import nltk when needed

.github/workflows/pip.yaml

@@ -14,7 +14,7 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: ["3.9", "3.10", "3.11"]
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
     steps:
     - uses: actions/checkout@v2
     - name: Set up Python
@@ -23,18 +23,17 @@ jobs:
         activate-environment: test
         auto-update-conda: true
         python-version: ${{ matrix.python-version }}
-        channels: conda-forge, ingeotec
+        channels: conda-forge
         allow-softlinks: true
         channel-priority: flexible
         show-channel-urls: true	
     - name: Install dependencies
       run: |
         conda install --yes pip
         pip install twine
-        pip install jieba
-        conda install --yes numpy scipy scikit-learn nltk nose microtc
+        conda install --yes numpy scipy scikit-learn microtc
         python setup.py sdist
-        python setup.py bdist_wheel  
+        python setup.py bdist_wheel
     - name: Pip
       if: ${{ runner.os == 'Linux' }}
       env:
@@ -44,6 +43,6 @@ jobs:
     - name: Wheel
       if: ${{ runner.os != 'Linux' }}
       env:
-        TWINE: ${{ secrets.TWINE }}              
+        TWINE: ${{ secrets.TWINE }}            
       run: |
         twine upload --skip-existing -u mgraffg -p $TWINE dist/*;

.github/workflows/test.yaml

@@ -14,7 +14,7 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: ["3.9", "3.10", "3.11"]
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
     steps:
     - uses: actions/checkout@v2
     - name: Set up Python
@@ -23,7 +23,7 @@ jobs:
         activate-environment: test
         auto-update-conda: true
         python-version: ${{ matrix.python-version }}
-        channels: conda-forge, ingeotec
+        channels: conda-forge
         allow-softlinks: true
         channel-priority: flexible
         show-channel-urls: true	
@@ -33,19 +33,18 @@ jobs:
         pip install coverage
         pip install coveralls
         pip install jieba
-        conda install --yes numpy scipy scikit-learn nltk nose microtc
-        python setup.py build_ext --inplace
+        conda install --yes numpy scipy scikit-learn nltk microtc pytest
     - name: Tests on Linux
       if: ${{ runner.os == 'Linux' }}
       run: |
         which python
         python --version
         which coverage
-        nosetests --verbose --with-coverage --cover-package=b4msa b4msa/tests
+        coverage run -m pytest b4msa/tests
     - name: Tests on macOS and Windows
       if: ${{ runner.os != 'Linux' }}
       run: |
-        nosetests --verbose b4msa/tests
+        python -m pytest b4msa/tests
     - name: coveralls
       if: ${{ runner.os == 'Linux' }}
       env:

b4msa/__init__.py

@@ -17,4 +17,6 @@
 
 The starting point is :py:class:`b4msa.textmodel.TextModel`
 """
-__version__ = "2.2.3"
+__version__ = "2.2.4"
+
+from b4msa.textmodel import TextModel

b4msa/lang_dependency.py

@@ -49,6 +49,8 @@ def get_lang(l):
 
 
 class LangDependencyError(Exception):
+    """Lang Dependency Error"""
+
     def __init__(self, message):
         self.message = message
 
@@ -74,19 +76,31 @@ def __init__(self, lang="spanish"):
         Initializes the parameters for specific language
         """
 
-        from nltk.stem.snowball import SnowballStemmer
+
         self.languages = ["spanish", "english", "italian", "german", "arabic"]
         self.lang = lang
 
-        if self.lang not in SnowballStemmer.languages and self.lang != 'chinese':
-            raise LangDependencyError("Language not supported for stemming: " + lang)
-        if self.lang == "english":
-            from nltk.stem.porter import PorterStemmer
-            self.stemmer = PorterStemmer()
-        elif self.lang == 'chinese':
-            self.stemmer = None
-        else:
-            self.stemmer = SnowballStemmer(self.lang)
+
+
+    @property
+    def stemmer(self):
+        """stemmer"""
+        from nltk.stem.snowball import SnowballStemmer        
+
+        try:
+            return self._stemmer
+        except AttributeError:
+            if self.lang not in SnowballStemmer.languages and self.lang != 'chinese':
+                _ = f"Language not supported for stemming: {self.lang}"
+                raise LangDependencyError(_)
+            if self.lang == "english":
+                from nltk.stem.porter import PorterStemmer
+
+                self.stemmer = PorterStemmer()
+            elif self.lang == 'chinese':
+                self.stemmer = None
+            else:
+                self.stemmer = SnowballStemmer(self.lang)
 
