deotter

deotter is a small JDBC connection utility with:

• A Java core for loading database aliases from a config file and creating JDBC connections.
• A Python wrapper (via JPype) that uses the Java ConnectionManager from Python.

Repository Layout

• java/main/: Java connection manager and provider interfaces/implementations.
• java/tests/: Java smoke test (TestConn.java) for alias-based connectivity.
• python/: Python package, packaging metadata, and tests.
• scripts/: helper scripts (project-specific automation).

How It Works

The Java ConnectionManager loads config from:

~/.config/deotter/config.properties

It expects a databases list plus per-alias fields:

• <alias>.url
• <alias>.user
• <alias>.password

Example:

databases=main,analytics

main.url=jdbc:postgresql://localhost:5432/app
main.user=app_user
main.password=secret

analytics.url=jdbc:mysql://localhost:3306/warehouse
analytics.user=warehouse_user
analytics.password=secret

Java Usage

Use the singleton manager and request connections by alias.

ConnectionManager mgr = ConnectionManager.getInstance();
Connection conn = mgr.getConnection("main");

Python Usage

The Python package lives under python/ and uses JPype to start a JVM and call Java classes.

Install dependencies from python/:

uv sync

Basic usage:

import deotter

with deotter.connect("main", autocommit=False) as conn:
    with conn.cursor() as cur:
        cur.execute("SELECT 1")
        rows = cur.fetchall()
        print(rows)

Optional JDBC drivers:

• Drop driver JARs into ~/.deotter/drivers.
• At runtime, the Python wrapper compiles Java classes into ~/.deotter/bin and adds all ~/.deotter/drivers/*.jar entries to the JVM classpath.

Development Notes

• Python dependency for JPype is jpype1 (not jpype).
• Python package metadata is in python/pyproject.toml.
• The project is currently lightweight and under active iteration.

AI Assistance Disclosure

Parts of this repository, including documentation and testing artifacts, were created with AI assistance and reviewed by a human maintainer before merge.

AI-generated output is treated as draft material and is validated through manual review and local test execution.

License

See LICENSE.