rename datasets-server to dataset-viewer (#6785)

* rename datasets-server to dataset-viewer

* use renamed error name

Author: severo

src/datasets/config.py

@@ -193,7 +193,7 @@
 )
 TIME_OUT_REMOTE_CODE = 15
 
-# Datasets-server
+# Dataset viewer API
 USE_PARQUET_EXPORT = True
 
 # Batch size constants. For more info, see:

src/datasets/load.py

@@ -71,7 +71,7 @@
     _hash_python_lines,
 )
 from .splits import Split
-from .utils import _datasets_server
+from .utils import _dataset_viewer
 from .utils.deprecation_utils import deprecated
 from .utils.file_utils import (
     OfflineModeIsEnabled,
@@ -1236,7 +1236,7 @@ def get_module(self) -> DatasetModule:
         metadata_configs = MetadataConfigs.from_dataset_card_data(dataset_card_data)
         dataset_infos = DatasetInfosDict.from_dataset_card_data(dataset_card_data)
         try:
-            exported_dataset_infos = _datasets_server.get_exported_dataset_infos(
+            exported_dataset_infos = _dataset_viewer.get_exported_dataset_infos(
                 dataset=self.name, revision=self.revision, token=self.download_config.token
             )
             exported_dataset_infos = DatasetInfosDict(
@@ -1245,7 +1245,7 @@ def get_module(self) -> DatasetModule:
                     for config_name in exported_dataset_infos
                 }
             )
-        except _datasets_server.DatasetsServerError:
+        except _dataset_viewer.DatasetViewerError:
             exported_dataset_infos = None
         if exported_dataset_infos:
             exported_dataset_infos.update(dataset_infos)
@@ -1372,10 +1372,10 @@ def __init__(
         increase_load_count(name, resource_type="dataset")
 
     def get_module(self) -> DatasetModule:
-        exported_parquet_files = _datasets_server.get_exported_parquet_files(
+        exported_parquet_files = _dataset_viewer.get_exported_parquet_files(
             dataset=self.name, revision=self.revision, token=self.download_config.token
         )
-        exported_dataset_infos = _datasets_server.get_exported_dataset_infos(
+        exported_dataset_infos = _dataset_viewer.get_exported_dataset_infos(
             dataset=self.name, revision=self.revision, token=self.download_config.token
         )
         dataset_infos = DatasetInfosDict(
@@ -1864,7 +1864,7 @@ def dataset_module_factory(
                         return HubDatasetModuleFactoryWithParquetExport(
                             path, download_config=download_config, revision=dataset_info.sha
                         ).get_module()
-                    except _datasets_server.DatasetsServerError:
+                    except _dataset_viewer.DatasetViewerError:
                         pass
                 # Otherwise we must use the dataset script if the user trusts it
                 return HubDatasetModuleFactoryWithScript(

src/datasets/utils/_datasets_server.py renamed to src/datasets/utils/_dataset_viewer.py

@@ -12,10 +12,10 @@
 logger = get_logger(__name__)
 
 
-class DatasetsServerError(DatasetsError):
-    """Dataset-server error.
+class DatasetViewerError(DatasetsError):
+    """Dataset viewer error.
 
-    Raised when trying to use the Datasets-server HTTP API and when trying to access:
+    Raised when trying to use the dataset viewer HTTP API and when trying to access:
     - a missing dataset, or
     - a private/gated dataset and the user is not authenticated.
     - unavailable /parquet or /info responses
@@ -27,10 +27,10 @@ def get_exported_parquet_files(dataset: str, revision: str, token: Optional[Unio
     Get the dataset exported parquet files
     Docs: https://huggingface.co/docs/datasets-server/parquet
     """
-    datasets_server_parquet_url = config.HF_ENDPOINT.replace("://", "://datasets-server.") + "/parquet?dataset="
+    dataset_viewer_parquet_url = config.HF_ENDPOINT.replace("://", "://datasets-server.") + "/parquet?dataset="
     try:
         parquet_data_files_response = http_get(
-            url=datasets_server_parquet_url + dataset,
+            url=dataset_viewer_parquet_url + dataset,
             temp_file=None,
             headers=get_authentication_headers_for_url(config.HF_ENDPOINT + f"datasets/{dataset}", token=token),
             timeout=100.0,
@@ -53,9 +53,9 @@ def get_exported_parquet_files(dataset: str, revision: str, token: Optional[Unio
                 logger.debug(
                     f"Parquet export for {dataset} is available but outdated (revision='{parquet_data_files_response.headers['X-Revision']}')"
                 )
-    except Exception as e:  # noqa catch any exception of the datasets-server and consider the parquet export doesn't exist
+    except Exception as e:  # noqa catch any exception of the dataset viewer API and consider the parquet export doesn't exist
         logger.debug(f"No parquet export for {dataset} available ({type(e).__name__}: {e})")
-    raise DatasetsServerError("No exported Parquet files available.")
+    raise DatasetViewerError("No exported Parquet files available.")
 
 
 def get_exported_dataset_infos(
@@ -65,10 +65,10 @@ def get_exported_dataset_infos(
     Get the dataset information, can be useful to get e.g. the dataset features.
     Docs: https://huggingface.co/docs/datasets-server/info
     """
-    datasets_server_info_url = config.HF_ENDPOINT.replace("://", "://datasets-server.") + "/info?dataset="
+    dataset_viewer_info_url = config.HF_ENDPOINT.replace("://", "://datasets-server.") + "/info?dataset="
     try:
         info_response = http_get(
-            url=datasets_server_info_url + dataset,
+            url=dataset_viewer_info_url + dataset,
             temp_file=None,
             headers=get_authentication_headers_for_url(config.HF_ENDPOINT + f"datasets/{dataset}", token=token),
             timeout=100.0,
@@ -91,6 +91,6 @@ def get_exported_dataset_infos(
                 logger.debug(
                     f"Dataset info for {dataset} is available but outdated (revision='{info_response.headers['X-Revision']}')"
                 )
-    except Exception as e:  # noqa catch any exception of the datasets-server and consider the dataset info doesn't exist
+    except Exception as e:  # noqa catch any exception of the dataset viewer API and consider the dataset info doesn't exist
         logger.debug(f"No dataset info for {dataset} available ({type(e).__name__}: {e})")
-    raise DatasetsServerError("No exported dataset infos available.")
+    raise DatasetViewerError("No exported dataset infos available.")

tests/test_load.py

@@ -46,7 +46,7 @@
 from datasets.packaged_modules.audiofolder.audiofolder import AudioFolder, AudioFolderConfig
 from datasets.packaged_modules.imagefolder.imagefolder import ImageFolder, ImageFolderConfig
 from datasets.packaged_modules.parquet.parquet import ParquetConfig
-from datasets.utils import _datasets_server
+from datasets.utils import _dataset_viewer
 from datasets.utils.logging import INFO, get_logger
 
 from .utils import (
@@ -862,7 +862,7 @@ def test_HubDatasetModuleFactoryWithParquetExport_errors_on_wrong_sha(self):
             download_config=self.download_config,
             revision="wrong_sha",
         )
-        with self.assertRaises(_datasets_server.DatasetsServerError):
+        with self.assertRaises(_dataset_viewer.DatasetViewerError):
             factory.get_module()
 
     @pytest.mark.integration