SPECS: add python-sentencepiece

SPECS/python-sentencepiece/python-sentencepiece.spec

@@ -0,0 +1,44 @@
+# SPDX-FileCopyrightText: (C) 2026 Institute of Software, Chinese Academy of Sciences (ISCAS)
+# SPDX-FileCopyrightText: (C) 2026 openRuyi Project Contributors
+# SPDX-FileContributor: Kimmy <yucheng.or@isrc.iscas.ac.cn>
+#
+# SPDX-License-Identifier: MulanPSL-2.0
+
+%global srcname sentencepiece
+
+Name:           python-%{srcname}
+Version:        0.2.1
+Release:        %autorelease
+Summary:        Unsupervised text tokenizer and detokenizer
+License:        Apache-2.0
+URL:            https://github.com/google/sentencepiece
+#!RemoteAsset:  sha256:8138cec27c2f2282f4a34d9a016e3374cd40e5c6e9cb335063db66a0a3b71fad
+Source:         https://files.pythonhosted.org/packages/source/s/%{srcname}/%{srcname}-%{version}.tar.gz
+BuildSystem:    pyproject
+
+BuildOption(install):  -l %{srcname} -L
+
+BuildRequires:  cmake
+BuildRequires:  pyproject-rpm-macros
+BuildRequires:  pkgconfig(python3)
+BuildRequires:  python3dist(setuptools)
+BuildRequires:  python3dist(protobuf)
+
+Provides:       python3-%{srcname} = %{version}-%{release}
+%python_provide python3-%{srcname}
+
+%description
+SentencePiece is an unsupervised text tokenizer and detokenizer mainly for
+Neural Network-based text generation systems. SentencePiece implements
+subword units (e.g., byte-pair-encoding (BPE) and unigram language model)
+with the extension of direct training from raw sentences. This package
+bundles the SentencePiece C++ library.
+
+%generate_buildrequires
+%pyproject_buildrequires
+
+%files -f %{pyproject_files}
+%doc README.md
+
+%changelog
+%autochangelog