path normalization and tempdir fixes for windows

.github/workflows/e2e-futurehouse-structure.yaml

@@ -12,7 +12,10 @@ jobs:
     # Dynamically determine if environment is needed only when workflow files change and contributor is external
     environment: ${{ (github.event_name == 'workflow_dispatch' || (contains(toJSON(github.event.pull_request.files.*.filename), '.github/workflows/') && github.event.pull_request.user.login != 'misrasaurabh1' && github.event.pull_request.user.login != 'KRRT7')) && 'external-trusted-contributors' || '' }}
 
-    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        os: [ubuntu-latest, windows-latest]
+    runs-on: ${{ matrix.os }}
     env:
       CODEFLASH_AIS_SERVER: prod
       POSTHOG_API_KEY: ${{ secrets.POSTHOG_API_KEY }}
@@ -22,6 +25,7 @@ jobs:
       RETRY_DELAY: 5
       EXPECTED_IMPROVEMENT_PCT: 10
       CODEFLASH_END_TO_END: 1
+      PYTHONIOENCODING: "utf8"
     steps:
       - name: 🛎️ Checkout
         uses: actions/checkout@v4
@@ -31,6 +35,7 @@ jobs:
           fetch-depth: 0
           token: ${{ secrets.GITHUB_TOKEN }}
       - name: Validate PR
+        shell: bash
         run: |
           # Check for any workflow changes
           if git diff --name-only "${{ github.event.pull_request.base.sha }}" "${{ github.event.pull_request.head.sha }}" | grep -q "^.github/workflows/"; then
@@ -56,13 +61,16 @@ jobs:
       - name: Set up Python 3.11 for CLI
         uses: astral-sh/setup-uv@v5
         with:
-          python-version: 3.11.6
+          python-version: 3.11
 
       - name: Install dependencies (CLI)
         run: |
           uv sync
+          uv add ruff
 
       - name: Run Codeflash to optimize code
         id: optimize_code
+        env:
+          PYTHONUTF8: 1
         run: |
-          uv run python tests/scripts/end_to_end_test_futurehouse.py
+          uv run tests/scripts/end_to_end_test_futurehouse.py

.github/workflows/unit-tests.yaml

@@ -12,8 +12,9 @@ jobs:
       fail-fast: false
       matrix:
         python-version: ["3.9", "3.10", "3.11", "3.12"]
+        os: [ubuntu-latest]
     continue-on-error: true
-    runs-on: ubuntu-latest
+    runs-on: ${{ matrix.os }}
     steps:
       - uses: actions/checkout@v4
         with:
@@ -24,10 +25,29 @@ jobs:
         uses: astral-sh/setup-uv@v5
         with:
           python-version: ${{ matrix.python-version }}
-          version: "0.5.30"
 
       - name: install dependencies
         run: uv sync
 
       - name: Unit tests
-        run: uv run pytest tests/ --benchmark-skip -m "not ci_skip"
+        run: uv run pytest tests/ --benchmark-skip -m "not ci_skip"
+
+  # unit-tests-windows:
+  #   runs-on: windows-latest
+  #   continue-on-error: true
+  #   steps:
+  #     - uses: actions/checkout@v4
+  #       with:
+  #         fetch-depth: 0
+  #         token: ${{ secrets.GITHUB_TOKEN }}
+
+  #     - name: Install uv
+  #       uses: astral-sh/setup-uv@v5
+  #       with:
+  #         python-version: "3.11"
+
+  #     - name: install dependencies
+  #       run: uv sync
+
+  #     - name: Unit tests
+  #       run: uv run pytest tests/ --benchmark-skip -m "not ci_skip"

