fix: make sure encoding is specified in more places

Signed-off-by: Henry Schreiner <[email protected]>

Author: henryiii

src/scikit_build_core/build/_editable.py

@@ -36,7 +36,7 @@ def editable_redirect(
     """
 
     editable_py = resources / "_editable_redirect.py"
-    editable_txt: str = editable_py.read_text(encoding="utf-8")
+    editable_txt: str = editable_py.read_text(encoding="utf-8-sig")
 
     arguments = (
         modules,

src/scikit_build_core/build/_file_processor.py

@@ -40,11 +40,13 @@ def each_unignored_file(
 
     for gi in [Path(".git/info/exclude"), Path(".gitignore")]:
         ignore_errs = [FileNotFoundError, NotADirectoryError]
-        with contextlib.suppress(*ignore_errs), gi.open(encoding="utf-8") as f:
+        with contextlib.suppress(*ignore_errs), gi.open(encoding="utf-8-sig") as f:
             exclude_lines += f.readlines()
 
     nested_excludes = {
-        p.parent: pathspec.GitIgnoreSpec.from_lines(p.read_text().splitlines())
+        p.parent: pathspec.GitIgnoreSpec.from_lines(
+            p.read_text(encoding="utf-8-sig").splitlines()
+        )
         for p in Path().rglob("**/.gitignore")
         if p != Path(".gitignore")
     }

src/scikit_build_core/build/_scripts.py

@@ -20,7 +20,9 @@ def __dir__() -> list[str]:
 def process_script_dir(script_dir: Path) -> None:
     for item in script_dir.iterdir():
         content = []
-        with contextlib.suppress(UnicodeDecodeError), item.open(encoding="utf-8") as f:
+        with contextlib.suppress(UnicodeDecodeError), item.open(
+            encoding="utf-8-sig"
+        ) as f:
             file_iter = iter(f)
             try:
                 # TODO: handle empty files
@@ -31,5 +33,5 @@ def process_script_dir(script_dir: Path) -> None:
             if match:
                 content = [f"#!python{match.group(1) or ''}\n", *file_iter]
         if content:
-            with item.open("w", encoding="utf-8") as f:
+            with item.open("w", encoding="utf-8-sig") as f:
                 f.writelines(content)

src/scikit_build_core/build/generate.py

@@ -26,7 +26,7 @@ def generate_file_contents(gen: GenerateSettings, metadata: StandardMetadata) ->
     ), f"One of template or template-path must be set for {gen.path}"
 
     if gen.template_path:
-        template = gen.template_path.read_text(encoding="utf-8")
+        template = gen.template_path.read_text(encoding="utf-8-sig")
     else:
         template = gen.template
 

src/scikit_build_core/cmake.py

@@ -100,7 +100,7 @@ def __post_init__(self) -> None:
             skbuild_info = self.build_dir / ".skbuild-info.json"
             # If building via SDist, this could be pre-filled, so delete it if it exists
             with contextlib.suppress(FileNotFoundError):
-                with skbuild_info.open("r", encoding="utf-8") as f:
+                with skbuild_info.open("r", encoding="utf-8-sig") as f:
                     info = json.load(f)
 
                 cached_source_dir = Path(info["source_dir"])
@@ -113,7 +113,7 @@ def __post_init__(self) -> None:
                     shutil.rmtree(self.build_dir)
                     self.build_dir.mkdir()
 
-            with skbuild_info.open("w", encoding="utf-8") as f:
+            with skbuild_info.open("w", encoding="utf-8-sig") as f:
                 json.dump(self._info_dict(), f, indent=2)
 
     def _info_dict(self) -> dict[str, str]:
@@ -132,7 +132,7 @@ def _info_dict(self) -> dict[str, str]:
     def init_cache(
         self, cache_settings: Mapping[str, str | os.PathLike[str] | bool]
     ) -> None:
-        with self.init_cache_file.open("w", encoding="utf-8") as f:
+        with self.init_cache_file.open("w", encoding="utf-8-sig") as f:
             for key, value in cache_settings.items():
                 if isinstance(value, bool):
                     str_value = "ON" if value else "OFF"
@@ -162,7 +162,7 @@ def init_cache(
                 )
                 f.write('set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE "BOTH" CACHE PATH "")\n')
 
-        contents = self.init_cache_file.read_text(encoding="utf-8").strip()
+        contents = self.init_cache_file.read_text(encoding="utf-8-sig").strip()
         logger.debug(
             "{}:\n{}",
             self.init_cache_file,

src/scikit_build_core/file_api/_cattrs_converter.py

@@ -40,7 +40,7 @@ def from_json_file(with_path: Dict[str, Any], t: Type[T]) -> T:
         if with_path["jsonFile"] is None:
             return converter.structure_attrs_fromdict({}, t)
         path = base_dir / Path(with_path["jsonFile"])
-        raw = json.loads(path.read_text(encoding="utf-8"))
+        raw = json.loads(path.read_text(encoding="utf-8-sig"))
         return converter.structure_attrs_fromdict(raw, t)
 
     converter.register_structure_hook(CodeModel, from_json_file)
@@ -57,7 +57,7 @@ def load_reply_dir(reply_dir: Path) -> Index:
         msg = f"index file not found in {reply_dir}"
         raise IndexError(msg)
     index_file = indexes[-1]
-    return converter.loads(index_file.read_text(), Index)
+    return converter.loads(index_file.read_text("utf-8-sig"), Index)
 
 
 if __name__ == "__main__":

src/scikit_build_core/file_api/reply.py

@@ -34,13 +34,13 @@ def load(self) -> Index:
         Load the newest index.json file and return the Index object.
         """
         index_file = sorted(self.base_dir.glob("index-*"))[-1]
-        with index_file.open(encoding="utf-8") as f:
+        with index_file.open(encoding="utf-8-sig") as f:
             data = json.load(f)
 
         return self.make_class(data, Index)
 
     def _load_from_json(self, name: Path, target: Type[T]) -> T:
-        with self.base_dir.joinpath(name).open(encoding="utf-8") as f:
+        with self.base_dir.joinpath(name).open(encoding="utf-8-sig") as f:
             data = json.load(f)
 
         return self.make_class(data, target)

src/scikit_build_core/metadata/regex.py

@@ -48,7 +48,7 @@ def dynamic_metadata(
     assert isinstance(result, str)
     remove = settings.get("result", "")
 
