Adding py-langid, a standalone Language Identification (LangID) tool.

5 files changed, 77 insertions, 0 deletions

diff --git a/textproc/Makefile b/textproc/Makefile
index ec58fc5d5084..bdde123b855d 100644
--- a/textproc/Makefile
+++ b/textproc/Makefile
@@ -1297,6 +1297,7 @@
     SUBDIR += py-jtextfsm
     SUBDIR += py-junit-xml
     SUBDIR += py-langdetect
+    SUBDIR += py-langid
     SUBDIR += py-laserhammer
     SUBDIR += py-libxml2
     SUBDIR += py-license-expression
diff --git a/textproc/py-langid/Makefile b/textproc/py-langid/Makefile
new file mode 100644
index 000000000000..9bce3b136baf
--- /dev/null
+++ b/textproc/py-langid/Makefile
@@ -0,0 +1,33 @@
+# Created by: Thierry Thomas <thierry@pompo.net>
+# $FreeBSD$
+
+PORTNAME=	langid
+DISTVERSION=	1.1.6-20170715
+CATEGORIES=	textproc devel python
+PKGNAMEPREFIX=	${PYTHON_PKGNAMEPREFIX}
+
+MAINTAINER=	thierry@FreeBSD.org
+COMMENT=	Standalone Language Identification (LangID) tool
+
+LICENSE=	BSD2CLAUSE
+LICENSE_FILE=	${WRKSRC}/LICENSE
+
+BUILD_DEPENDS=	${PYNUMPY}
+RUN_DEPENDS=	${PYNUMPY}
+
+USE_GITHUB=	yes
+GH_ACCOUNT=	saffsd
+GH_PROJECT=	${PORTNAME}.py
+GH_TAGNAME=	4153583
+
+USES=		python:3.6+ shebangfix
+USE_PYTHON=	distutils
+SHEBANG_GLOB=	*.py
+
+NO_ARCH=	yes
+
+post-extract:
+	${MKDIR} ${WRKDIR}/unsupported-Python-2.7
+	${MV} ${WRKSRC}/langid/train ${WRKDIR}/unsupported-Python-2.7
+
+.include <bsd.port.mk>
diff --git a/textproc/py-langid/distinfo b/textproc/py-langid/distinfo
new file mode 100644
index 000000000000..7bf7e79e3177
--- /dev/null
+++ b/textproc/py-langid/distinfo
@@ -0,0 +1,3 @@
+TIMESTAMP = 1606746121
+SHA256 (saffsd-langid.py-1.1.6-20170715-4153583_GH0.tar.gz) = 04b005bd607fcf54f9b06b41a20f968fcb5bdd7d96ec5471177c88ef858ffc9d
+SIZE (saffsd-langid.py-1.1.6-20170715-4153583_GH0.tar.gz) = 1959856
diff --git a/textproc/py-langid/pkg-descr b/textproc/py-langid/pkg-descr
new file mode 100644
index 000000000000..5f3d76514028
--- /dev/null
+++ b/textproc/py-langid/pkg-descr
@@ -0,0 +1,17 @@
+langid.py is a standalone Language Identification (LangID) tool.
+
+The design principles are as follows:
+
+    Fast
+    Pre-trained over a large number of languages (currently 97)
+    Not sensitive to domain-specific features (e.g. HTML/XML markup)
+    Single .py file with minimal dependencies
+    Deployable as a web service
+
+Remark: the main script langid/langid.py is cross-compatible with both Python2
+and Python3, but the accompanying training tools are still Python2-only, hence
+not installed by this port.
+
+See also the port textproc/py-langdetect for a similar program.
+
+WWW: https://github.com/saffsd/langid.py
diff --git a/textproc/py-langid/pkg-plist b/textproc/py-langid/pkg-plist
new file mode 100644
index 000000000000..4773b9c32de8
--- /dev/null
+++ b/textproc/py-langid/pkg-plist
@@ -0,0 +1,23 @@
+bin/langid
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/PKG-INFO
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/SOURCES.txt
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/dependency_links.txt
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/entry_points.txt
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/not-zip-safe
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/requires.txt
+%%PYTHON_SITELIBDIR%%/langid-1.1.6-py%%PYTHON_VER%%.egg-info/top_level.txt
+%%PYTHON_SITELIBDIR%%/langid/__init__.py
+%%PYTHON_SITELIBDIR%%/langid/__pycache__/__init__.cpython-%%PYTHON_SUFFIX%%.opt-1.pyc
+%%PYTHON_SITELIBDIR%%/langid/__pycache__/__init__.cpython-%%PYTHON_SUFFIX%%.pyc
+%%PYTHON_SITELIBDIR%%/langid/__pycache__/langid.cpython-%%PYTHON_SUFFIX%%.opt-1.pyc
+%%PYTHON_SITELIBDIR%%/langid/__pycache__/langid.cpython-%%PYTHON_SUFFIX%%.pyc
+%%PYTHON_SITELIBDIR%%/langid/langid.py
+%%PYTHON_SITELIBDIR%%/langid/tools/__init__.py
+%%PYTHON_SITELIBDIR%%/langid/tools/__pycache__/__init__.cpython-%%PYTHON_SUFFIX%%.opt-1.pyc
+%%PYTHON_SITELIBDIR%%/langid/tools/__pycache__/__init__.cpython-%%PYTHON_SUFFIX%%.pyc
+%%PYTHON_SITELIBDIR%%/langid/tools/__pycache__/featWeights.cpython-%%PYTHON_SUFFIX%%.opt-1.pyc
+%%PYTHON_SITELIBDIR%%/langid/tools/__pycache__/featWeights.cpython-%%PYTHON_SUFFIX%%.pyc
+%%PYTHON_SITELIBDIR%%/langid/tools/__pycache__/printfeats.cpython-%%PYTHON_SUFFIX%%.opt-1.pyc
+%%PYTHON_SITELIBDIR%%/langid/tools/__pycache__/printfeats.cpython-%%PYTHON_SUFFIX%%.pyc
+%%PYTHON_SITELIBDIR%%/langid/tools/featWeights.py
+%%PYTHON_SITELIBDIR%%/langid/tools/printfeats.py