From 680d31a72dbe8188cc1a81a2eb8eb4140a9caa37 Mon Sep 17 00:00:00 2001
From: Albert Villanova del Moral
 <8515462+albertvillanova@users.noreply.github.com>
Date: Thu, 31 Aug 2023 10:38:22 +0200
Subject: [PATCH 1/3] Test split order in DataFilesDict

---
 tests/test_data_files.py | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/tests/test_data_files.py b/tests/test_data_files.py
index 34bfb26332e..d617a9c7c19 100644
--- a/tests/test_data_files.py
+++ b/tests/test_data_files.py
@@ -1,3 +1,4 @@
+import copy
 import os
 from pathlib import Path, PurePath
 from typing import List
@@ -385,6 +386,13 @@ def test_DataFilesList_from_patterns_raises_FileNotFoundError(complex_data_dir):
         DataFilesList.from_patterns(["file_that_doesnt_exist.txt"], complex_data_dir)
 
 
+class TestDataFilesDict:
+    def test_key_order_after_copy(self):
+        data_files = DataFilesDict({"train": "train.csv", "test": "test.csv"})
+        copied_data_files = copy.deepcopy(data_files)
+        assert list(copied_data_files.keys()) == list(data_files.keys())  # test split order with list()
+
+
 @pytest.mark.parametrize("pattern", _TEST_PATTERNS)
 def test_DataFilesDict_from_patterns_in_dataset_repository(
     hub_dataset_repo_path, hub_dataset_repo_patterns_results, pattern

From 416e9b777ba457acd8371154d8bb9c714a347f96 Mon Sep 17 00:00:00 2001
From: Albert Villanova del Moral
 <8515462+albertvillanova@users.noreply.github.com>
Date: Thu, 31 Aug 2023 10:39:10 +0200
Subject: [PATCH 2/3] Remove key sorting in DataFilesDict

---
 src/datasets/data_files.py | 11 -----------
 1 file changed, 11 deletions(-)

diff --git a/src/datasets/data_files.py b/src/datasets/data_files.py
index fd1eb739c99..77c3ae18284 100644
--- a/src/datasets/data_files.py
+++ b/src/datasets/data_files.py
@@ -682,17 +682,6 @@ def from_patterns(
             )
         return out
 
-    def __reduce__(self):
-        """
-        To make sure the order of the keys doesn't matter when pickling and hashing:
-
-        >>> from datasets.data_files import DataFilesDict
-        >>> from datasets.fingerprint import Hasher
-        >>> assert Hasher.hash(DataFilesDict(a=[], b=[])) == Hasher.hash(DataFilesDict(b=[], a=[]))
-
-        """
-        return DataFilesDict, (dict(sorted(self.items())),)
-
     def filter_extensions(self, extensions: List[str]) -> "DataFilesDict":
         out = type(self)()
         for key, data_files_list in self.items():

From 0f3b6eaf69d3352394d3bf3c4d6ed01dd2af5860 Mon Sep 17 00:00:00 2001
From: Albert Villanova del Moral
 <8515462+albertvillanova@users.noreply.github.com>
Date: Thu, 31 Aug 2023 11:21:31 +0200
Subject: [PATCH 3/3] Fix test_cache_dir_for_data_files

---
 tests/test_builder.py | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/tests/test_builder.py b/tests/test_builder.py
index 45b7574ffc7..54bae47ae08 100644
--- a/tests/test_builder.py
+++ b/tests/test_builder.py
@@ -759,10 +759,6 @@ def test_cache_dir_for_data_files(self):
                 cache_dir=tmp_dir, data_files={"train": [dummy_data1], "test": dummy_data2}
             )
             self.assertEqual(builder.cache_dir, other_builder.cache_dir)
-            other_builder = DummyGeneratorBasedBuilder(
-                cache_dir=tmp_dir, data_files={"test": dummy_data2, "train": dummy_data1}
-            )
-            self.assertEqual(builder.cache_dir, other_builder.cache_dir)
             other_builder = DummyGeneratorBasedBuilder(
                 cache_dir=tmp_dir, data_files={"train": dummy_data1, "validation": dummy_data2}
             )