textproc/py-semchunk: New port: Python library for splitting text into semantically coherent chunks

textproc/Makefile

@@ -1653,6 +1653,7 @@
     SUBDIR += py-scour
     SUBDIR += py-segments
     SUBDIR += py-segno
+    SUBDIR += py-semchunk
     SUBDIR += py-sense2vec
     SUBDIR += py-sentencepiece
     SUBDIR += py-simplebayes

textproc/py-semchunk/Makefile

@@ -0,0 +1,23 @@
+PORTNAME=	semchunk
+DISTVERSION=	4.1.1
+CATEGORIES=	textproc python
+MASTER_SITES=	PYPI
+PKGNAMEPREFIX=	${PYTHON_PKGNAMEPREFIX}
+
+MAINTAINER=	yuri@FreeBSD.org
+COMMENT=	Python library for splitting text into semantically coherent chunks
+WWW=		https://github.com/isaacus-dev/semchunk
+
+LICENSE=	MIT
+
+BUILD_DEPENDS=	${PYTHON_PKGNAMEPREFIX}hatchling>0:devel/py-hatchling@${PY_FLAVOR}
+RUN_DEPENDS=	${PYTHON_PKGNAMEPREFIX}tqdm>0:misc/py-tqdm@${PY_FLAVOR}
+
+USES=		python
+USE_PYTHON=	autoplist concurrent pep517
+NO_ARCH=	yes
+
+post-extract:
+	@${REINPLACE_CMD} -e 's/^license = "\([^"]*\)"/license = {text = "\1"}/' -e '/^license-files/d' ${WRKSRC}/pyproject.toml
+
+.include <bsd.port.mk>

textproc/py-semchunk/distinfo

@@ -0,0 +1,3 @@
+TIMESTAMP = 1783063884
+SHA256 (semchunk-4.1.1.tar.gz) = f27dc85716a0e9509a3f99c78b961cc476b3120efe711ee9a9a950ed1bd0f8ba
+SIZE (semchunk-4.1.1.tar.gz) = 24910

textproc/py-semchunk/pkg-descr

@@ -0,0 +1,4 @@
+Semchunk splits text into smaller chunks while preserving as much
+local semantic context as possible. It is designed for use in RAG
+and LLM applications where text must be divided into segments that
+maintain coherent meaning for embedding and retrieval.