-    with Path(input_filename).open(encoding="utf-8") as f:
+    with Path(input_filename).open(encoding="utf-8-sig") as f:
         match = re.search(regex, f.read(), re.MULTILINE)
 
     if not match:

src/scikit_build_core/settings/skbuild_schema.py

@@ -212,7 +212,7 @@ def get_skbuild_schema(tool_name: str = "scikit-build") -> dict[str, Any]:
     "Get the stored complete schema for scikit-build settings."
     assert tool_name == "scikit-build", "Only scikit-build is supported."
 
-    with resources.joinpath("scikit-build.schema.json").open(encoding="utf-8") as f:
+    with resources.joinpath("scikit-build.schema.json").open(encoding="utf-8-sig") as f:
         return json.load(f)  # type: ignore[no-any-return]
 
 

tests/packages/navigate_editable/python/shared_pkg/py_module.py

@@ -19,10 +19,10 @@ def py_method():
 def read_py_data_txt():
     root = files("shared_pkg.data")
     py_data = root / "py_data.txt"
-    print(py_data.read_text())
+    print(py_data.read_text(encoding="utf-8-sig"))
 
 
 def read_c_generated_txt():
     root = files("shared_pkg.data")
     c_generated_txt = root / "c_generated.txt"
-    print(c_generated_txt.read_text())
+    print(c_generated_txt.read_text(encoding="utf-8-sig"))

tests/test_cmake_config.py

@@ -107,7 +107,7 @@ def test_init_cache(
     cmake_init = config.build_dir / "CMakeInit.txt"
     source_dir_str = str(config.source_dir).replace("\\", "/")
     assert (
-        cmake_init.read_text()
+        cmake_init.read_text(encoding="utf-8-sig")
         == f"""\
 set(SKBUILD ON CACHE BOOL "" FORCE)
 set(SKBUILD_VERSION [===[1.0.0]===] CACHE STRING "" FORCE)

tests/test_custom_modules.py

@@ -20,7 +20,7 @@ def test_ep(isolated):
 
     script = isolated.run("script1", capture=True).strip()
     pysys = isolated.execute("import sys; print(sys.executable)").strip()
-    contents = Path(script).read_text()
+    contents = Path(script).read_text(encoding="utf-8-sig")
     assert contents.startswith(f"#!{pysys}")
 
     if sys.version_info >= (3, 8):

tests/test_editable_unit.py

@@ -84,7 +84,7 @@ def editable_package(
 
             from importlib.resources import files
 
-            read_file = files("pkg.resources").joinpath("file.txt").read_text(encoding="utf-8")
+            read_file = files("pkg.resources").joinpath("file.txt").read_text(encoding="utf-8-sig")
             assert read_file == "hello"
             """
         )
@@ -98,7 +98,7 @@ def editable_package(
             """\
             from importlib.resources import files
 
-            read_file = files("pkg.iresources").joinpath("file.txt").read_text(encoding="utf-8")
+            read_file = files("pkg.iresources").joinpath("file.txt").read_text(encoding="utf-8-sig")
             assert read_file == "hi"
             """
         )

tests/test_process_scripts.py

@@ -33,18 +33,35 @@ def test_script_dir(tmp_path: Path) -> None:
 
     process_script_dir(script_dir)
 
-    assert script_1.read_text() == "#!python\n\nprint('hello world')"
+    assert (
+        script_1.read_text(encoding="utf-8-sig") == "#!python\n\nprint('hello world')"
+    )
     assert script_1.stat().st_mode == orig_mode_1
 
-    assert script_2.read_text() == "#!python\n\nprint('hello world')"
+    assert (
+        script_2.read_text(encoding="utf-8-sig") == "#!python\n\nprint('hello world')"
+    )
     assert script_2.stat().st_mode == orig_mode_2
 
-    assert script_3.read_text() == "#!python\n\nprint('hello world')"
+    assert (
+        script_3.read_text(encoding="utf-8-sig") == "#!python\n\nprint('hello world')"
+    )
 
-    assert script_4.read_text() == "#!python\n\nprint('hello world')"
+    assert (
+        script_4.read_text(encoding="utf-8-sig") == "#!python\n\nprint('hello world')"
+    )
 
-    assert script_5.read_text() == "#!/usr/bin/other\n\nprint('hello world')"
+    assert (
+        script_5.read_text(encoding="utf-8-sig")
+        == "#!/usr/bin/other\n\nprint('hello world')"
+    )
 
-    assert script_6.read_text() == "#!python other\n\nprint('hello world')"
+    assert (
+        script_6.read_text(encoding="utf-8-sig")
+        == "#!python other\n\nprint('hello world')"
+    )
 
-    assert script_7.read_text() == "#!/usr/bin/env other\n\nprint('hello world')"
+    assert (
+        script_7.read_text(encoding="utf-8-sig")
+        == "#!/usr/bin/env other\n\nprint('hello world')"
+    )