Fix failed pyproject.toml read when Unicode characters present (#1422)

Copilot · nathanjmcdougall · web-flow · commit c15d4588be8c · 2026-03-19T15:29:17.000+13:00
* Initial plan

* Fix unicode encoding issue in file read/write operations

Co-authored-by: nathanjmcdougall &lt;18602289+nathanjmcdougall@users.noreply.github.com&gt;

---------

Co-authored-by: copilot-swe-agent[bot] &lt;198982749+Copilot@users.noreply.github.com&gt;
Co-authored-by: nathanjmcdougall &lt;18602289+nathanjmcdougall@users.noreply.github.com&gt;
diff --git a/src/usethis/_file/yaml/io_.py b/src/usethis/_file/yaml/io_.py
@@ -449,7 +449,7 @@ def read_yaml(
     guess_indent: bool = True,
 ) -> Generator[YAMLDocument, None, None]:
     """A context manager to read a YAML file."""
-    with yaml_path.open(mode="r") as f:
+    with yaml_path.open(mode="r", encoding="utf-8") as f:
         try:
             yaml_document = _get_yaml_document(f, guess_indent=guess_indent)
         except YAMLError as err:
diff --git a/src/usethis/_init.py b/src/usethis/_init.py
@@ -44,7 +44,8 @@ def project_init():
                 f"""\
 def hello() -> str:
     return "Hello from {project_name}!"
-"""
+""",
+                encoding="utf-8",
             )
         (src_dir / pkg_name / "py.typed").touch(exist_ok=True)
     else:
@@ -115,7 +116,8 @@ def ensure_pyproject_toml(*, author: bool = True) -> None:
 [build-system]
 requires = ["hatchling"]
 build-backend = "hatchling.build"
-"""
+""",
+            encoding="utf-8",
         )
     else:
         assert_never(backend)
diff --git a/src/usethis/_integrations/mkdocs/core.py b/src/usethis/_integrations/mkdocs/core.py
@@ -21,5 +21,6 @@ def add_docs_dir() -> None:
 # {get_project_name()}
 
 Welcome to the documentation for {get_project_name()}.
-"""
+""",
+            encoding="utf-8",
         )
diff --git a/src/usethis/_integrations/pytest/core.py b/src/usethis/_integrations/pytest/core.py
@@ -19,7 +19,7 @@ def add_pytest_dir() -> None:
 
     tick_print("Writing '/tests/conftest.py'.")
     (tests_dir / "conftest.py").write_text(
-        "collect_ignore_glob = []\npytest_plugins = []\n"
+        "collect_ignore_glob = []\npytest_plugins = []\n", encoding="utf-8"
     )
 
 
diff --git a/src/usethis/_integrations/sonarqube/config.py b/src/usethis/_integrations/sonarqube/config.py
@@ -24,7 +24,9 @@ def get_sonar_project_properties() -> str:
     # Get Python version
     try:
         python_version = PythonVersion.from_string(
-            (usethis_config.cpd() / ".python-version").read_text().strip()
+            (usethis_config.cpd() / ".python-version")
+            .read_text(encoding="utf-8")
+            .strip()
         ).to_short_string()
     except (FileNotFoundError, PythonVersionParseError):
         python_version = PythonVersion.from_interpreter().to_short_string()
diff --git a/src/usethis/_io.py b/src/usethis/_io.py
@@ -116,7 +116,7 @@ def write_file(self) -> None:
         if not self.path.exists():
             return
 
-        self.path.write_text(self._dump_content())
+        self.path.write_text(self._dump_content(), encoding="utf-8")
 
     def read_file(self) -> DocumentT:
         """Read the document from disk and store it in memory.
@@ -132,7 +132,7 @@ def read_file(self) -> DocumentT:
             )
             raise UnexpectedFileIOError(msg)
         try:
-            document = self._parse_content(self.path.read_text())
+            document = self._parse_content(self.path.read_text(encoding="utf-8"))
         except FileNotFoundError:
             msg = f"'{self.name}' not found in the current directory at '{self.path}'."
             raise FileNotFoundError(msg) from None
diff --git a/tests/usethis/_file/pyproject_toml/test_pyproject_toml_io_.py b/tests/usethis/_file/pyproject_toml/test_pyproject_toml_io_.py
@@ -76,6 +76,23 @@ def test_read_file_invalid_toml(self, tmp_path: Path):
         with pytest.raises(PyprojectTOMLDecodeError), manager:
             manager.read_file()
 
+    def test_read_file_with_unicode_chars(self, tmp_path: Path):
+        # Unicode characters like 'ā' in comments should not cause read failures.
+
+        # Arrange
+        with change_cwd(tmp_path):
+            (tmp_path / "pyproject.toml").write_text(
+                '# Comment with unicode: ā\n[project]\nname = "foo"\n',
+                encoding="utf-8",
+            )
+            manager = PyprojectTOMLManager()
+
+        # Act & Assert (should not raise)
+        with manager:
+            result = manager.get()
+
+        assert result.value["project"]["name"] == "foo"
+
     def test_commit_and_get(self, tmp_path: Path):
         with change_cwd(tmp_path):
             # Arrange

Original file line number	Diff line number	Diff line change
`@@ -21,5 +21,6 @@ def add_docs_dir() -> None:`
`21`	`21`	`# {get_project_name()}`
`22`	`22`
`23`	`23`	`Welcome to the documentation for {get_project_name()}.`
`24`		`-"""`
	`24`	`+""",`
	`25`	`+ encoding="utf-8",`
`25`	`26`	`)`
Original file line number	Diff line number	Diff line change
`@@ -19,7 +19,7 @@ def add_pytest_dir() -> None:`
`19`	`19`
`20`	`20`	`tick_print("Writing '/tests/conftest.py'.")`
`21`	`21`	`(tests_dir / "conftest.py").write_text(`
`22`		`- "collect_ignore_glob = []\npytest_plugins = []\n"`
	`22`	`+ "collect_ignore_glob = []\npytest_plugins = []\n", encoding="utf-8"`
`23`	`23`	`)`
`24`	`24`
`25`	`25`