Thanks for helping build datasheet2spice. The project needs power electronics
users, simulator users, PDF/OCR people, frontend contributors, and anyone who
can bring real datasheet cases.
Useful small contributions include:
- add a public datasheet case to
validation/public_datasheet_cases.json, - add a golden case under
validation/golden/, - improve a README or tutorial for your simulator,
- add one extraction fixture for a table layout that currently fails,
- run
benchmark-modelon a device and attach the report to an issue, - improve error messages or review UI copy.
python -m pip install -e .[pdf]
$env:PYTHONPATH="$PWD\src"
python -m unittest discover -s tests -vBefore pushing, run:
python -m unittest discover -s tests -v
datasheet2spice score-case examples/demo_sic_diode/device.json validation/golden/demo_sic_diode.case.json
datasheet2spice benchmark-model examples/demo_sic_diode/device.json --out build/bench-diode --model diode-basic --model diode-abm-dynamic --dialect all
git diff --checkIf you change frontend files under web/, also run:
python tools/sync_web_frontend.py- Keep core runtime dependencies minimal.
- Put optional AGPL/GPL/proprietary integrations behind plugins.
- Do not commit confidential datasheets, vendor encrypted models, simulator raw files, or large generated bundles.
- Store third-party PDF/model URLs and expected values, not redistributed vendor files.
- Keep generated model claims modest and traceable.
- Add tests for new emitters, extractors, validators, and quality gates.
- Public source URL is included.
- Vendor and part numbers are clear.
- Component profile is set, such as
mosfet.powerordiode.power. - Layout tags describe the challenge, such as
series-datasheet,scanned-pdf,capacitance-curve, orvendor-spice-model. - Expected fields include units and tolerances.
- Any vendor SPICE model license constraints are noted.
- Registers through
datasheet2spice.plugins.emitter. - Supports at least one documented dialect.
- Emits deterministic text.
- Has tests and a benchmark path.
- Documents convergence and accuracy limitations.
Open a draft PR early if you want design feedback. A good PR description states:
- what changed,
- how it was tested,
- what model/extraction limitations remain,
- whether any generated evidence files are included.