python-harness 0.0.3__tar.gz → 0.0.6__tar.gz

{python_harness-0.0.3/python_harness.egg-info → python_harness-0.0.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-harness
-Version: 0.0.3
+Version: 0.0.6
 Summary: An agentic codebase evaluation and evolution tool for Python projects.
 Author-email: Mingli Yuan <mingli.yuan@gmail.com>
 License: MIT

{python_harness-0.0.3 → python_harness-0.0.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "python-harness"
-version = "0.0.3"
+version = "0.0.6"
 description = "An agentic codebase evaluation and evolution tool for Python projects."
 requires-python = ">=3.10"
 readme = "README.md"

{python_harness-0.0.3 → python_harness-0.0.6}/python_harness/cli.py

@@ -104,7 +104,7 @@ def measure(path: str = typer.Argument(".", help="The path to evaluate")) -> Non
         if hard_results["mypy"]["status"] != "success":
             output = hard_results["mypy"].get("output", "")
             console.print(f"[red]Mypy issues found:[/red]\n{output}")
-        if hard_results["ty"]["status"] != "success":
+        if hard_results["ty"]["status"] not in ("success", "warning"):
             output = hard_results["ty"].get("output", "")
             # ty might print to stderr instead of stdout, or it might be missing
             error_msg = hard_results["ty"].get("error_message", "")
@@ -116,6 +116,9 @@ def measure(path: str = typer.Argument(".", help="The path to evaluate")) -> Non
                 console.print(
                     "[red]Ty failed, but no standard output was captured.[/red]"
                 )
+        elif hard_results["ty"]["status"] == "warning":
+            msg = hard_results["ty"].get("error_message", "ty not found")
+            console.print(f"[yellow]Ty warning:[/yellow] {msg}")
         if hard_results["radon_cc"]["status"] != "success":
             issues = hard_results["radon_cc"].get("issues", [])
             console.print(
@@ -127,6 +130,16 @@ def measure(path: str = typer.Argument(".", help="The path to evaluate")) -> Non
                     f"  - {issue['file']}: {issue['type']} '{issue['name']}' "
                     f"has CC {issue['complexity']}"
                 )
+            
+            # If radon failed for another reason 
+            # (e.g. radon not installed or syntax error)
+            if not issues and hard_results["radon_cc"].get("error_message"):
+                err_msg = hard_results['radon_cc'].get('error_message')
+                console.print(f"[red]Radon CC Error:[/red] {err_msg}")
+            elif not issues:
+                console.print(
+                    "[red]Radon CC failed but no specific issues were parsed.[/red]"
+                )
         sys.exit(1)
         
     console.print("[bold green]Hard Evaluation Passed![/bold green]")

{python_harness-0.0.3 → python_harness-0.0.6}/python_harness/hard_evaluator.py

@@ -63,6 +63,7 @@ class HardEvaluator:
     def run_ty(self) -> dict[str, Any]:
         """
         Run ty language server checks.
+        If ty is not installed, fail gracefully rather than crashing.
         """
         try:
             result = subprocess.run(
@@ -79,7 +80,18 @@ class HardEvaluator:
                 "output": output,
                 "return_code": result.returncode,
             }
+        except FileNotFoundError:
+            return {
+                "status": "warning", 
+                "error_message": "ty executable not found. Skipping ty checks."
+            }
         except Exception as e:
+            # Handle cases where ty is found but fails to run or throws other OS errors
+            if "No such file or directory: 'ty'" in str(e):
+                return {
+                    "status": "warning", 
+                    "error_message": "ty executable not found. Skipping ty checks."
+                }
             return {"status": "error", "error_message": str(e)}
 
     def run_radon_cc(self) -> dict[str, Any]:
@@ -111,14 +123,18 @@ class HardEvaluator:
                                     "complexity": block.get('complexity')
                                 })
             
-            if issues:
+            if result.returncode != 0:
+                # E.g. syntax error in target code preventing radon from parsing
+                status = "failed"
+            elif issues:
                 status = "failed"
                 
             return {
                 "status": status,
                 "issues": issues,
                 "return_code": result.returncode,
-                "output": result.stdout
+                "output": result.stdout,
+                "error_message": result.stderr if result.returncode != 0 else ""
             }
         except Exception as e:
             return {"status": "error", "error_message": str(e)}
@@ -188,7 +204,7 @@ class HardEvaluator:
         all_passed = (
             ruff_res.get("status") == "success" and 
             mypy_res.get("status") == "success" and
-            ty_res.get("status") == "success" and
+            ty_res.get("status") in ("success", "warning") and
             radon_cc_res.get("status") == "success"
         )
 

