crawlee 1.0.1b12__py3-none-any.whl → 1.0.2__py3-none-any.whl

crawlee/storage_clients/_file_system/_dataset_client.py

@@ -9,7 +9,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
 from pydantic import ValidationError
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._consts import METADATA_FILENAME
 from crawlee._utils.crypto import crypto_random_object_id
@@ -94,7 +94,7 @@ class FileSystemDatasetClient(DatasetClient):
         name: str | None,
         alias: str | None,
         configuration: Configuration,
-    ) -> FileSystemDatasetClient:
+    ) -> Self:
         """Open or create a file system dataset client.
 
         This method attempts to open an existing dataset from the file system. If a dataset with the specified ID

crawlee/storage_clients/_file_system/_key_value_store_client.py

@@ -10,7 +10,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
 from pydantic import ValidationError
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._consts import METADATA_FILENAME
 from crawlee._utils.crypto import crypto_random_object_id
@@ -93,7 +93,7 @@ class FileSystemKeyValueStoreClient(KeyValueStoreClient):
         name: str | None,
         alias: str | None,
         configuration: Configuration,
-    ) -> FileSystemKeyValueStoreClient:
+    ) -> Self:
         """Open or create a file system key-value store client.
 
         This method attempts to open an existing key-value store from the file system. If a KVS with the specified

crawlee/storage_clients/_file_system/_request_queue_client.py

@@ -11,7 +11,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING
 
 from pydantic import BaseModel, ValidationError
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee import Request
 from crawlee._consts import METADATA_FILENAME
@@ -144,7 +144,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         name: str | None,
         alias: str | None,
         configuration: Configuration,
-    ) -> FileSystemRequestQueueClient:
+    ) -> Self:
         """Open or create a file system request queue client.
 
         This method attempts to open an existing request queue from the file system. If a queue with the specified

crawlee/storage_clients/_memory/_dataset_client.py

@@ -4,7 +4,7 @@ from datetime import datetime, timezone
 from logging import getLogger
 from typing import TYPE_CHECKING, Any
 
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._utils.crypto import crypto_random_object_id
 from crawlee._utils.raise_if_too_many_kwargs import raise_if_too_many_kwargs
@@ -55,7 +55,7 @@ class MemoryDatasetClient(DatasetClient):
         id: str | None,
         name: str | None,
         alias: str | None,
-    ) -> MemoryDatasetClient:
+    ) -> Self:
         """Open or create a new memory dataset client.
 
         This method creates a new in-memory dataset instance. Unlike persistent storage implementations, memory

crawlee/storage_clients/_memory/_key_value_store_client.py

@@ -4,7 +4,7 @@ import sys
 from datetime import datetime, timezone
 from typing import TYPE_CHECKING, Any
 
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._utils.crypto import crypto_random_object_id
 from crawlee._utils.file import infer_mime_type
@@ -53,7 +53,7 @@ class MemoryKeyValueStoreClient(KeyValueStoreClient):
         id: str | None,
         name: str | None,
         alias: str | None,
-    ) -> MemoryKeyValueStoreClient:
+    ) -> Self:
         """Open or create a new memory key-value store client.
 
         This method creates a new in-memory key-value store instance. Unlike persistent storage implementations,

crawlee/storage_clients/_memory/_request_queue_client.py

@@ -6,7 +6,7 @@ from datetime import datetime, timezone
 from logging import getLogger
 from typing import TYPE_CHECKING
 
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee import Request
 from crawlee._utils.crypto import crypto_random_object_id
@@ -65,7 +65,7 @@ class MemoryRequestQueueClient(RequestQueueClient):
         id: str | None,
         name: str | None,
         alias: str | None,
-    ) -> MemoryRequestQueueClient:
+    ) -> Self:
         """Open or create a new memory request queue client.
 
         This method creates a new in-memory request queue instance. Unlike persistent storage implementations,

crawlee/storage_clients/_sql/_dataset_client.py

@@ -4,7 +4,7 @@ from logging import getLogger
 from typing import TYPE_CHECKING, Any
 
 from sqlalchemy import Select, insert, select
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee.storage_clients._base import DatasetClient
 from crawlee.storage_clients.models import DatasetItemsListPage, DatasetMetadata
@@ -78,7 +78,7 @@ class SqlDatasetClient(DatasetClient, SqlClientMixin):
         name: str | None,
         alias: str | None,
         storage_client: SqlStorageClient,
-    ) -> SqlDatasetClient:
+    ) -> Self:
         """Open an existing dataset or create a new one.
 
