datachain 0.11.11__tar.gz → 0.13.0__tar.gz

{datachain-0.11.11 → datachain-0.13.0}/.github/workflows/tests.yml

@@ -80,14 +80,6 @@ jobs:
 
       - name: Set up FFmpeg
         uses: AnimMouse/setup-ffmpeg@v1
-        id: ffmpeg-install
-        continue-on-error: ${{ runner.os == 'macOS' }}
-
-      # https://github.com/AnimMouse/setup-ffmpeg/issues/5
-      - if: steps.ffmpeg-install.outcome == 'failure' && runner.os == 'macOS'
-        run: brew install ffmpeg
-        env:
-          HOMEBREW_NO_AUTO_UPDATE: "1"
 
       - name: Set up Python ${{ matrix.pyv }}
         uses: actions/setup-python@v5
@@ -117,7 +109,7 @@ jobs:
         shell: bash
 
       - name: Run E2E tests
-        run: nox -s tests-${{ matrix.pyv }} -- -m "e2e" --cov-append $DISABLE_REMOTES_ARG
+        run: nox -s e2e-${{ matrix.pyv }}
         shell: bash
 
       - name: Upload coverage report
@@ -141,11 +133,13 @@ jobs:
       matrix:
         os: [ubuntu-latest, windows-latest]
         pyv: ['3.9', '3.13']
-        group: ['get_started', 'computer_vision', 'llm_and_nlp', 'multimodal']
+        group: ['get_started', 'computer_vision', 'multimodal']
         exclude:
           - {os: ubuntu-latest, pyv: '3.9', group: 'multimodal'}
           - {os: ubuntu-latest, pyv: '3.13', group: 'multimodal'}
         include:
+          # HF runs against actual API - thus run it only once
+          - {os: ubuntu-latest, pyv: "3.13", group: llm_and_nlp}
           - {os: ubuntu-latest-4-cores, pyv: "3.9", group: multimodal}
           - {os: ubuntu-latest-4-cores, pyv: "3.13", group: multimodal}
 
@@ -169,9 +163,8 @@ jobs:
       - name: Install nox
         run: uv pip install nox --system
 
-      # HF runs against actual API - thus run it only once
       - name: Set hf token
-        if: matrix.os == 'ubuntu-latest' && matrix.pyv == '3.13'
+        if: matrix.group == 'llm_and_nlp'
         run: echo 'HF_TOKEN=${{ secrets.HF_TOKEN }}' >> "$GITHUB_ENV"
 
       - name: Run examples

{datachain-0.11.11 → datachain-0.13.0}/.pre-commit-config.yaml

@@ -24,7 +24,7 @@ repos:
       - id: trailing-whitespace
         exclude: '^LICENSES/'
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: 'v0.9.9'
+    rev: 'v0.11.0'
     hooks:
       - id: ruff
         args: [--fix, --exit-non-zero-on-fix]

{datachain-0.11.11 → datachain-0.13.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: datachain
-Version: 0.11.11
+Version: 0.13.0
 Summary: Wrangle unstructured AI data at scale
 Author-email: Dmitry Petrov <support@dvc.org>
 License: Apache-2.0
@@ -71,6 +71,7 @@ Requires-Dist: usearch; extra == "vector"
 Provides-Extra: hf
 Requires-Dist: numba>=0.60.0; extra == "hf"
 Requires-Dist: datasets[audio,vision]>=2.21.0; extra == "hf"
+Requires-Dist: fsspec>=2024.12.0; extra == "hf"
 Provides-Extra: video
 Requires-Dist: ffmpeg-python; extra == "video"
 Requires-Dist: imageio[ffmpeg,pyav]>=2.37.0; extra == "video"
@@ -90,6 +91,7 @@ Requires-Dist: hypothesis; extra == "tests"
 Requires-Dist: aiotools>=1.7.0; extra == "tests"
 Requires-Dist: requests-mock; extra == "tests"
 Requires-Dist: scipy; extra == "tests"
+Requires-Dist: ultralytics; extra == "tests"
 Provides-Extra: dev
 Requires-Dist: datachain[docs,tests]; extra == "dev"
 Requires-Dist: mypy==1.15.0; extra == "dev"
@@ -103,7 +105,7 @@ Requires-Dist: datachain[tests]; extra == "examples"
 Requires-Dist: defusedxml; extra == "examples"
 Requires-Dist: accelerate; extra == "examples"
 Requires-Dist: huggingface_hub[hf_transfer]; extra == "examples"
-Requires-Dist: ultralytics==8.3.82; extra == "examples"
+Requires-Dist: ultralytics; extra == "examples"
 Requires-Dist: open_clip_torch; extra == "examples"
 
