three steps

  1. Join the Discord server and download the latest XDPYHARBOUR package from the #anuncios (announcements) channel.
  2. Run the sample programs to create a virtual environment, install a test package (for example requests) and execute a simple Python script from Harbour.
  3. Integrate into your own PRGs – start with one “assistant” task like PDF parsing, AI text generation or an external API call, and grow from there.

The project is actively maintained and evolving – feedback and real-world stories from Harbour developers are very welcome.

Open Discord & Downloads

Why this matters for legacy Harbour apps

Many business systems built on Clipper/Harbour/DBF are still mission-critical. Rewriting them in a new stack is risky and expensive. A controlled bridge to Python lets you add new capabilities – especially AI and analytics – while keeping your stable core intact.

In other words: Harbour stays the captain, Python becomes a powerful first officer.