{python_harness-0.0.3 → python_harness-0.0.6/python_harness.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-harness
-Version: 0.0.3
+Version: 0.0.6
 Summary: An agentic codebase evaluation and evolution tool for Python projects.
 Author-email: Mingli Yuan <mingli.yuan@gmail.com>
 License: MIT

python_harness-0.0.6/tests/test_hard_evaluator.py

@@ -0,0 +1,95 @@
+"""
+Tests for hard evaluation logic.
+"""
+
+from pathlib import Path
+from typing import Any
+
+from python_harness.hard_evaluator import HardEvaluator
+
+
+def test_hard_evaluator_methods() -> None:
+    """
+    Test methods of HardEvaluator.
+    """
+    evaluator = HardEvaluator(".")
+    
+    ruff_result = evaluator.run_ruff()
+    assert "status" in ruff_result
+    
+    mypy_result = evaluator.run_mypy()
+    assert "status" in mypy_result
+    
+    ty_result = evaluator.run_ty()
+    assert "status" in ty_result
+
+    # pytest_result = evaluator.run_pytest() # Causes infinite loop when run in test
+    # assert "status" in pytest_result
+    
+    eval_result = evaluator.evaluate()
+    assert "ruff" in eval_result
+    assert "mypy" in eval_result
+    assert "ty" in eval_result
+
+def test_ty_fallback_behavior(monkeypatch: Any) -> None:
+    """
+    Test that run_ty gracefully falls back to a warning when 'ty' is not installed.
+    """
+    # Create a mock subprocess.run that always raises FileNotFoundError
+    # to simulate 'ty' not being on the system PATH
+    def mock_run(*args: Any, **kwargs: Any) -> Any:
+        raise FileNotFoundError("[Errno 2] No such file or directory: 'ty'")
+    
+    monkeypatch.setattr("subprocess.run", mock_run)
+    
+    evaluator = HardEvaluator(".")
+    result = evaluator.run_ty()
+    
+    assert result["status"] == "warning"
+    assert "ty executable not found" in result["error_message"]
+
+def test_ty_fallback_behavior_oserror(monkeypatch: Any) -> None:
+    """
+    Test that run_ty gracefully falls back to a warning when a generic Exception 
+    containing the Errno 2 string is thrown.
+    """
+    def mock_run(*args: Any, **kwargs: Any) -> Any:
+        raise Exception("[Errno 2] No such file or directory: 'ty'")
+    
+    monkeypatch.setattr("subprocess.run", mock_run)
+    
+    evaluator = HardEvaluator(".")
+    result = evaluator.run_ty()
+    
+    assert result["status"] == "warning"
+    assert "ty executable not found" in result["error_message"]
+
+def test_radon_cc_syntax_error(monkeypatch: Any, tmp_path: Path) -> None:
+    """
+    Test that run_radon_cc correctly captures and reports stderr when radon 
+    fails (e.g. due to syntax errors in the target codebase).
+    """
+    # Create a mock subprocess.run that simulates radon exiting with code 1
+    # and writing an error to stderr (which happens when there are syntax errors)
+    import subprocess
+    original_run = subprocess.run
+    
+    def mock_run(args: Any, **kwargs: Any) -> Any:
+        if args and args[0] == "radon" and args[1] == "cc":
+            # Simulate radon failing on syntax error
+            class MockResult:
+                returncode = 1
+                stdout = ""
+                stderr = "ERROR: SyntaxError in bad.py"
+            return MockResult()
+        return original_run(args, **kwargs)
+        
+    monkeypatch.setattr("subprocess.run", mock_run)
+    
+    evaluator = HardEvaluator(str(tmp_path))
+    result = evaluator.run_radon_cc()
+    
+    assert result["status"] == "failed"
+    assert len(result["issues"]) == 0
+    # Radon should output to stderr because of the syntax error
+    assert "SyntaxError" in result["error_message"] or result["return_code"] != 0

python_harness-0.0.3/tests/test_hard_evaluator.py

@@ -1,29 +0,0 @@
-"""
-Tests for hard evaluation logic.
-"""
-
-from python_harness.hard_evaluator import HardEvaluator
-
-
-def test_hard_evaluator_methods() -> None:
-    """
-    Test methods of HardEvaluator.
-    """
-    evaluator = HardEvaluator(".")
-    
-    ruff_result = evaluator.run_ruff()
-    assert "status" in ruff_result
-    
-    mypy_result = evaluator.run_mypy()
-    assert "status" in mypy_result
-    
-    ty_result = evaluator.run_ty()
-    assert "status" in ty_result
-
-    # pytest_result = evaluator.run_pytest() # Causes infinite loop when run in test
-    # assert "status" in pytest_result
-    
-    eval_result = evaluator.evaluate()
-    assert "ruff" in eval_result
-    assert "mypy" in eval_result
-    assert "ty" in eval_result