 ================

{datachain-0.11.11 → datachain-0.13.0}/docs/examples.md

@@ -13,7 +13,7 @@ title: Examples
     For example, let us consider the New Yorker Cartoon caption contest dataset, where cartoons are matched against the potential titles. Let us imagine we want to augment this dataset with synthetic scene descriptions coming from an AI model. The below code takes images from the cloud, and applies PaliGemma model to caption the first five of them and put the results in the column “scene”:
 
     ```python
-    from datachain.lib.dc import Column, DataChain, File # (1)!
+    from datachain import Column, DataChain, File # (1)!
     from transformers import AutoProcessor, PaliGemmaForConditionalGeneration # (2)!
 
     images = DataChain.from_storage("gs://datachain-demo/newyorker_caption_contest/images", type="image")
@@ -80,12 +80,10 @@ In the below example, we are calling a Mixtral 8x22b model to judge the “servi
 # $ export MISTRAL_API_KEY='your key'
 
 import os
-from datachain.lib.feature import Feature
-from datachain.lib.dc import Column, DataChain
+from datachain import Column, DataChain, DataModel, Feature
 from mistralai.client import MistralClient
 from mistralai.models.chat_completion import ChatMessage
 from mistralai.models.chat_completion import ChatCompletionResponse as MistralModel
-from datachain.lib.data_model import DataModel
 
 prompt = "Was this dialog successful? Describe the 'result' as 'Yes' or 'No' in a short JSON"
 api_key = os.environ["MISTRAL_API_KEY"]
@@ -189,7 +187,7 @@ DataChain library understands common annotation formats (JSON, CSV, webdataset a
 Here is an example of reading a simple CSV file where schema is heuristically derived from the header:
 
 ```python
-from datachain.lib.dc import DataChain
+from datachain import DataChain
 
 uri="gs://datachain-demo/chatbot-csv/"
 csv_dataset = DataChain.from_csv(uri)
@@ -234,7 +232,7 @@ However, Datachain can easily parse the entire COCO structure via several readin
 
 ```python
 
-from datachain.lib.dc import Column, DataChain
+from datachain import Column, DataChain
 
 images_uri="gs://datachain-demo/coco2017/images/val/"
 captions_uri="gs://datachain-demo/coco2017/annotations/captions_val2017.json"

{datachain-0.11.11 → datachain-0.13.0}/docs/quick-start.md

@@ -138,7 +138,7 @@ chain = (
 )
 
 successful_chain = chain.filter(Column("is_success") == True)
-successful_chain.export_files("./output_mistral")
+successful_chain.to_storage("./output_mistral")
 
