Fluvo

Fluvo gets data into Odoo, fast. A high-performance, Polars-backed ETL toolkit that replaces fragile, manual data prep with declarative, repeatable, configuration-as-code workflows.

---

Why Fluvo?

• ⚡ Fast at scale — a Polars-backed, multi-threaded engine for importing and exporting large Odoo datasets.
• 🛟 Imports that don't lose data — a load → create fallback rescues the good records from a failing batch and writes the rest to a fail file with the exact error, so a few bad rows never sink the whole import.
• 🔗 No import-order headaches — automatically detects relational and parent/child data and runs a two-pass import, so you don't have to hand-order your files.
• 🧩 Configuration-as-code — define transforms as plain Python with a rich mapper library: readable, repeatable, and version-controlled.
• 🔄 Server-to-server migration — export, transform, and import between two Odoo instances in a single in-memory step.
• ✅ Clean data before it lands — built-in cleaning and validation catch bad values before they ever reach Odoo.

Installation

You can install Fluvo via uv or pip from PyPI:

$ uv pip install fluvo

Quick Usage Example

The core workflow involves two simple steps:

1. Transform your source data with a Python script. Create a transform.py file to define the mapping from your source file to Odoo's format.

# transform.py
from fluvo.lib.transform import Processor
from fluvo.lib import mapper

my_mapping = {
    'id': mapper.m2o('product', 'SKU'),  # external id, e.g. product.A1
    'name': mapper.val('ProductName'),
    'list_price': mapper.num('Price'),
}

processor = Processor(my_mapping, source_filename='origin/products.csv', separator=',')
processor.process('data/products_clean.csv', {'model': 'product.product'})
processor.write_to_file("load.sh")

...

$ python transform.py

2. Load the clean data into Odoo using the CLI. The transform.py script generates a load.sh file containing the correct CLI command.

# Contents of the generated load.sh
fluvo import --connection-file conf/connection.conf --file data/products_clean.csv --model product.product ...

Then execute the script.

$ bash load.sh

When the import command runs, it automatically detects the data structure. If it finds relational data like parent_id fields, it will automatically switch to a robust two-pass strategy to ensure the import succeeds.

Documentation

For a complete user guide, tutorials, and API reference, please see the full documentation on Read the Docs. Please see the Command-line Reference for details.

Contributing

Contributions are very welcome. To learn more, see the Contributor Guide.

License

Distributed under the terms of the LGPL 3.0 license, Fluvo is free and open source software.

Issues

If you encounter any problems, please file an issue along with a detailed description.

Credits

This development of project is financially supported by stefcy.com.