     @property
     def lang(self):

pyproject.toml

@@ -0,0 +1,46 @@
+[project]
+name = "b4msa"
+description = "Baselines for Multilingual Sentiment Analysis"
+readme = "README.rst"
+dependencies = [
+    "numpy",
+    "scikit-learn>=1.3.0",
+    "microtc"
+]
+dynamic = ["version"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Environment :: Console",
+    "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: Apache Software License",
+    "Natural Language :: English",
+    "Operating System :: MacOS :: MacOS X",
+    "Operating System :: POSIX :: Linux",
+    "Programming Language :: Python :: 3",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence"    
+]
+
+authors = [
+  {name = "Eric S. Tellez"},
+  {name = "Sabino Miranda-Jiménez"},
+  {name = "Mario Graff"},
+  {name = "Daniela Moctezuma"},
+  {name = "Ranyart R. Suárez"},
+  {name = "Oscar S. Siordia"}
+]
+
+maintainers = [
+  {name = "Mario Graff", email = "mgraffg@ieee.org"}
+]
+
+[tool.setuptools.dynamic]
+version = {attr = 'b4msa.__version__'}
+
+[project.urls]
+Homepage = "https://b4msa.readthedocs.io"
+Repository = "https://github.com/INGEOTEC/b4msa"
+Issues = "https://github.com/INGEOTEC/b4msa/issues"
+
+[build-system]
+requires = ["setuptools >= 61.0"]
+build-backend = "setuptools.build_meta"

setup.py

@@ -12,50 +12,5 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 from setuptools import setup
-import b4msa
 
-
-with open('README.rst') as fpt:
-    long_desc = fpt.read()
-
-
-setup(
-    name="b4msa",
-    description="""Baselines for Multilingual Sentiment Analysis""",
-    long_description=long_desc,
-    version=b4msa.__version__,
-    classifiers=[
-        "Development Status :: 3 - Alpha",
-        "Environment :: Console",
-        "Intended Audience :: Science/Research",
-        "License :: OSI Approved :: Apache Software License",
-        "Natural Language :: English",
-        "Operating System :: MacOS :: MacOS X",
-        "Operating System :: POSIX :: Linux",
-        'Programming Language :: Python :: 3',
-        "Topic :: Scientific/Engineering :: Artificial Intelligence"],
-    # url='https://github.com/mgraffg/EvoDAG',
-    # author="Mario Graff",
-    # author_email="mgraffg@ieee.org",
-    # cmdclass={"build_ext": build_ext, "clean": Clean},
-    # ext_modules=ext_modules,
-    packages=['b4msa', 'b4msa/tests', 'b4msa/tools', 'b4msa/resources'],
-    include_package_data=True,
-    zip_safe=False,
-    package_data={'b4msa/resources': ['arabic.stopwords',
-                                      'spanish.neg.stopwords',
-                                      'spanish.stopwords',
-                                      'english.stopwords',
-                                      'english.neg.stopwords',
-                                      'emoticons.json',
-                                      'italian.stopwords',
-                                      'italian.neg.stopwords'],
-                  'b4msa/tests': ['text.json']},
-    install_requires=['microtc'],
-    scripts=['b4msa/tools/b4msa-train',
-             'b4msa/tools/b4msa-test',
-             'b4msa/tools/b4msa-params',
-             'b4msa/tools/b4msa-perf',
-             'b4msa/tools/b4msa-kfolds',
-             'b4msa/tools/b4msa-textModel']
-)
+setup()