Update Dataset Path in DuplicatedKeysError

The DuplicatedKeysError class was updated to always except
duplicate_key_indices and old Error message is removed.

Path to Dataset in Error updated using dataset name to make it more
user friendly.

Author: VijayKalmath

src/datasets/builder.py

@@ -49,6 +49,7 @@
 from .fingerprint import Hasher
 from .info import DatasetInfo, DatasetInfosDict, PostProcessedInfo
 from .iterable_dataset import ExamplesIterable, IterableDataset, _generate_examples_from_tables_wrapper
+from .keyhash import DuplicatedKeysError
 from .naming import camelcase_to_snakecase
 from .splits import Split, SplitDict, SplitGenerator
 from .streaming import extend_dataset_builder_for_streaming
@@ -798,7 +799,10 @@ def _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs
                     + "\nOriginal error:\n"
                     + str(e)
                 ) from None
-
+            # If check_duplicates is set to True , then except DuplicatedKeysError
+            except DuplicatedKeysError as e:
+                e.args = (e.args[0].replace("<Path to Dataset>", f"datasets/{self.name}/{self.name}.py"),)
+                raise (e) from None
             dl_manager.manage_extracted_files()
 
         if verify_infos:

src/datasets/keyhash.py

@@ -70,17 +70,11 @@ def __init__(self, hash_data):
 class DuplicatedKeysError(Exception):
     """Raise an error when duplicate key found."""
 
-    def __init__(self, key, duplicate_key_indices=None):
-        if duplicate_key_indices:
-            self.prefix = f"Found multiple examples with duplicate key: {key}"
-            self.err_msg = f"\nThe following examples {' ,'.join(duplicate_key_indices)} have the same key {key} "
-            self.suffix = "\nPlease fix the dataset script at <Path to Dataset>"
-            super().__init__(f"{self.prefix}{self.err_msg}{self.suffix}")
-        else:
-            self.prefix = "FAILURE TO GENERATE DATASET !"
-            self.err_msg = f"\nFound duplicate Key: {key}"
-            self.suffix = "\nKeys should be unique and deterministic in nature"
-            super().__init__(f"{self.prefix}{self.err_msg}{self.suffix}")
+    def __init__(self, key, duplicate_key_indices):
+        self.prefix = f"Found multiple examples generated with the same key"
+        self.err_msg = f"\nThe following examples {', '.join(duplicate_key_indices)} have the key {key}"
+        self.suffix = "\nPlease fix the dataset script at <Path to Dataset> to avoid duplicate keys"
+        super().__init__(f"{self.prefix}{self.err_msg}{self.suffix}")
 
 
 class KeyHasher: