fix(cli): add entrypoint and utf-8 output

Refs #379, #383.

Author: lepture

pyproject.toml

@@ -35,6 +35,9 @@ Documentation = "https://mistune.lepture.com/"
 Source = "https://github.com/lepture/mistune"
 Donate = "https://github.com/sponsors/lepture"
 
+[project.scripts]
+mistune = "mistune.__main__:cli"
+
 [build-system]
 requires = ["setuptools"]
 build-backend = "setuptools.build_meta"

src/mistune/__main__.py

@@ -35,9 +35,10 @@ def _md(args: argparse.Namespace) -> "Markdown":
 
 def _output(text: str, args: argparse.Namespace) -> None:
     if args.output:
-        with open(args.output, "w") as f:
+        with open(args.output, "w", encoding="utf-8") as f:
             f.write(text)
     else:
+        _ensure_stdout_utf8()
         print(text)
 
 
@@ -123,6 +124,12 @@ def cli() -> None:
         sys.exit(1)
 
 
+def _ensure_stdout_utf8() -> None:
+    reconfigure = getattr(sys.stdout, "reconfigure", None)
+    if reconfigure is not None:
+        reconfigure(encoding="utf-8")
+
+
 def read_stdin() -> Optional[str]:
     is_stdin_pipe = not sys.stdin.isatty()
     if is_stdin_pipe:

tests/test_misc.py

@@ -309,3 +309,21 @@ def test_def_list_plugin_redos_candidate(self):
         md = mistune.create_markdown(escape=False, plugins=["def_list"])
         result = md("x\n" * 8000)
         self.assertTrue(result.startswith("<p>x\nx\n"))
+
+    def test_cli_output_file_uses_utf8(self):
+        from argparse import Namespace
+        from pathlib import Path
+        from tempfile import TemporaryDirectory
+
+        from mistune.__main__ import _output
+
+        with TemporaryDirectory() as tmpdir:
+            path = Path(tmpdir) / "out.html"
+            _output("★", Namespace(output=str(path)))
+            self.assertEqual(path.read_text(encoding="utf-8"), "★")
+
+    def test_project_script_entrypoint(self):
+        from pathlib import Path
+
+        pyproject = Path("pyproject.toml").read_text(encoding="utf-8")
+        self.assertIn('mistune = "mistune.__main__:cli"', pyproject)