 print(f"{successful_chain.count()} files were exported")
 ```

{datachain-0.11.11 → datachain-0.13.0}/examples/computer_vision/openimage-detect.py

@@ -22,13 +22,9 @@ def openimage_detect(args):
         detections = json.load(stream_json).get("detections", [])
 
     for i, detect in enumerate(detections):
-        bbox = model.BBox.from_list(
-            [
-                detect["XMin"] * img.width,
-                detect["XMax"] * img.width,
-                detect["YMin"] * img.height,
-                detect["YMax"] * img.height,
-            ]
+        bbox = model.BBox.from_albumentations(
+            [detect[k] for k in ("XMin", "YMin", "XMax", "YMax")],
+            img_size=(img.width, img.height),
         )
 
         fstream = File(

{datachain-0.11.11 → datachain-0.13.0}/examples/computer_vision/ultralytics-bbox.py

@@ -1,11 +1,3 @@
-import os
-
-os.environ["YOLO_VERBOSE"] = "false"
-
-
-from io import BytesIO
-
-from PIL import Image
 from ultralytics import YOLO
 
 from datachain import C, DataChain, File
@@ -13,7 +5,7 @@ from datachain.model.ultralytics import YoloBBoxes
 
 
 def process_bboxes(yolo: YOLO, file: File) -> YoloBBoxes:
-    results = yolo(Image.open(BytesIO(file.read())))
+    results = yolo(file.as_image_file().read(), verbose=False)
     return YoloBBoxes.from_results(results)
 
 

{datachain-0.11.11 → datachain-0.13.0}/examples/computer_vision/ultralytics-pose.py

@@ -1,11 +1,3 @@
-import os
-
-os.environ["YOLO_VERBOSE"] = "false"
-
-
-from io import BytesIO
-
-from PIL import Image
 from ultralytics import YOLO
 
 from datachain import C, DataChain, File
@@ -13,7 +5,7 @@ from datachain.model.ultralytics import YoloPoses
 
 
 def process_poses(yolo: YOLO, file: File) -> YoloPoses:
-    results = yolo(Image.open(BytesIO(file.read())))
+    results = yolo(file.as_image_file().read(), verbose=False)
     return YoloPoses.from_results(results)
 
 

{datachain-0.11.11 → datachain-0.13.0}/examples/computer_vision/ultralytics-segment.py

@@ -1,11 +1,3 @@
-import os
-
-os.environ["YOLO_VERBOSE"] = "false"
-
-
-from io import BytesIO
-
-from PIL import Image
 from ultralytics import YOLO
 
 from datachain import C, DataChain, File
@@ -13,7 +5,7 @@ from datachain.model.ultralytics import YoloSegments
 
 
 def process_segments(yolo: YOLO, file: File) -> YoloSegments:
-    results = yolo(Image.open(BytesIO(file.read())))
+    results = yolo(file.as_image_file().read(), verbose=False)
     return YoloSegments.from_results(results)
 
 

{datachain-0.11.11 → datachain-0.13.0}/noxfile.py

@@ -56,6 +56,20 @@ def tests(session: nox.Session) -> None:
     )
 
 
+@nox.session(python=python_versions)
+def e2e(session: nox.Session) -> None:
+    session.install(".[tests]")
+    session.run(
+        "pytest",
+        "--durations=0",
+        "--numprocesses=logical",
+        "--dist=loadgroup",
+        "-m",
+        "e2e",
+        *session.posargs,
+    )
+
+
 @nox.session
 def lint(session: nox.Session) -> None:
     session.install("pre-commit")

{datachain-0.11.11 → datachain-0.13.0}/pyproject.toml

@@ -80,7 +80,8 @@ vector = [
 ]
 hf = [
   "numba>=0.60.0",
-  "datasets[audio,vision]>=2.21.0"
+  "datasets[audio,vision]>=2.21.0",
+  "fsspec>=2024.12.0"
 ]
 video = [
   "ffmpeg-python",
@@ -101,7 +102,8 @@ tests = [
   "hypothesis",
   "aiotools>=1.7.0",
   "requests-mock",
-  "scipy"
+  "scipy",
+  "ultralytics"
 ]
 dev = [
   "datachain[docs,tests]",
@@ -117,7 +119,7 @@ examples = [
   "defusedxml",
   "accelerate",
   "huggingface_hub[hf_transfer]",
-  "ultralytics==8.3.82",
+  "ultralytics",
   "open_clip_torch"
 ]
 

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/catalog/catalog.py

@@ -25,7 +25,6 @@ from typing import (
 )
 from uuid import uuid4
 
-import requests
 import sqlalchemy as sa
 from sqlalchemy import Column
 from tqdm.auto import tqdm
@@ -54,7 +53,6 @@ from datachain.error import (
 from datachain.lib.listing import get_listing
 from datachain.node import DirType, Node, NodeWithPath
 from datachain.nodes_thread_pool import NodesThreadPool
-from datachain.remote.studio import StudioClient
 from datachain.sql.types import DateTime, SQLType
 from datachain.utils import DataChainDir
 
@@ -162,6 +160,8 @@ class DatasetRowsFetcher(NodesThreadPool):
         max_threads: int = PULL_DATASET_MAX_THREADS,
         progress_bar=None,
     ):
+        from datachain.remote.studio import StudioClient
+
         super().__init__(max_threads)
         self._check_dependencies()
         self.metastore = metastore
@@ -234,6 +234,8 @@ class DatasetRowsFetcher(NodesThreadPool):
         return df.drop("sys__id", axis=1)
 
     def get_parquet_content(self, url: str):
