fix(learning): suppress dayfirst-coercion warning in column diff

pandas <2's Series.eq() between an object (string) column and a
datetime64 column implicitly attempts to parse the string side as
dates for comparison, which can emit a dayfirst-ambiguity UserWarning;
pandas >=2 dropped that implicit coercion. Since freshdata's pytest
config promotes freshdata-sourced UserWarnings to errors, this broke
the Python 3.9 (pandas<2) CI matrix job across most of tests/learning/.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Author: JohnnyWilson16

src/freshdata/learning/diff.py

@@ -7,6 +7,8 @@
 
 from __future__ import annotations
 
+import warnings
+
 import pandas as pd
 
 from .types import AlignedPair, DiffSummary, RowDiffSummary, SchemaDiffSummary, ValueDiff
@@ -26,9 +28,13 @@ def _column_diffs(messy: pd.Series, clean: pd.Series, column: str) -> list[Value
     raw_na = messy.isna()
     clean_na = clean.isna()
     # A cell differs when values are unequal, except both-missing which is equal.
-    unequal = ~(messy.eq(clean) | (raw_na & clean_na))
-    # eq() treats NaN != NaN, and mixed dtypes can raise inside eq; the mask
-    # above already covers the common path, but guard object-dtype surprises.
+    # eq() between an object (string) column and a datetime64 column makes
+    # pandas <2 attempt an implicit to_datetime coercion on the string side,
+    # which can emit a dayfirst-ambiguity UserWarning; pandas >=2 dropped that
+    # coercion. Suppress here rather than requiring every caller to know why.
+    with warnings.catch_warnings():
+        warnings.simplefilter("ignore", UserWarning)
+        unequal = ~(messy.eq(clean) | (raw_na & clean_na))
     if not unequal.any():
         return []