Installation

dicom-dre requires Python 3.12 and uses uv for package management.

Install the package

Install from a checkout of the repository:

uv pip install -e .

This installs the library and the dicom-dre command-line entry point.

Verify the install:

dicom-dre --help

Optional: build the JPEG DCT C extension

The JPEG DCT-domain scrubber has a pure-Python implementation and an optional CFFI-based C extension. The extension is not required; when it is absent, the engine falls back to the pure-Python path. Building it reduces the time spent zeroing DCT blocks in JPEG Baseline images.

Build the extension with the just recipe:

just build-ext

The recipe compiles src/dicom_dre/_jpeg_dct_accel.c and reports the path of the resulting shared object. To remove the compiled artifacts:

just clean-ext

To confirm the pure-Python fallback still works without the compiled extension, run:

just test-fallback

Development environment

Development uses a Dev Container so every contributor has the same toolchain (Python 3.12, uv, just, dcmtk, and the build dependencies for the C extension) regardless of host operating system.

Prerequisites

• Visual Studio Code

• Docker

• The Dev Containers VS Code extension

Open the repository in the container

1. Clone the repository and open the folder in VS Code.

2. When VS Code detects the Dev Container configuration, select “Reopen in Container”. You can also run “Dev Containers: Reopen in Container” from the command palette.

3. The first build downloads the base image and installs the container features; later starts reuse the cached image.

The container’s postCreateCommand runs .devcontainer/post-create.sh, which:

• runs uv sync to install all dependency groups,

• compiles the CFFI JPEG-DCT acceleration extension into src/dicom_dre/, and

• installs the pre-commit, pre-push, and commit-msg git hooks.

No manual bootstrap step is required. After the script finishes, run the test suite:

uv run pytest

See Testing for the full set of test recipes.