+        import requests
+
         while True:
             if self.should_check_for_status():
                 self.check_for_status()
@@ -775,6 +777,8 @@ class Catalog:
         validate_version: Optional[bool] = True,
         listing: Optional[bool] = False,
         uuid: Optional[str] = None,
+        description: Optional[str] = None,
+        labels: Optional[list[str]] = None,
     ) -> "DatasetRecord":
         """
         Creates new dataset of a specific version.
@@ -801,6 +805,8 @@ class Catalog:
                 query_script=query_script,
                 schema=schema,
                 ignore_if_exists=True,
+                description=description,
+                labels=labels,
             )
 
         version = version or default_version
@@ -1130,6 +1136,8 @@ class Catalog:
         raise DatasetNotFoundError(f"Dataset with version uuid {uuid} not found.")
 
     def get_remote_dataset(self, name: str) -> DatasetRecord:
+        from datachain.remote.studio import StudioClient
+
         studio_client = StudioClient()
 
         info_response = studio_client.dataset_info(name)
@@ -1164,8 +1172,27 @@ class Catalog:
 
         return direct_dependencies
 
-    def ls_datasets(self, include_listing: bool = False) -> Iterator[DatasetListRecord]:
-        datasets = self.metastore.list_datasets()
+    def ls_datasets(
+        self, include_listing: bool = False, studio: bool = False
+    ) -> Iterator[DatasetListRecord]:
+        from datachain.remote.studio import StudioClient
+
+        if studio:
+            client = StudioClient()
+            response = client.ls_datasets()
+            if not response.ok:
+                raise DataChainError(response.message)
+            if not response.data:
+                return
+
+            datasets: Iterator[DatasetListRecord] = (
+                DatasetListRecord.from_dict(d)
+                for d in response.data
+                if not d.get("name", "").startswith(QUERY_DATASET_PREFIX)
+            )
+        else:
+            datasets = self.metastore.list_datasets()
+
         for d in datasets:
             if not d.is_bucket_listing or include_listing:
                 yield d
@@ -1173,9 +1200,12 @@ class Catalog:
     def list_datasets_versions(
         self,
         include_listing: bool = False,
+        studio: bool = False,
     ) -> Iterator[tuple[DatasetListRecord, "DatasetListVersion", Optional["Job"]]]:
         """Iterate over all dataset versions with related jobs."""
-        datasets = list(self.ls_datasets(include_listing=include_listing))
+        datasets = list(
+            self.ls_datasets(include_listing=include_listing, studio=studio)
+        )
 
         # preselect dataset versions jobs from db to avoid multiple queries
         jobs_ids: set[str] = {
@@ -1345,6 +1375,8 @@ class Catalog:
         if cp and not output:
             raise ValueError("Please provide output directory for instantiation")
 
+        from datachain.remote.studio import StudioClient
+
         studio_client = StudioClient()
 
         try:
@@ -1580,7 +1612,7 @@ class Catalog:
             except TerminationSignal as exc:
                 signal.signal(signal.SIGTERM, orig_sigterm_handler)
                 signal.signal(signal.SIGINT, orig_sigint_handler)
-                logging.info("Shutting down process %s, received %r", proc.pid, exc)
+                logger.info("Shutting down process %s, received %r", proc.pid, exc)
                 # Rather than forwarding the signal to the child, we try to shut it down
                 # gracefully. This is because we consider the script to be interactive
                 # and special, so we give it time to cleanup before exiting.
@@ -1595,7 +1627,7 @@ class Catalog:
                 if thread:
                     thread.join()  # wait for the reader thread
 
-        logging.info("Process %s exited with return code %s", proc.pid, proc.returncode)
+        logger.info("Process %s exited with return code %s", proc.pid, proc.returncode)
         if proc.returncode == QUERY_SCRIPT_CANCELED_EXIT_CODE:
             raise QueryScriptCancelError(
                 "Query script was canceled by user",

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/catalog/loader.py

@@ -1,19 +1,13 @@
 import os
 from importlib import import_module
-from typing import Any, Optional
-
-from datachain.catalog import Catalog
-from datachain.data_storage import (
-    AbstractMetastore,
-    AbstractWarehouse,
-)
-from datachain.data_storage.serializer import deserialize
-from datachain.data_storage.sqlite import (
-    SQLiteMetastore,
-    SQLiteWarehouse,
-)
+from typing import TYPE_CHECKING, Any, Optional