codeflash/benchmarking/replay_test.py

@@ -162,7 +162,7 @@ def create_trace_replay_test_code(
                 benchmark_function_name=benchmark_function_name,
                 orig_function_name=function_name,
                 function_name=alias,
-                file_path=file_path,
+                file_path=Path(file_path).as_posix(),
                 max_run_count=max_run_count,
             )
         else:
@@ -176,7 +176,7 @@ def create_trace_replay_test_code(
                 test_body = test_class_method_body.format(
                     benchmark_function_name=benchmark_function_name,
                     orig_function_name=function_name,
-                    file_path=file_path,
+                    file_path=Path(file_path).as_posix(),
                     class_name_alias=class_name_alias,
                     class_name=class_name,
                     method_name=method_name,
@@ -187,7 +187,7 @@ def create_trace_replay_test_code(
                 test_body = test_static_method_body.format(
                     benchmark_function_name=benchmark_function_name,
                     orig_function_name=function_name,
-                    file_path=file_path,
+                    file_path=Path(file_path).as_posix(),
                     class_name_alias=class_name_alias,
                     class_name=class_name,
                     method_name=method_name,
@@ -198,7 +198,7 @@ def create_trace_replay_test_code(
                 test_body = test_method_body.format(
                     benchmark_function_name=benchmark_function_name,
                     orig_function_name=function_name,
-                    file_path=file_path,
+                    file_path=Path(file_path).as_posix(),
                     class_name_alias=class_name_alias,
                     class_name=class_name,
                     method_name=method_name,

codeflash/code_utils/code_utils.py

@@ -171,8 +171,9 @@ def get_run_tmp_file(file_path: Path) -> Path:
 
 
 def path_belongs_to_site_packages(file_path: Path) -> bool:
+    file_path_resolved = file_path.resolve()
     site_packages = [Path(p) for p in site.getsitepackages()]
-    return any(file_path.resolve().is_relative_to(site_package_path) for site_package_path in site_packages)
+    return any(file_path_resolved.is_relative_to(site_package_path) for site_package_path in site_packages)
 
 
 def is_class_defined_in_file(class_name: str, file_path: Path) -> bool:

codeflash/code_utils/coverage_utils.py

@@ -45,7 +45,7 @@ def generate_candidates(source_code_path: Path) -> list[str]:
     current_path = source_code_path.parent
 
     while current_path != current_path.parent:
-        candidate_path = str(Path(current_path.name) / candidates[-1])
+        candidate_path = (Path(current_path.name) / candidates[-1]).as_posix()
         candidates.append(candidate_path)
         current_path = current_path.parent
 

codeflash/code_utils/env_utils.py

@@ -18,18 +18,17 @@ def check_formatter_installed(formatter_cmds: list[str], exit_on_failure: bool =
     if formatter_cmds[0] == "disabled":
         return return_code
     tmp_code = """print("hello world")"""
-    with tempfile.NamedTemporaryFile(mode="w", encoding="utf-8", suffix=".py") as f:
-        f.write(tmp_code)
-        f.flush()
-        tmp_file = Path(f.name)
+    with tempfile.TemporaryDirectory() as tmpdir:
+        tmp_file = Path(tmpdir) / "test_codeflash_formatter.py"
+        tmp_file.write_text(tmp_code, encoding="utf-8")
         try:
             format_code(formatter_cmds, tmp_file, print_status=False, exit_on_failure=exit_on_failure)
         except Exception:
             exit_with_message(
                 "⚠️ Codeflash requires a code formatter to be installed in your environment, but none was found. Please install a supported formatter, verify the formatter-cmds in your codeflash pyproject.toml config and try again.",
                 error_on_exit=True,
             )
-    return return_code
+        return return_code
 
 
 @lru_cache(maxsize=1)

codeflash/code_utils/instrument_existing_tests.py

@@ -212,7 +212,9 @@ def visit_FunctionDef(self, node: ast.FunctionDef, test_class_name: str | None =
                                     args=[
                                         ast.JoinedStr(
                                             values=[
-                                                ast.Constant(value=f"{get_run_tmp_file(Path('test_return_values_'))}"),
+                                                ast.Constant(
+                                                    value=f"{get_run_tmp_file(Path('test_return_values_')).as_posix()}"
+                                                ),
                                                 ast.FormattedValue(
                                                     value=ast.Name(id="codeflash_iteration", ctx=ast.Load()),
                                                     conversion=-1,

codeflash/models/models.py

@@ -147,7 +147,7 @@ def markdown(self) -> str:
         """Returns the markdown representation of the code, including the file path where possible."""
         return "\n".join(
             [
-                f"```python{':' + str(code_string.file_path) if code_string.file_path else ''}\n{code_string.code.strip()}\n```"
+                f"```python{':' + code_string.file_path.as_posix() if code_string.file_path else ''}\n{code_string.code.strip()}\n```"
                 for code_string in self.code_strings
             ]
         )

