add test workflow

Author: youyupei

.github/workflows/unittest.yml

@@ -0,0 +1,28 @@
+name: Python application test with Unittest
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  build:
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, macos-latest]
+        python-version: ['3.7', '3.8', '3.9', '3.10']
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v2
+      with:
+        python-version: ${{ matrix.python-version }}
+    - name: Install blaze
+      run: |
+        python -m pip install --upgrade pip
+        pip install .
+    - name: Run test with unittest
+      run: |
+        python -m unittest discover -s test/

setup.py

@@ -10,6 +10,7 @@
     author_email="[email protected]",
     description='Barcode identification from Long reads for AnalyZing single cell gene Expression',
     packages=find_packages(),
+    test_suite='test',
     long_description=long_description,
     long_description_content_type='text/markdown',
     package_data={'blaze': ['10X_bc/3M-february-2018.zip', '10X_bc/737K-august-2016.txt']},

test/test_blaze.py

@@ -0,0 +1,84 @@
+import unittest
+import filecmp
+from blaze import blaze
+import blaze.helper as helper
+from blaze.config import *
+import os
+import gzip
+import shutil
+import difflib
+
+class TestBlazeMain(unittest.TestCase):
+
+    def setUp(self):
+        self.argv = \
+            ' --expect-cells=500 --threads=8 --output-prefix test_ test/data/'
+        self.expected_normal_files = [
+            'test_' + DEFAULT_GRB_OUT_RAW_BC,
+            'test_' + DEFAULT_GRB_OUT_WHITELIST,
+            'test_' + DEFAULT_EMPTY_DROP_FN,
+            'test_' + DEFAULT_KNEE_PLOT_FN,
+            'test_' + DEFAULT_BC_STAT_FN
+        ]
+        self.expected_gz_files = ['test_' + DEFAULT_GRB_OUT_FASTQ]
+
+        self.expected_dir = 'test/expect_output/'
+
+    def decompress_gz_file(self, file_path):
+        with gzip.open(file_path, 'rb') as f_in:
+            with open(file_path+'.tmp', 'wb') as f_out:  # remove .gz extension
+                shutil.copyfileobj(f_in, f_out)
+        return file_path+'.tmp'
+
+    def compare_gz_files(self, file1, file2):
+        file1_decompressed = self.decompress_gz_file(file1)
+        file2_decompressed = self.decompress_gz_file(file2)
+        comparison_result = filecmp.cmp(file1_decompressed, file2_decompressed)
+        
+        # Clean up decompressed files
+        os.remove(file1_decompressed)
+        os.remove(file2_decompressed)
+        if comparison_result:
+            return comparison_result
+        else:
+            self.diff_files(file1_decompressed, file2_decompressed)
+            return comparison_result
+    
+    def diff_files(self, file1, file2):
+        with open(file1, 'r') as f1, open(file2, 'r') as f2:
+            diff = difflib.unified_diff(
+                f1.readlines(),  # Read lines from the first file
+                f2.readlines(),  # Read lines from the second file
+                fromfile=file1,
+                tofile=file2,
+            )
+            # Output the difference to console or write to a file
+            for i, line in enumerate(diff):
+                if i > 10:
+                    break
+                print(line, end='')
+
+    def test_main(self):
+        # Run the main function
+        blaze(self.argv)
+
+        # Check that the expected files were created and are correct
+        for fn in self.expected_normal_files:
+            self.assertTrue(os.path.exists(fn), f"File {fn} was not created")
+            if not filecmp.cmp(fn, self.expected_dir + fn):
+                self.diff_files(fn, self.expected_dir + fn)
+            self.assertTrue(filecmp.cmp(fn, self.expected_dir + fn), f"File {fn} does not match the expected output (({self.expected_dir + fn}))")
+            
+         
+        for fn in self.expected_gz_files:
+            self.assertTrue(os.path.exists(fn), f"File {fn} was not created")
+            self.assertTrue(self.compare_gz_files(fn, self.expected_dir + fn) , f"File {fn} does not match the expected output ({self.expected_dir + fn})")
+
+    def tearDown(self):
+        # Remove created files after tests
+        for fn in self.expected_normal_files + self.expected_gz_files:
+            if os.path.exists(fn):
+                os.remove(fn)
+
+if __name__ == '__main__':
+    unittest.main()