+
 from datachain.utils import get_envs_by_prefix
 
+if TYPE_CHECKING:
+    from datachain.catalog import Catalog
+    from datachain.data_storage import AbstractMetastore, AbstractWarehouse
+
 METASTORE_SERIALIZED = "DATACHAIN__METASTORE"
 METASTORE_IMPORT_PATH = "DATACHAIN_METASTORE"
 METASTORE_ARG_PREFIX = "DATACHAIN_METASTORE_ARG_"
@@ -27,6 +21,9 @@ IN_MEMORY_ERROR_MESSAGE = "In-memory is only supported on SQLite"
 
 
 def get_metastore(in_memory: bool = False) -> "AbstractMetastore":
+    from datachain.data_storage import AbstractMetastore
+    from datachain.data_storage.serializer import deserialize
+
     metastore_serialized = os.environ.get(METASTORE_SERIALIZED)
     if metastore_serialized:
         metastore_obj = deserialize(metastore_serialized)
@@ -45,6 +42,8 @@ def get_metastore(in_memory: bool = False) -> "AbstractMetastore":
     }
 
     if not metastore_import_path:
+        from datachain.data_storage.sqlite import SQLiteMetastore
+
         metastore_args["in_memory"] = in_memory
         return SQLiteMetastore(**metastore_args)
     if in_memory:
@@ -62,6 +61,9 @@ def get_metastore(in_memory: bool = False) -> "AbstractMetastore":
 
 
 def get_warehouse(in_memory: bool = False) -> "AbstractWarehouse":
+    from datachain.data_storage import AbstractWarehouse
+    from datachain.data_storage.serializer import deserialize
+
     warehouse_serialized = os.environ.get(WAREHOUSE_SERIALIZED)
     if warehouse_serialized:
         warehouse_obj = deserialize(warehouse_serialized)
@@ -80,6 +82,8 @@ def get_warehouse(in_memory: bool = False) -> "AbstractWarehouse":
     }
 
     if not warehouse_import_path:
+        from datachain.data_storage.sqlite import SQLiteWarehouse
+
         warehouse_args["in_memory"] = in_memory
         return SQLiteWarehouse(**warehouse_args)
     if in_memory:
@@ -121,7 +125,7 @@ def get_distributed_class(**kwargs):
 
 def get_catalog(
     client_config: Optional[dict[str, Any]] = None, in_memory: bool = False
-) -> Catalog:
+) -> "Catalog":
     """
     Function that creates Catalog instance with appropriate metastore
     and warehouse classes. Metastore class can be provided with env variable
@@ -133,6 +137,8 @@ def get_catalog(
     and name of variable after, e.g. if it accepts team_id as kwargs
     we can provide DATACHAIN_METASTORE_ARG_TEAM_ID=12345 env variable.
     """
+    from datachain.catalog import Catalog
+
     return Catalog(
         metastore=get_metastore(in_memory=in_memory),
         warehouse=get_warehouse(in_memory=in_memory),

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/cli/__init__.py

@@ -6,7 +6,6 @@ from multiprocessing import freeze_support
 from typing import Optional
 
 from datachain.cli.utils import get_logging_level
-from datachain.telemetry import telemetry
 
 from .commands import (
     clear_cache,
@@ -70,6 +69,8 @@ def main(argv: Optional[list[str]] = None) -> int:
         error, return_code = handle_general_exception(exc, args, logging_level)
         return return_code
     finally:
+        from datachain.telemetry import telemetry
+
         telemetry.send_cli_call(args.command, error=error)
 
 

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/cli/commands/ls.py

@@ -38,11 +38,12 @@ def ls_local(
 ):
     from datachain import DataChain
 
-    if catalog is None:
-        from datachain.catalog import get_catalog
-
-        catalog = get_catalog(client_config=client_config)
     if sources:
+        if catalog is None:
+            from datachain.catalog import get_catalog
+
+            catalog = get_catalog(client_config=client_config)
+
         actual_sources = list(ls_urls(sources, catalog=catalog, long=long, **kwargs))
         if len(actual_sources) == 1:
             for _, entries in actual_sources:
@@ -61,8 +62,9 @@ def ls_local(
                 for entry in entries:
                     print(format_ls_entry(entry))
     else:
-        chain = DataChain.listings()
-        for ls in chain.collect("listing"):
+        # Collect results in a list here to prevent interference from `tqdm` and `print`
+        listing = list(DataChain.listings().collect("listing"))
+        for ls in listing:
             print(format_ls_entry(f"{ls.uri}@v{ls.version}"))  # type: ignore[union-attr]
 
 

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/cli/commands/show.py

@@ -40,6 +40,13 @@ def show(
         .offset(offset)
     )
     records = query.to_db_records()
+    print("Name: ", name)
+    if dataset.description:
+        print("Description: ", dataset.description)
+    if dataset.labels:
+        print("Labels: ", ",".join(dataset.labels))
+    print("\n")
+