codeflash/result/create_pr.py

@@ -101,7 +101,7 @@ def existing_tests_source_for(
                 if greater:
                     rows.append(
                         [
-                            f"`{print_filename}::{qualified_name}`",
+                            f"`{print_filename.as_posix()}::{qualified_name}`",
                             f"{print_original_runtime}",
                             f"{print_optimized_runtime}",
                             f"⚠️{perf_gain}%",
@@ -110,7 +110,7 @@ def existing_tests_source_for(
                 else:
                     rows.append(
                         [
-                            f"`{print_filename}::{qualified_name}`",
+                            f"`{print_filename.as_posix()}::{qualified_name}`",
                             f"{print_original_runtime}",
                             f"{print_optimized_runtime}",
                             f"✅{perf_gain}%",

codeflash/verification/instrument_codeflash_capture.py

@@ -33,7 +33,7 @@ def instrument_codeflash_capture(
     modified_code = add_codeflash_capture_to_init(
         target_classes={class_parent.name},
         fto_name=function_to_optimize.function_name,
-        tmp_dir_path=str(get_run_tmp_file(Path("test_return_values"))),
+        tmp_dir_path=get_run_tmp_file(Path("test_return_values")).as_posix(),
         code=original_code,
         tests_root=tests_root,
         is_fto=True,
@@ -46,7 +46,7 @@ def instrument_codeflash_capture(
         modified_code = add_codeflash_capture_to_init(
             target_classes=helper_classes,
             fto_name=function_to_optimize.function_name,
-            tmp_dir_path=str(get_run_tmp_file(Path("test_return_values"))),
+            tmp_dir_path=get_run_tmp_file(Path("test_return_values")).as_posix(),
             code=original_code,
             tests_root=tests_root,
             is_fto=False,
@@ -124,7 +124,7 @@ def visit_ClassDef(self, node: ast.ClassDef) -> ast.ClassDef:
             keywords=[
                 ast.keyword(arg="function_name", value=ast.Constant(value=f"{node.name}.__init__")),
                 ast.keyword(arg="tmp_dir_path", value=ast.Constant(value=self.tmp_dir_path)),
-                ast.keyword(arg="tests_root", value=ast.Constant(value=str(self.tests_root))),
+                ast.keyword(arg="tests_root", value=ast.Constant(value=self.tests_root.as_posix())),
                 ast.keyword(arg="is_fto", value=ast.Constant(value=self.is_fto)),
             ],
         )

tests/scripts/end_to_end_test_utilities.py

@@ -94,6 +94,7 @@ def run_codeflash_command(
 
     output = []
     for line in process.stdout:
+        line = line.strip().encode("utf-8").decode("utf-8")
         logging.info(line.strip())
         output.append(line)
 
@@ -122,7 +123,7 @@ def build_command(
 ) -> list[str]:
     python_path = "../../../codeflash/main.py" if "code_directories" in str(cwd) else "../codeflash/main.py"
 
-    base_command = ["python", python_path, "--file", config.file_path, "--no-pr"]
+    base_command = ["uv", "run", "--no-project",  python_path, "--file", config.file_path, "--no-pr"]
 
     if config.function_name:
         base_command.extend(["--function", config.function_name])
@@ -187,13 +188,14 @@ def validate_stdout_in_candidate(stdout: str, expected_in_stdout: list[str]) ->
 def run_trace_test(cwd: pathlib.Path, config: TestConfig, expected_improvement_pct: int) -> bool:
     test_root = cwd / "tests" / (config.test_framework or "")
     clear_directory(test_root)
-    command = ["python", "-m", "codeflash.main", "optimize", "workload.py"]
+    command = ["uv", "run", "--no-project", "-m", "codeflash.main", "optimize", "workload.py"]
     process = subprocess.Popen(
         command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True, cwd=str(cwd), env=os.environ.copy()
     )
 
     output = []
     for line in process.stdout:
+        line = line.strip().encode("utf-8").decode("utf-8")
         logging.info(line.strip())
         output.append(line)