Windows: type-check the POSIX file lock by platform; read files as UTF-8

GitHub CI (first run): mypy on Windows rejected fcntl.flock in the spending
ledger, because fcntl is POSIX-only. The lock is now taken under a
sys.platform check that mypy understands; runtime behaviour is unchanged
(file lock on POSIX, thread lock on Windows, as documented). Also read JSON
files as UTF-8 explicitly, so Windows does not fall back to its locale code
page (Spider loader, scripts, tests). No change to evaluation behaviour.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01E1YjKkuhoLd53dFQNTgHfH

Author: Amine-LG

intentsql/benchmark/spider_ir.py

@@ -26,7 +26,7 @@ class Unconvertible(ValueError):
 @lru_cache(maxsize=4)
 def _tables(split: str) -> dict[str, dict]:
     path = spider_root() / SPLIT_FILES[split][2]
-    return {db["db_id"]: db for db in json.loads(Path(path).read_text())}
+    return {db["db_id"]: db for db in json.loads(Path(path).read_text(encoding="utf-8"))}
 
 
 class _Converter:

intentsql/spending.py

@@ -21,6 +21,7 @@
 
 import json
 import os
+import sys
 import threading
 import time
 from collections.abc import Iterator
@@ -77,11 +78,9 @@ def _locked(self) -> Iterator[dict[str, Any]]:
                 return
             self.ledger.parent.mkdir(parents=True, exist_ok=True)
             with open(self.ledger.with_suffix(".lock"), "a+") as lock_file:
-                try:
+                if sys.platform != "win32":            # Windows: the thread lock only
                     import fcntl
                     fcntl.flock(lock_file, fcntl.LOCK_EX)
-                except ImportError:                    # Windows: thread lock only
-                    pass
                 try:
                     state = json.loads(self.ledger.read_text(encoding="utf-8"))
                     if not isinstance(state, dict):

scripts/check_ir_coverage.py

@@ -21,7 +21,7 @@
 
 def main(split: str, limit: int | None = None, show: int = 8) -> None:
     root = spider_root()
-    data = json.loads((root / SPLIT_FILES[split][0]).read_text())[:limit]
+    data = json.loads((root / SPLIT_FILES[split][0]).read_text(encoding="utf-8"))[:limit]
     outcome: Counter[str] = Counter()
     shown = 0
     for index, item in enumerate(data):

scripts/failures.py

@@ -4,7 +4,7 @@
 import json
 import sys
 
-report = json.load(open(sys.argv[1]))
+report = json.load(open(sys.argv[1], encoding="utf-8"))
 if report["meta"].get("split") == "test":
     sys.exit("refusing to print held-out test cases")
 filt = sys.argv[2] if len(sys.argv) > 2 else None

scripts/fit_calibration.py

@@ -20,7 +20,7 @@
 def main(provider: str, paths: list[str]) -> None:
     points = []
     for path in paths:
-        report = json.load(open(path))
+        report = json.load(open(path, encoding="utf-8"))
         meta = report.get("meta", {})
         if meta.get("split") != "train":
             sys.exit(f"{path}: calibration is fitted on train reports only")

tests/test_custom_models.py

@@ -115,7 +115,7 @@ def test_add_test_edit_and_remove(self):
         moved = self.add(id="lab-model", api_key=None, url="https://elsewhere.example/decide").json()
         self.assertFalse(self.card(moved)["has_key"])
 
-        stored = json.loads((Path(self.tmp.name) / "connection.json").read_text())
+        stored = json.loads((Path(self.tmp.name) / "connection.json").read_text(encoding="utf-8"))
         if os.name == "posix":                                     # Windows has no owner-only mode bits
             self.assertEqual((Path(self.tmp.name) / "connection.json").stat().st_mode & 0o777, 0o600)
         self.assertNotIn("custom:lab-model", stored["keys"])
@@ -297,7 +297,7 @@ def test_hand_edited_settings_cannot_crash_or_bypass_the_rules(self):
                         json.dumps({"custom": {"evil": {"label": "E", "url": "http://evil.example/x", "model": "m",
                                                         "max_chars": -5}},
                                     "keys": {"custom:evil": "k-0123456789", "typesafe": 12345}})):
-            path.write_text(content)
+            path.write_text(content, encoding="utf-8")
             config.settings.cache_clear()
             response = self.client.get("/api/connection")
             self.assertEqual(response.status_code, 200, content)

tests/test_spending.py

@@ -42,7 +42,7 @@ def test_ledger_persists_and_its_limit_cannot_be_raised(self):
             second.reserve(0.7)
         lower = Budget(0.5, self.ledger)
         self.assertEqual(Budget(None, self.ledger).limit, 0.5)
-        self.assertEqual(json.loads(self.ledger.read_text())["requests"], 1)
+        self.assertEqual(json.loads(self.ledger.read_text(encoding="utf-8"))["requests"], 1)
         self.assertIsNotNone(lower)
 
     def test_concurrent_reservations_never_pass_the_limit(self):