     show_records(records, collapse_columns=not no_collapse, hidden_fields=hidden_fields)
 
     if schema and dataset_version.feature_schema:

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/cli/parser/studio.py

@@ -63,19 +63,31 @@ def add_auth_parser(subparsers, parent_parser) -> None:
         default=False,
         help="Use code-based authentication without browser",
     )
+    login_parser.add_argument(
+        "--local",
+        action="store_true",
+        default=False,
+        help="Save the token in the local project config",
+    )
 
     auth_logout_help = "Log out from Studio"
     auth_logout_description = (
         "Remove the Studio authentication token from global config."
     )
 
-    auth_subparser.add_parser(
+    logout_parser = auth_subparser.add_parser(
         "logout",
         parents=[parent_parser],
         description=auth_logout_description,
         help=auth_logout_help,
         formatter_class=CustomHelpFormatter,
     )
+    logout_parser.add_argument(
+        "--local",
+        action="store_true",
+        default=False,
+        help="Remove the token from the local project config",
+    )
 
     auth_team_help = "Set default team for Studio operations"
     auth_team_description = "Set the default team for Studio operations."

{datachain-0.11.11 → datachain-0.13.0}/src/datachain/client/fsspec.py

@@ -17,10 +17,10 @@ from typing import (
     ClassVar,
     NamedTuple,
     Optional,
+    Union,
 )
 from urllib.parse import urlparse
 
-from botocore.exceptions import ClientError
 from dvc_objects.fs.system import reflink
 from fsspec.asyn import get_loop, sync
 from fsspec.callbacks import DEFAULT_CALLBACK, Callback
@@ -28,7 +28,6 @@ from tqdm.auto import tqdm
 
 from datachain.cache import Cache
 from datachain.client.fileslice import FileWrapper
-from datachain.error import ClientError as DataChainClientError
 from datachain.nodes_fetcher import NodesFetcher
 from datachain.nodes_thread_pool import NodeChunk
 
@@ -83,19 +82,17 @@ class Client(ABC):
         self.uri = self.get_uri(self.name)
 
     @staticmethod
-    def get_implementation(url: str) -> type["Client"]:
+    def get_implementation(url: Union[str, os.PathLike[str]]) -> type["Client"]:
         from .azure import AzureClient
         from .gcs import GCSClient
         from .hf import HfClient
         from .local import FileClient
         from .s3 import ClientS3
 
-        protocol = urlparse(url).scheme
+        protocol = urlparse(str(url)).scheme
 
-        if not protocol or _is_win_local_path(url):
+        if not protocol or _is_win_local_path(str(url)):
             return FileClient
-
-        protocol = protocol.lower()
         if protocol == ClientS3.protocol:
             return ClientS3
         if protocol == GCSClient.protocol:
@@ -121,9 +118,11 @@ class Client(ABC):
         return cls.get_uri(storage_name), rel_path
 
     @staticmethod
-    def get_client(source: str, cache: Cache, **kwargs) -> "Client":
+    def get_client(
+        source: Union[str, os.PathLike[str]], cache: Cache, **kwargs
+    ) -> "Client":
         cls = Client.get_implementation(source)
-        storage_url, _ = cls.split_url(source)
+        storage_url, _ = cls.split_url(str(source))
         if os.name == "nt":
             storage_url = storage_url.removeprefix("/")
 
@@ -209,7 +208,7 @@ class Client(ABC):
 
     async def get_current_etag(self, file: "File") -> str:
         kwargs = {}
-        if self.fs.version_aware:
+        if getattr(self.fs, "version_aware", False):
             kwargs["version_id"] = file.version
         info = await self.fs._info(
             self.get_full_path(file.path, file.version), **kwargs
@@ -286,11 +285,6 @@ class Client(ABC):
                     worker.cancel()
             if excs:
                 raise excs[0]
-        except ClientError as exc:
-            raise DataChainClientError(
-                exc.response.get("Error", {}).get("Message") or exc,
-                exc.response.get("Error", {}).get("Code"),
-            ) from exc
         finally:
             # This ensures the progress bar is closed before any exceptions are raised
             progress_bar.close()
@@ -333,7 +327,9 @@ class Client(ABC):
         return not (key.startswith("/") or key.endswith("/") or "//" in key)
 
     async def ls_dir(self, path):
-        return await self.fs._ls(path, detail=True, versions=True)
+        if getattr(self.fs, "version_aware", False):
+            kwargs = {"versions": True}
+        return await self.fs._ls(path, detail=True, **kwargs)
 
     def rel_path(self, path: str) -> str:
         return self.fs.split_path(path)[1]