add tests

python/cudf/cudf/core/index.py

@@ -1443,6 +1443,11 @@ def __repr__(self):
                     output[:break_idx].replace("'", "") + output[break_idx:]
                 )
             else:
+                # Too many non-unique categories will cause
+                # the output to take too long. In this case, we
+                # split the categories into data and categories
+                # and generate the repr separately and
+                # merge them.
                 pd_cats = pd.Categorical(
                     preprocess.astype(preprocess.categories.dtype).to_pandas()
                 )

python/cudf/cudf/testing/_utils.py

@@ -1,6 +1,7 @@
 # Copyright (c) 2020-2024, NVIDIA CORPORATION.
 
 import itertools
+import signal
 import string
 from collections import abc
 from contextlib import contextmanager
@@ -368,3 +369,23 @@ def sv_to_udf_str_testing_lowering(context, builder, sig, args):
     return cast_string_view_to_udf_string(
         context, builder, sig.args[0], sig.return_type, args[0]
     )
+
+
+class cudf_timeout:
+    """
+    Context manager to raise a TimeoutError after a specified number of seconds.
+    """
+
+    def __init__(self, seconds, *, timeout_message=""):
+        self.seconds = int(seconds)
+        self.timeout_message = timeout_message
+
+    def _timeout_handler(self, signum, frame):
+        raise TimeoutError(self.timeout_message)
+
+    def __enter__(self):
+        signal.signal(signal.SIGALRM, self._timeout_handler)
+        signal.alarm(self.seconds)
+
+    def __exit__(self, type, value, traceback):
+        signal.alarm(0)

python/cudf/cudf/tests/test_repr.py

@@ -1480,3 +1480,14 @@ def test_interval_index_repr():
     gi = cudf.from_pandas(pi)
 
     assert repr(pi) == repr(gi)
+
+
+def test_large_unique_categories_repr():
+    # Unfortunately, this is a long running test (takes about 1 minute)
+    # and there is no way we can reduce the time
+    pi = pd.CategoricalIndex(range(100_000_000))
+    gi = cudf.CategoricalIndex(range(100_000_000))
+    expected_repr = repr(pi)
+    with utils.cudf_timeout(2, timeout_message="Failed to repr fast enough"):
+        actual_repr = repr(gi)
+    assert expected_repr == actual_repr