pyproject.toml

# Packaging for HJCFIT.
#
# scikit-build-core drives the existing CMake build rather than replacing it, so
# there is one build system rather than two descriptions of the same thing that
# can drift apart.
#
# MPFR is off for wheels. It is the arbitrary-precision fallback for root
# bracketing, it has never been available on Windows, and it cannot be built on
# macOS at all because HJCFITConfig.h.in instantiates std::complex<mpfr::mpreal>,
# which the standard does not define (F26). Shipping it on Linux alone would mean
# wheels that compute slightly different things on different platforms, which is
# worse than not shipping it. Build from source with -DHJCFIT_USE_MPFR=ON if you
# need it.

[build-system]
# swig is a build tool, not a library, and the PyPI package ships a real
# binary so a version can be pinned -- Homebrew, yum and chocolatey each serve
# whatever is current. SWIG 4.5.0 generates Python 2 C-API calls into the
# generated wrapper and the build fails to compile (#188). Putting it here
# rather than in a CI step is what the package documents, and is what puts it
# in the isolated build environment where the console script can import its
# own module.
requires = ["scikit-build-core>=0.10", "numpy>=1.25", "swig<4.5"]
build-backend = "scikit_build_core.build"

[project]
name = "hjcfit"
version = "1.2.0"
description = "Maximum likelihood fitting of ion-channel mechanisms with exact missed-events correction"
readme = "README.md"
license = { file = "LICENSE" }
requires-python = ">=3.10"

authors = [
    { name = "Mayeul d'Avezac" },
    { name = "Jens Hedegaard Nielsen" },
    { name = "Raquel Alegre" },
    { name = "Remigijus Lape" },
]

keywords = [
    "ion channels", "single channel", "patch clamp", "electrophysiology",
    "maximum likelihood", "Q-matrix", "missed events", "HJC",
]

classifiers = [
    "Development Status :: 4 - Beta",
    "Intended Audience :: Science/Research",
    "License :: OSI Approved :: GNU General Public License v3 (GPLv3)",
    "Operating System :: MacOS",
    "Operating System :: Microsoft :: Windows",
    "Operating System :: POSIX :: Linux",
    "Programming Language :: C++",
    "Programming Language :: Python :: 3",
    "Topic :: Scientific/Engineering :: Bio-Informatics",
]

dependencies = [
    "numpy>=1.25",
    "scipy>=1.10",
]

# The likelihood needs numpy and scipy and nothing else, and a test asserts
# that importing HJCFIT leaves dcio and scalcs out of sys.modules. Reading a
# record, defining a mechanism and running a fit need more than that, so they
# are an extra rather than a dependency: someone who wants Log10Likelihood on
# a Q matrix they already have should not acquire a mechanism library and a
# plotting stack to get it.
#
#   pip install hjcfit             the likelihood
#   pip install hjcfit[fitting]    records, mechanisms, fits, displays
#
# scalcs is GPL-2.0-or-later, which HJCFIT (GPL-3.0-or-later) can combine
# with. It could not have been combined with GPL-2.0-only, which is what the
# metadata claimed until DCPROGS/SCALCS#24.
[project.optional-dependencies]
fitting = [
    "dcio>=0.1.2",      # records: formats, dead time, periods, bursts
                        # 0.1.2 because 0.1.1 reports the wrong __version__
    "scalcs>=1.2.0",    # mechanisms, rate constants, constraints, curves
                        # 1.2.0 because the fitter needs Rate.is_free, which
                        # 1.1.0 does not have: the looser pin let the extra
                        # resolve and the fit then fail on a missing attribute
    "matplotlib>=3.6",  # the displays; nothing in the likelihood draws
    # A fit specification is TOML, which is in the standard library from 3.11.
    # tomli is the same parser by the same author, for 3.10.
    "tomli>=1.1; python_version < '3.11'",
]
# EKDIST -- exponential mixtures and t_crit estimation -- is deliberately not
# an extra. It is not on PyPI, so declaring it would make
# `pip install hjcfit[...]` fail outright rather than install what it can;
# EKDIST's own dcio dependency was unresolvable in exactly that way until dcio
# was published. Install it from source when a display needs it.

# `hjcfit fit spec.toml` -- the command over a fit specification. It needs the
# [fitting] extra to do anything, and says so rather than failing on an import;
# `pip install hjcfit` alone still installs the script, which is better than a
# script whose presence depends on how the package was installed.
[project.scripts]
hjcfit = "HJCFIT.likelihood.cli:main"

[project.urls]
Homepage = "https://github.com/DCPROGS/HJCFIT"
Documentation = "http://dcprogs.github.io/HJCFIT"
Repository = "https://github.com/DCPROGS/HJCFIT"
Issues = "https://github.com/DCPROGS/HJCFIT/issues"

[tool.scikit-build]
cmake.version = ">=3.16"
build-dir = "build/{wheel_tag}"

# Tests need googletest fetched from the network and behave installed; neither
# belongs in a wheel build. The documentation targets need doxygen, sphinx,
# pandoc and latex.
cmake.define.tests = "OFF"
cmake.define.compileDocs = "OFF"
cmake.define.executenotebooks = "OFF"
cmake.define.HJCFIT_USE_MPFR = "OFF"

# The CMake install rules place everything; there is no source package tree to
# copy in on top of them.
wheel.packages = []

# Jupyter's editor state, checked in years ago, duplicates the notebooks it
# shadows -- two copies of a 371 KB file in every source distribution.
sdist.exclude = [".ipynb_checkpoints"]