3 files changed, 33 insertions, 0 deletions
diff --git a/textproc/py-sacremoses/Makefile b/textproc/py-sacremoses/Makefile
new file mode 100644
index 000000000000..56dba0f3f173
--- /dev/null
+++ b/textproc/py-sacremoses/Makefile
@@ -0,0 +1,29 @@
+PORTNAME=	sacremoses
+DISTVERSION=	0.1.1
+CATEGORIES=	textproc python # machine-learning
+#MASTER_SITES=	PYPI # no tests
+PKGNAMEPREFIX=	${PYTHON_PKGNAMEPREFIX}
+
+MAINTAINER=	yuri@FreeBSD.org
+COMMENT=	Python port of Moses tokenizer, truecaser and normalizer
+WWW=		https://github.com/hplt-project/sacremoses
+
+LICENSE=	MIT
+LICENSE_FILE=	${WRKSRC}/LICENSE
+
+RUN_DEPENDS=	${PYTHON_PKGNAMEPREFIX}click>=7.1.2:devel/py-click@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}joblib>=1.1.0:devel/py-joblib@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}regex>=2020.1.8:textproc/py-regex@${PY_FLAVOR} \
+		${PYTHON_PKGNAMEPREFIX}tqdm>=4.64.0:misc/py-tqdm@${PY_FLAVOR}
+
+USES=		python
+USE_PYTHON=	distutils autoplist pytest
+
+USE_GITHUB=	yes
+GH_ACCOUNT=	hplt-project
+
+NO_ARCH=	yes
+
+# tests as of 0.1.1: 1 failed, 35 passed, 1 skipped in 15.70s, see https://github.com/hplt-project/sacremoses/issues/153
+
+.include <bsd.port.mk>
diff --git a/textproc/py-sacremoses/distinfo b/textproc/py-sacremoses/distinfo
new file mode 100644
index 000000000000..b06ad3a3bf45
--- /dev/null
+++ b/textproc/py-sacremoses/distinfo
@@ -0,0 +1,3 @@
+TIMESTAMP = 1755454968
+SHA256 (hplt-project-sacremoses-0.1.1_GH0.tar.gz) = 139faad4800a5da04498e4b5c297a5a8e1c958877b59090715d8a3f90e0b9b76
+SIZE (hplt-project-sacremoses-0.1.1_GH0.tar.gz) = 889511
diff --git a/textproc/py-sacremoses/pkg-descr b/textproc/py-sacremoses/pkg-descr
new file mode 100644
index 000000000000..b2027d4eba1c
--- /dev/null
+++ b/textproc/py-sacremoses/pkg-descr
@@ -0,0 +1 @@
+sacremoses is a Python port of Moses tokenizer, truecaser and normalizer.