From f6c00ff376a7affe561e44f4c1af09f717262016 Mon Sep 17 00:00:00 2001
From: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>
Date: Tue, 20 Feb 2024 05:11:08 -1000
Subject: [PATCH] Deprecate datelike isin casting strings to dates to match
 pandas 2.2 (#15046)

Matching https://github.com/pandas-dev/pandas/pull/56427

Authors:
  - Matthew Roeschke (https://github.com/mroeschke)

Approvers:
  - GALI PREM SAGAR (https://github.com/galipremsagar)

URL: https://github.com/rapidsai/cudf/pull/15046
---
 python/cudf/cudf/core/tools/datetimes.py | 10 +++++++
 python/cudf/cudf/tests/test_index.py     | 34 ++++++++++++------------
 python/cudf/cudf/tests/test_series.py    |  8 ++++--
 3 files changed, 33 insertions(+), 19 deletions(-)

diff --git a/python/cudf/cudf/core/tools/datetimes.py b/python/cudf/cudf/core/tools/datetimes.py
index 529296da6a2..0e0df4ecf6e 100644
--- a/python/cudf/cudf/core/tools/datetimes.py
+++ b/python/cudf/cudf/core/tools/datetimes.py
@@ -767,10 +767,20 @@ def _isin_datetimelike(
     rhs = None
     try:
         rhs = cudf.core.column.as_column(values)
+        was_string = len(rhs) and rhs.dtype.kind == "O"
 
         if rhs.dtype.kind in {"f", "i", "u"}:
             return cudf.core.column.full(len(lhs), False, dtype="bool")
         rhs = rhs.astype(lhs.dtype)
+        if was_string:
+            warnings.warn(
+                f"The behavior of 'isin' with dtype={lhs.dtype} and "
+                "castable values (e.g. strings) is deprecated. In a "
+                "future version, these will not be considered matching "
+                "by isin. Explicitly cast to the appropriate dtype before "
+                "calling isin instead.",
+                FutureWarning,
+            )
         res = lhs._isin_earlystop(rhs)
         if res is not None:
             return res
diff --git a/python/cudf/cudf/tests/test_index.py b/python/cudf/cudf/tests/test_index.py
index 3cbfea8063f..defd42b3d00 100644
--- a/python/cudf/cudf/tests/test_index.py
+++ b/python/cudf/cudf/tests/test_index.py
@@ -2497,19 +2497,12 @@ def test_index_nan_as_null(data, nan_idx, NA_idx, nan_as_null):
 
 
 @pytest.mark.parametrize(
-    "data",
+    "index",
     [
-        [],
-        pd.Series(
-            ["this", "is", None, "a", "test"], index=["a", "b", "c", "d", "e"]
-        ),
-        pd.Series([0, 15, 10], index=[0, None, 9]),
-        pd.Series(
-            range(25),
-            index=pd.date_range(
-                start="2019-01-01", end="2019-01-02", freq="h"
-            ),
-        ),
+        pd.Index([]),
+        pd.Index(["a", "b", "c", "d", "e"]),
+        pd.Index([0, None, 9]),
+        pd.date_range("2019-01-01", periods=3),
     ],
 )
 @pytest.mark.parametrize(
@@ -2521,12 +2514,19 @@ def test_index_nan_as_null(data, nan_idx, NA_idx, nan_as_null):
         ["2019-01-01 04:00:00", "2019-01-01 06:00:00", "2018-03-02 10:00:00"],
     ],
 )
-def test_isin_index(data, values):
-    psr = pd.Series(data)
-    gsr = cudf.Series.from_pandas(psr)
+def test_isin_index(index, values):
+    pidx = index
+    gidx = cudf.Index.from_pandas(pidx)
 
-    got = gsr.index.isin(values)
-    expected = psr.index.isin(values)
+    is_dt_str = (
+        next(iter(values), None) == "2019-01-01 04:00:00"
+        and len(pidx)
+        and pidx.dtype.kind == "M"
+    )
+    with expect_warning_if(is_dt_str):
+        got = gidx.isin(values)
+    with expect_warning_if(PANDAS_GE_220 and is_dt_str):
+        expected = pidx.isin(values)
 
     assert_eq(got, expected)
 
diff --git a/python/cudf/cudf/tests/test_series.py b/python/cudf/cudf/tests/test_series.py
index 14006f90b45..252343391be 100644
--- a/python/cudf/cudf/tests/test_series.py
+++ b/python/cudf/cudf/tests/test_series.py
@@ -15,6 +15,7 @@
 
 import cudf
 from cudf.api.extensions import no_default
+from cudf.core._compat import PANDAS_GE_220
 from cudf.errors import MixedTypeError
 from cudf.testing._utils import (
     NUMERIC_TYPES,
@@ -1795,8 +1796,11 @@ def test_isin_datetime(data, values):
     psr = pd.Series(data)
     gsr = cudf.Series.from_pandas(psr)
 
-    got = gsr.isin(values)
-    expected = psr.isin(values)
+    is_len_str = isinstance(next(iter(values), None), str) and len(data)
+    with expect_warning_if(is_len_str):
+        got = gsr.isin(values)
+    with expect_warning_if(PANDAS_GE_220 and is_len_str):
+        expected = psr.isin(values)
     assert_eq(got, expected)