Move info_utils errors to exceptions module (#6952)

* Move info_utils errors to exceptions module

* Create new errors and deprecate old ones

* Replace deprecated errors

* Make deprecation backward compatible

* Test deprecated and non-deprecated errors

* Fix non-deprecated errors

src/datasets/__init__.py

@@ -58,7 +58,9 @@
 # Deprecated modules
 from . import arrow_dataset as _arrow_dataset
 from . import utils as _utils
+from .exceptions import ExpectedMoreDownloadedFiles, ExpectedMoreSplits, UnexpectedDownloadedFile, UnexpectedSplits
 from .utils import download_manager as _deprecated_download_manager
+from .utils import info_utils as _deprecated_info_utils
 
 
 _arrow_dataset.concatenate_datasets = concatenate_datasets
@@ -68,5 +70,10 @@
 _deprecated_download_manager.DownloadConfig = DownloadConfig
 _deprecated_download_manager.DownloadMode = DownloadMode
 _deprecated_download_manager.DownloadManager = DownloadManager
+_deprecated_info_utils.ExpectedMoreDownloadedFiles = ExpectedMoreDownloadedFiles
+_deprecated_info_utils.ExpectedMoreSplits = ExpectedMoreSplits
+_deprecated_info_utils.UnexpectedDownloadedFile = UnexpectedDownloadedFile
+_deprecated_info_utils.UnexpectedSplits = UnexpectedSplits
 
 del _arrow_dataset, _utils, _deprecated_download_manager
+del _deprecated_info_utils, ExpectedMoreDownloadedFiles, ExpectedMoreSplits, UnexpectedDownloadedFile, UnexpectedSplits

src/datasets/exceptions.py

@@ -6,6 +6,7 @@
 
 from . import config
 from .table import CastError
+from .utils.deprecation_utils import deprecated
 from .utils.track import TrackedIterable, tracked_list, tracked_str
 
 
@@ -83,3 +84,113 @@ def from_cast_error(
             explanation_message += f"\n\nThis happened while the {builder_name} dataset builder was generating data using\n\n{', '.join(formatted_tracked_gen_kwargs)}"
         help_message = "\n\nPlease either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)"
         return cls("An error occurred while generating the dataset" + explanation_message + help_message)
+
+
+@deprecated("Use 'ChecksumVerificationError' instead.")
+class ChecksumVerificationException(Exception):
+    """Exceptions during checksums verifications of downloaded files.
+
+    <Deprecated version="2.20.0">
+
+    Use `ChecksumVerificationError` instead.
+
+    </Deprecated>
+    """
+
+
+class ChecksumVerificationError(DatasetsError, ChecksumVerificationException):
+    """Error raised during checksums verifications of downloaded files."""
+
+    def __init__(self, *args, **kwargs):
+        DatasetsError.__init__(self, *args, **kwargs)
+
+
+@deprecated("Use 'UnexpectedDownloadedFileError' instead.")
+class UnexpectedDownloadedFile(ChecksumVerificationException):
+    """Some downloaded files were not expected.
+
+    <Deprecated version="2.20.0">
+
+    Use `UnexpectedDownloadedFileError` instead.
+
+    </Deprecated>
+    """
+
+
+class UnexpectedDownloadedFileError(ChecksumVerificationError, UnexpectedDownloadedFile):
+    """Some downloaded files were not expected."""
+
+
+@deprecated("Use 'ExpectedMoreDownloadedFilesError' instead.")
+class ExpectedMoreDownloadedFiles(ChecksumVerificationException):
+    """Some files were supposed to be downloaded but were not.
+
+    <Deprecated version="2.20.0">
+
+    Use `ExpectedMoreDownloadedFilesError` instead.
+
+    </Deprecated>
+    """
+
+
+class ExpectedMoreDownloadedFilesError(ChecksumVerificationError, ExpectedMoreDownloadedFiles):
+    """Some files were supposed to be downloaded but were not."""
+
+
+class NonMatchingChecksumError(ChecksumVerificationError):
+    """The downloaded file checksum don't match the expected checksum."""
+
+
+@deprecated("Use 'SplitsVerificationError' instead.")
+class SplitsVerificationException(Exception):
+    """Exceptions during splits verifications.
+
+    <Deprecated version="2.20.0">
+
+    Use `SplitsVerificationError` instead.
+
+    </Deprecated>
+    """
+
+
+class SplitsVerificationError(DatasetsError, SplitsVerificationException):
+    """Error raised during splits verifications."""
+
+    def __init__(self, *args, **kwargs):
+        DatasetsError.__init__(self, *args, **kwargs)
+
+
+@deprecated("Use 'UnexpectedSplitsError' instead.")
+class UnexpectedSplits(SplitsVerificationException):
+    """The expected splits of the downloaded file is missing.
+
+    <Deprecated version="2.20.0">
+
+    Use `UnexpectedSplitsError` instead.
+
+    </Deprecated>
+    """
+
+
+class UnexpectedSplitsError(SplitsVerificationError, UnexpectedSplits):
+    """The expected splits of the downloaded file is missing."""
+
+
+@deprecated("Use 'ExpectedMoreSplitsError' instead.")
+class ExpectedMoreSplits(SplitsVerificationException):
+    """Some recorded splits are missing.
+
+    <Deprecated version="2.20.0">
+
+    Use `ExpectedMoreSplitsError` instead.
+
+    </Deprecated>
+    """
+
+
+class ExpectedMoreSplitsError(SplitsVerificationError, ExpectedMoreSplits):
+    """Some recorded splits are missing."""
+
+
+class NonMatchingSplitsSizesError(SplitsVerificationError):
+    """The splits sizes don't match the expected splits sizes."""

