From 82a52940157b6b2f35efb28e83f5acb74b1a87b4 Mon Sep 17 00:00:00 2001
From: ABrandes <79119643+AMBRA7592@users.noreply.github.com>
Date: Sun, 19 Jul 2026 06:16:58 +0200
Subject: [PATCH] Adapters: normalize MHS float categories

---
 adapters/mhs.py | 12 +++++++++++-
 test_claims.py  | 36 ++++++++++++++++++++++++++++++++++++
 2 files changed, 47 insertions(+), 1 deletion(-)

diff --git a/adapters/mhs.py b/adapters/mhs.py
index 4073a90..a4ab4a2 100644
--- a/adapters/mhs.py
+++ b/adapters/mhs.py
@@ -7,6 +7,7 @@
 """
 
 import json
+import math
 from collections import Counter, OrderedDict
 
 
@@ -68,7 +69,16 @@ def _label(value, index):
         raise MHSInputError(
             "record {} hatespeech must be null or one of 0, 1, 2".format(index)
         )
-    normalized = str(value)
+    if isinstance(value, float):
+        if not math.isfinite(value) or not value.is_integer():
+            raise MHSInputError(
+                "record {} hatespeech must be null or one of 0, 1, 2".format(
+                    index
+                )
+            )
+        normalized = str(int(value))
+    else:
+        normalized = str(value)
     if normalized not in LABELS:
         raise MHSInputError(
             "record {} hatespeech must be null or one of 0, 1, 2".format(index)
diff --git a/test_claims.py b/test_claims.py
index 399c2a4..378ed14 100644
--- a/test_claims.py
+++ b/test_claims.py
@@ -1411,6 +1411,42 @@ def test_frozen_eligibility_cohorts_primary_filter_and_halt(self):
         self.assertEqual(halted["counts"]["primary_items"], 49)
         self.assertIn("below the frozen minimum 50", halted["halt_reason"])
 
+    def test_label_normalizes_only_finite_integral_parquet_floats(self):
+        self.assertIsNone(mhs_adapter._label(None, 1))
+        for value, expected in (
+            (0.0, "0"),
+            (1.0, "1"),
+            (2.0, "2"),
+            (0, "0"),
+            (1, "1"),
+            (2, "2"),
+            ("0", "0"),
+            ("1", "1"),
+            ("2", "2"),
+        ):
+            self.assertEqual(mhs_adapter._label(value, 1), expected)
+
+        for value in (
+            0.5,
+            3.0,
+            -1.0,
+            math.nan,
+            math.inf,
+            -math.inf,
+            True,
+            False,
+            3,
+            -1,
+            "0.0",
+            "3",
+            "-1",
+        ):
+            with self.subTest(value=value):
+                with self.assertRaisesRegex(
+                    mhs_adapter.MHSInputError, "one of 0, 1, 2"
+                ):
+                    mhs_adapter._label(value, 1)
+
     def test_synthetic_tool_wiring_and_frozen_metric_results(self):
         with tempfile.TemporaryDirectory(prefix="groundless-mhs-") as temp:
             temp_path = Path(temp)