%global pypi_name torchtext
%global pypi_version 0.16.0

# So models are not removed with testing locally
%bcond_with test

Name:           python-%{pypi_name}
Version:        %{pypi_version}
Release:        1%{?dist}
Summary:        Data loaders and abstractions for language processing, powered by PyTorch

# licensecheck reports
# Apache License 2.0
# ------------------
# text-0.16.0/torchtext/csrc/bert_tokenizer.cpp
#
# BSD 3-Clause License
# --------------------
# text-0.16.0/LICENSE
# This is the main license
#
# MIT License
# -----------
# text-0.16.0/run-clang-format.py
# This file is not distributed
License:        BSD-3-Clause AND Apache-2.0

URL:            https://github.com/pytorch/text
Source0:        %{url}/archive/refs/tags/v%{version}.tar.gz#/text-%{version}.tar.gz
# this is a tarball, not a git repo
Patch0:         0001-Prepare-torchtext-setup-for-fedora.patch
# We have removed the third_party dir
# so needs to be versioned
Patch1:         0001-Prepare-torchtext-cmake-for-Fedora.patch
%if %{without test}
# Remove models/ from submodules
Patch3:         0001-remove-torchtext-pretrained-models.patch
%endif

# Limit to these because that is what torch is on
ExclusiveArch:  x86_64 aarch64
%global toolchain clang

BuildRequires:  cmake
BuildRequires:  double-conversion-devel
BuildRequires:  clang
BuildRequires:  ninja-build
BuildRequires:  python3-devel
BuildRequires:  python3-pybind11
BuildRequires:  python3-torch-devel
BuildRequires:  re2-devel
BuildRequires:  sentencepiece-devel
BuildRequires:  utf8proc-devel

BuildRequires:  python3dist(nltk)
BuildRequires:  python3dist(requests)
BuildRequires:  python3dist(tqdm)

%description
torchtext is a complementary python package to torch, providing torch
with text processing functionality needed for AI.

%package -n     python3-%{pypi_name}
Summary:        %{summary}

%description -n python3-%{pypi_name}
torchtext is a complementary python package to torch, providing torch
with text processing functionality needed for AI.

%prep
%autosetup -p1 -n text-%{version}

rm -rf third_party/*

# Remove pretrained models
# https://github.com/pytorch/vision/issues/2597
# If we are testing, include the models.
%if %{without test}
rm -rf torchtext/models
%endif

%build
# Building uses python3_sitearch/torch/utils/cpp_extension.py
# cpp_extension.py does a general linking with all the pytorch libs which
# leads warnings being reported by rpmlint.
%py3_build

%install
%py3_install

%files -n python3-%{pypi_name}
%license LICENSE
%doc README.rst
%{python3_sitearch}/%{pypi_name}
%{python3_sitearch}/%{pypi_name}-*.egg-info/


%changelog
* Sat Dec 23 2023 Tom Rix <trix@redhat.com> - 0.16.0-1
- Initial package.