src/datasets/utils/info_utils.py

@@ -5,6 +5,14 @@
 from huggingface_hub.utils import insecure_hashlib
 
 from .. import config
+from ..exceptions import (
+    ExpectedMoreDownloadedFilesError,
+    ExpectedMoreSplitsError,
+    NonMatchingChecksumError,
+    NonMatchingSplitsSizesError,
+    UnexpectedDownloadedFileError,
+    UnexpectedSplitsError,
+)
 from .logging import get_logger
 
 
@@ -33,30 +41,14 @@ class VerificationMode(enum.Enum):
     NO_CHECKS = "no_checks"
 
 
-class ChecksumVerificationException(Exception):
-    """Exceptions during checksums verifications of downloaded files."""
-
-
-class UnexpectedDownloadedFile(ChecksumVerificationException):
-    """Some downloaded files were not expected."""
-
-
-class ExpectedMoreDownloadedFiles(ChecksumVerificationException):
-    """Some files were supposed to be downloaded but were not."""
-
-
-class NonMatchingChecksumError(ChecksumVerificationException):
-    """The downloaded file checksum don't match the expected checksum."""
-
-
 def verify_checksums(expected_checksums: Optional[dict], recorded_checksums: dict, verification_name=None):
     if expected_checksums is None:
         logger.info("Unable to verify checksums.")
         return
     if len(set(expected_checksums) - set(recorded_checksums)) > 0:
-        raise ExpectedMoreDownloadedFiles(str(set(expected_checksums) - set(recorded_checksums)))
+        raise ExpectedMoreDownloadedFilesError(str(set(expected_checksums) - set(recorded_checksums)))
     if len(set(recorded_checksums) - set(expected_checksums)) > 0:
-        raise UnexpectedDownloadedFile(str(set(recorded_checksums) - set(expected_checksums)))
+        raise UnexpectedDownloadedFileError(str(set(recorded_checksums) - set(expected_checksums)))
     bad_urls = [url for url in expected_checksums if expected_checksums[url] != recorded_checksums[url]]
     for_verification_name = " for " + verification_name if verification_name is not None else ""
     if len(bad_urls) > 0:
@@ -68,30 +60,14 @@ def verify_checksums(expected_checksums: Optional[dict], recorded_checksums: dic
     logger.info("All the checksums matched successfully" + for_verification_name)
 
 
-class SplitsVerificationException(Exception):
-    """Exceptions during splis verifications"""
-
-
-class UnexpectedSplits(SplitsVerificationException):
-    """The expected splits of the downloaded file is missing."""
-
-
-class ExpectedMoreSplits(SplitsVerificationException):
-    """Some recorded splits are missing."""
-
-
-class NonMatchingSplitsSizesError(SplitsVerificationException):
-    """The splits sizes don't match the expected splits sizes."""
-
-
 def verify_splits(expected_splits: Optional[dict], recorded_splits: dict):
     if expected_splits is None:
         logger.info("Unable to verify splits sizes.")
         return
     if len(set(expected_splits) - set(recorded_splits)) > 0:
-        raise ExpectedMoreSplits(str(set(expected_splits) - set(recorded_splits)))
+        raise ExpectedMoreSplitsError(str(set(expected_splits) - set(recorded_splits)))
     if len(set(recorded_splits) - set(expected_splits)) > 0:
-        raise UnexpectedSplits(str(set(recorded_splits) - set(expected_splits)))
+        raise UnexpectedSplitsError(str(set(recorded_splits) - set(expected_splits)))
     bad_splits = [
         {"expected": expected_splits[name], "recorded": recorded_splits[name]}
         for name in expected_splits

tests/test_exceptions.py

@@ -0,0 +1,58 @@
+import warnings
+
+import pytest
+
+import datasets.utils.deprecation_utils
+from datasets.exceptions import (
+    ChecksumVerificationError,
+    ChecksumVerificationException,
+    ExpectedMoreDownloadedFiles,
+    ExpectedMoreDownloadedFilesError,
+    ExpectedMoreSplits,
+    ExpectedMoreSplitsError,
+    NonMatchingChecksumError,
+    NonMatchingSplitsSizesError,
+    SplitsVerificationError,
+    SplitsVerificationException,
+    UnexpectedDownloadedFile,
+    UnexpectedDownloadedFileError,
+    UnexpectedSplits,
+    UnexpectedSplitsError,
+)
+
+
+@pytest.mark.parametrize(
+    "error",
+    [
+        ChecksumVerificationException,
+        UnexpectedDownloadedFile,
+        ExpectedMoreDownloadedFiles,
+        SplitsVerificationException,
+        UnexpectedSplits,
+        ExpectedMoreSplits,
+    ],
+)
+def test_error_deprecated(error, monkeypatch):
+    monkeypatch.setattr(datasets.utils.deprecation_utils, "_emitted_deprecation_warnings", set())
+    with pytest.deprecated_call():
+        error()
+
+
+@pytest.mark.parametrize(
+    "error",
+    [
+        ChecksumVerificationError,
+        UnexpectedDownloadedFileError,
+        ExpectedMoreDownloadedFilesError,
+        NonMatchingChecksumError,
+        SplitsVerificationError,
+        UnexpectedSplitsError,
+        ExpectedMoreSplitsError,
+        NonMatchingSplitsSizesError,
+    ],
+)
+def test_error_not_deprecated(error, monkeypatch):
+    monkeypatch.setattr(datasets.utils.deprecation_utils, "_emitted_deprecation_warnings", set())
+    with warnings.catch_warnings():
+        warnings.simplefilter("error")
+        error()