fix(tools): write progress history files with utf-8 encoding

tools/progress_history.py

@@ -26,7 +26,7 @@ def git(*args):
 
 def daily_commits(ref):
     seen = {}
-    for line in git("log", "--first-parent", "--format=%H %ct", ref).decode().splitlines():
+    for line in git("log", "--first-parent", "--format=%H %ct", ref).decode("utf-8").splitlines():
         sha, timestamp = line.split()
         seen.setdefault(datetime.fromtimestamp(int(timestamp), timezone.utc).date().isoformat(), sha)
     return sorted(seen.items())
@@ -221,12 +221,12 @@ def page(points):
     ]) + "\n"
 
 
-if __name__ == "__main__":
+def main(argv=None):
     parser = argparse.ArgumentParser()
     parser.add_argument("output", type=Path)
     parser.add_argument("--ref", default="HEAD", help="branch or commit whose first-parent history is measured")
     parser.add_argument("--cache", type=Path, help="reuse per-commit counts while the scanner, ISA list and Python version match")
-    args = parser.parse_args()
+    args = parser.parse_args(argv)
     try:
         if args.cache:
             outputs = [args.output / name for name in ("history.json", "history.svg", "history.html")]
@@ -239,7 +239,11 @@ if __name__ == "__main__":
     if not points:
         parser.error("no commit with a measurable tree")
     args.output.mkdir(parents=True, exist_ok=True)
-    (args.output / "history.json").write_text(json.dumps(points, indent=2))
-    (args.output / "history.svg").write_text(chart(points) + "\n")
-    (args.output / "history.html").write_text(page(points))
+    (args.output / "history.json").write_text(json.dumps(points, indent=2), encoding="utf-8")
+    (args.output / "history.svg").write_text(chart(points) + "\n", encoding="utf-8")
+    (args.output / "history.html").write_text(page(points), encoding="utf-8")
     print(f"{len(points)} days, {points[0]['date']} .. {points[-1]['date']}")
+
+
+if __name__ == "__main__":
+    main()

tools/tests/test_progress_history.py

@@ -1,3 +1,4 @@
+import json
 import os
 import subprocess
 import sys
@@ -17,9 +18,9 @@ OPCODES = Path("core/shader/recompiler/RdnaDecoder/include/RdnaDecoder/RdnaOpcod
 def commit(root, day, message, body, time="12:00:00Z"):
     library = root / "core/libs/prx/libSceExample"
     library.mkdir(parents=True, exist_ok=True)
-    (library / "Export.cpp").write_text(body)
+    (library / "Export.cpp").write_text(body, encoding="utf-8")
     (root / OPCODES).parent.mkdir(parents=True, exist_ok=True)
-    (root / OPCODES).write_text("enum class RdnaOpcode : int {\n    SEndpgm,\n};\n")
+    (root / OPCODES).write_text("enum class RdnaOpcode : int {\n    SEndpgm,\n};\n", encoding="utf-8")
     env = {**os.environ, "GIT_AUTHOR_DATE": f"{day}T{time}", "GIT_COMMITTER_DATE": f"{day}T{time}"}
     subprocess.run(["git", "-C", str(root), "add", "-A"], check=True)
     subprocess.run(["git", "-C", str(root), "-c", "user.name=t", "-c", "user.email=t@t", "commit", "-q", "-m", message],
@@ -140,6 +141,21 @@ class ProgressHistoryTests(unittest.TestCase):
         for label in ("Aug 3", "Aug 10", "Aug 17", "Aug 24", "Aug 31"):
             self.assertIn(f">{label}<", svg)
 
+    def test_main_writes_utf8_output_files(self):
+        point = {"date": "2026-01-01", "commit": "abc1234",
+                 "libraries": {"done": 1, "total": 2, "percent": 50.0},
+                 "shaders": {"done": 3, "total": 4, "percent": 75.0}}
+        with tempfile.TemporaryDirectory() as directory:
+            output = Path(directory) / "site"
+            with patch.object(progress_history, "collect", return_value=[point]):
+                progress_history.main([str(output)])
+            svg = (output / "history.svg").read_text(encoding="utf-8")
+            self.assertIn("\u25cf", svg)
+            html = (output / "history.html").read_text(encoding="utf-8")
+            self.assertIn("<!DOCTYPE html>", html)
+            data = json.loads((output / "history.json").read_text(encoding="utf-8"))
+            self.assertEqual(data, [point])
+
 
 
 if __name__ == "__main__":