midas-zipper 0.1.3

Standalone zarr-zip generation for MIDAS FF/PF workflows (no C-binary or source-tree dependency).

midas-zipper

Standalone zarr-zip generation for MIDAS FF/PF workflows.

pip install midas-zipper works with no MIDAS source tree and no compiled C binaries — it depends only on numpy, h5py, zarr, numcodecs, numba, and pillow, plus the system zip/bzip2 tools.

It packages the zarr ingest that the FF and PF pipelines need to turn raw detector data (HDF5/GE/TIFF) into the *.MIDAS.zip files the rest of MIDAS consumes — previously the source-tree script utils/ffGenerateZipRefactor.py, which broke pip-only installs because it lived outside any package.

Notebooks

Worked-example Jupyter notebooks live in notebooks/. They are not shipped with pip install — get them by cloning the MIDAS repository.

CLI

# FF zarr-zip generation (formerly ffGenerateZipRefactor.py)
midas-ff-zip -resultFolder <dir> -paramFN <params.txt> -LayerNr <n>

# Update a key inside an existing Zarr.zip
midas-update-zarr -fn <file.MIDAS.zip> -keyToUpdate <key> -updatedValue <...>

Both are also runnable as modules:

python -m midas_zipper.ff_zip   -resultFolder ... -paramFN ... -LayerNr ...
python -m midas_zipper.update_zarr -fn ... -keyToUpdate ... -updatedValue ...

Python API

from midas_zipper import generate_ff_zip

generate_ff_zip(
    result_folder="/path/to/scan",
    param_file="/path/to/Parameters.txt",
    layer_nr=1,
)

Scope

This package intentionally excludes AutoCalibrateZarr.py: that tool drives the C/OpenMP calibration binaries (CalibrantPanelShiftsOMP, CalibrantIntegratorOMP, GetHKLList) and so cannot be pip-portable. It stays with the calibration workflow.