ENRGDAQ: Data acquisition framework for neutrino physics experiments.
Handles readout, processing, and storage at multi-Gbps rates.

Quick start

git clone https://github.com/ENRG-tr/enrgdaq.git
cd enrgdaq
uv sync
uv run python src/run.py --daq-job-config-path configs/examples/

How it works

Everything runs as independent OS processes called DAQJobs. A supervisor spawns them from TOML config files, monitors their health, and restarts them if they crash. Jobs communicate through a ZMQ pub/sub broker inside the supervisor. For bulk data like waveforms and images, the producer writes directly to a shared memory ring buffer and sends a tiny handle over ZMQ. The consumer reads the data with zero copies.

Storage backends

CSV, ROOT, HDF5, MySQL, Redis, Raw, Memory. Mix multiple backends on the same job by stacking [store_config.*] sections in the TOML file.

Custom jobs

Subclass DAQJob, define a config, implement start(). The system auto-discovers your class anywhere under src/enrgdaq/.

from enrgdaq.daq.base import DAQJob

class DAQJobMySensor(DAQJob):
    config_type = MySensorConfig

    def start(self):
        while not self._has_been_freed:
            data = read_hardware()
            self._put_message_out(DAQJobMessageStoreRaw(
                data=data, store_config=self.config.store_config
            ))

Commands

• uv run python src/run.py starts the supervisor
• uv run pytest src/tests/ -v runs the test suite
• uv run ruff check src/ lints the code

Platform notes

Linux and macOS are fully supported. Windows support is partial (no fork, no shared memory ring buffers).