Switching from using diff cover executable to diff cover import

.github/workflows/ci_pipeline.yml

@@ -135,7 +135,7 @@ jobs:
 
     - name: Install Dependencies
       run: |
-        pip install poetry wandb tree_sitter
+        pip install poetry wandb tree_sitter diff-cover
         poetry install
     - name: Build Executable
       run: make installer

Makefile

@@ -1,5 +1,6 @@
 # Makefile
 SITE_PACKAGES=$(shell python3 -c "import wandb, os; print(os.path.dirname(wandb.__file__))")
+DIFF_COVER_TEMPLATES=$(shell python -c "import diff_cover, os; print(os.path.join(os.path.dirname(diff_cover.__file__), 'templates'))")
 TOML_FILES=$(shell find cover_agent/settings -name "*.toml" | sed 's/.*/-\-add-data "&:."/' | tr '\n' ' ')
 
 .PHONY: test build installer
@@ -23,6 +24,7 @@ installer:
 		$(TOML_FILES) \
 		--add-data "$(SITE_PACKAGES)/vendor:wandb/vendor" \
 		--add-data "build_helpers/anthropic_tokenizer.json:litellm/litellm_core_utils/tokenizers" \
+		--add-data "$(DIFF_COVER_TEMPLATES):diff_cover/templates" \
 		--hidden-import=tiktoken_ext.openai_public \
 		--hidden-import=tiktoken_ext \
 		--hidden-import=wandb \
@@ -31,3 +33,4 @@ installer:
 		--onefile \
 		--name cover-agent \
 		cover_agent/main.py
+

cover_agent/UnitTestValidator.py

@@ -4,6 +4,8 @@
 import logging
 import os
 
+from diff_cover.diff_cover_tool import main as diff_cover_main
+
 from cover_agent.AICaller import AICaller
 from cover_agent.CustomLogger import CustomLogger
 from cover_agent.FilePreprocessor import FilePreprocessor
@@ -690,20 +692,32 @@ def post_process_coverage_report(self, time_of_test_command: int):
         return report
 
     def generate_diff_coverage_report(self):
-        # Run the diff-cover command to generate a JSON diff coverage report
-        coverage_filename = os.path.basename(self.code_coverage_report_path)
-        coverage_command = f"diff-cover --json-report {self.diff_coverage_report_name} --compare-branch={self.comparison_branch} {coverage_filename}"
-        # Log and execute the diff coverage command
-        self.logger.info(f'Running diff coverage command: "{coverage_command}"')
-        stdout, stderr, exit_code, _ = Runner.run_command(
-            command=coverage_command, cwd=self.test_command_dir
-        )
-
-        # Ensure the diff command executed successfully
-        assert exit_code == 0, (
-            f'Fatal: Error running diff coverage command. Are you sure the command is correct? "{coverage_command}"'
-            f"\nExit code {exit_code}. \nStdout: \n{stdout} \nStderr: \n{stderr}"
-        )
+        """
+        Generates a JSON diff coverage report using the diff-cover tool.
+        This method runs the diff-cover command with the specified arguments to generate
+        a JSON report that shows the coverage differences between the current branch and
+        the specified comparison branch.
+        Args:
+            None
+        Returns:
+            None
+        Raises:
+            Exception: If an error occurs while running the diff-cover command.
+        """
+
+        diff_cover_args = [
+            "diff-cover",
+            "--json-report",
+            self.diff_cover_report_path,
+            "--compare-branch={}".format(self.comparison_branch),
+            self.code_coverage_report_path
+        ]
+
+        self.logger.info(f'Running diff coverage module with args: "{diff_cover_args}"')
+        try:
+            diff_cover_main(diff_cover_args)
+        except Exception as e:
+            self.logger.error(f"Error running diff-cover: {e}")
 
     def get_current_coverage(self):
-        return self.current_coverage_report.total_coverage
+            return self.current_coverage_report.total_coverage

cover_agent/version.txt

@@ -1 +1 @@
-0.2.13
+0.2.15

tests/test_UnitTestValidator.py

@@ -102,7 +102,7 @@ def test_validate_test_pass_no_coverage_increase_with_prompt(self):
                 result = generator.validate_test(test_to_validate)
 
                 assert result["status"] == "FAIL"
-                assert result["reason"] == "Coverage did not increase. Maybe the test did run but did not increase coverage, or maybe the test execution was skipped due to some problem"
+                assert "Coverage did not increase" in result["reason"]
                 assert result["exit_code"] == 0
 
     def test_initial_test_suite_analysis_with_prompt_builder(self):
@@ -182,16 +182,40 @@ def test_post_process_coverage_report_without_flags(self):
                 coverage_report = generator.post_process_coverage_report(datetime.datetime.now())
                 assert coverage_report.total_coverage == 0.7
 
-    def test_generate_diff_coverage_report(self):
+
+    def test_generate_diff_coverage_report_success(self):
         with tempfile.NamedTemporaryFile(suffix=".py", delete=False) as temp_source_file:
             generator = UnitTestValidator(
                 source_file_path=temp_source_file.name,
                 test_file_path="test_test.py",
                 code_coverage_report_path="coverage.xml",
                 test_command="pytest",
                 llm_model="gpt-3",
-                diff_coverage=True
+                diff_coverage=True,
+                comparison_branch="main"
             )
-            with patch.object(Runner, 'run_command', return_value=("", "", 0, datetime.datetime.now())):
+            with patch("cover_agent.UnitTestValidator.diff_cover_main") as mock_diff_cover_main:
+                generator.generate_diff_coverage_report()
+                mock_diff_cover_main.assert_called_once_with([
+                    "diff-cover",
+                    "--json-report",
+                    generator.diff_cover_report_path,
+                    "--compare-branch=main",
+                    "coverage.xml"
+                ])
+
+    def test_generate_diff_coverage_report_failure(self):
+        with tempfile.NamedTemporaryFile(suffix=".py", delete=False) as temp_source_file:
+            generator = UnitTestValidator(
+                source_file_path=temp_source_file.name,
+                test_file_path="test_test.py",
+                code_coverage_report_path="coverage.xml",
+                test_command="pytest",
+                llm_model="gpt-3",
+                diff_coverage=True,
+                comparison_branch="main"
+            )
+            with patch("cover_agent.UnitTestValidator.diff_cover_main", side_effect=Exception("Mock exception")), \
+                    patch.object(generator.logger, 'error') as mock_logger_error:
                 generator.generate_diff_coverage_report()
-                assert generator.diff_cover_report_path.endswith("diff-cover-report.json")
+                mock_logger_error.assert_called_once_with("Error running diff-cover: Mock exception")