         Args:

crawlee/storage_clients/_sql/_key_value_store_client.py

@@ -5,7 +5,7 @@ from logging import getLogger
 from typing import TYPE_CHECKING, Any
 
 from sqlalchemy import delete, select
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._utils.file import infer_mime_type
 from crawlee.storage_clients._base import KeyValueStoreClient
@@ -77,7 +77,7 @@ class SqlKeyValueStoreClient(KeyValueStoreClient, SqlClientMixin):
         name: str | None,
         alias: str | None,
         storage_client: SqlStorageClient,
-    ) -> SqlKeyValueStoreClient:
+    ) -> Self:
         """Open or create a SQL key-value store client.
 
         This method attempts to open an existing key-value store from the SQL database. If a KVS with the specified

crawlee/storage_clients/_sql/_request_queue_client.py

@@ -10,7 +10,7 @@ from typing import TYPE_CHECKING, Any
 from sqlalchemy import func, or_, select, update
 from sqlalchemy.exc import SQLAlchemyError
 from sqlalchemy.orm import load_only
-from typing_extensions import NotRequired, override
+from typing_extensions import NotRequired, Self, override
 
 from crawlee import Request
 from crawlee._utils.crypto import crypto_random_object_id
@@ -119,7 +119,7 @@ class SqlRequestQueueClient(RequestQueueClient, SqlClientMixin):
         name: str | None,
         alias: str | None,
         storage_client: SqlStorageClient,
-    ) -> SqlRequestQueueClient:
+    ) -> Self:
         """Open an existing request queue or create a new one.
 
         This method first tries to find an existing queue by ID or name.

crawlee/storages/_base.py

@@ -44,7 +44,9 @@ class Storage(ABC):
 
         Args:
             id: The storage ID.
-            name: The storage name (global scope, persists across runs).
+            name: The storage name (global scope, persists across runs). Name can only contain letters "a" through "z",
+                the digits "0" through "9", and the hyphen ("-") but only in the middle of the string
+                (e.g. "my-value-1").
             alias: The storage alias (run scope, creates unnamed storage).
             configuration: Configuration object used during the storage creation or restoration process.
             storage_client: Underlying storage client to use. If not provided, the default global storage client

crawlee/storages/_dataset.py

@@ -12,6 +12,7 @@ from crawlee._utils.file import export_csv_to_stream, export_json_to_stream
 
 from ._base import Storage
 from ._key_value_store import KeyValueStore
+from ._utils import validate_storage_name
 
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator
@@ -75,6 +76,8 @@ class Dataset(Storage):
             id: The unique identifier of the storage.
             name: The name of the storage, if available.
         """
+        validate_storage_name(name)
+
         self._client = client
         self._id = id
         self._name = name

crawlee/storages/_key_value_store.py

@@ -15,6 +15,7 @@ from crawlee._utils.recoverable_state import RecoverableState
 from crawlee.storage_clients.models import KeyValueStoreMetadata
 
 from ._base import Storage
+from ._utils import validate_storage_name
 
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator
@@ -84,6 +85,8 @@ class KeyValueStore(Storage):
             id: The unique identifier of the storage.
             name: The name of the storage, if available.
         """
+        validate_storage_name(name)
+
         self._client = client
         self._id = id
         self._name = name

crawlee/storages/_request_queue.py

@@ -13,6 +13,7 @@ from crawlee._utils.wait import wait_for_all_tasks_for_finish
 from crawlee.request_loaders import RequestManager
 
 from ._base import Storage
+from ._utils import validate_storage_name
 
 if TYPE_CHECKING:
     from collections.abc import Sequence
@@ -80,6 +81,8 @@ class RequestQueue(Storage, RequestManager):
             id: The unique identifier of the storage.
             name: The name of the storage, if available.
         """
+        validate_storage_name(name)
+
         self._client = client
         self._id = id
         self._name = name

crawlee/storages/_storage_instance_manager.py

@@ -8,6 +8,8 @@ from typing import TYPE_CHECKING, TypeVar
 from crawlee._utils.raise_if_too_many_kwargs import raise_if_too_many_kwargs
 from crawlee.storage_clients._base import DatasetClient, KeyValueStoreClient, RequestQueueClient
 
+from ._utils import validate_storage_name
+
 if TYPE_CHECKING:
     from ._base import Storage
 
@@ -90,7 +92,9 @@ class StorageInstanceManager:
         Args:
             cls: The storage class to instantiate.
             id: Storage ID.
-            name: Storage name. (global scope, persists across runs).
+            name: Storage name. (global scope, persists across runs). Name can only contain letters "a" through "z",
+                the digits "0" through "9", and the hyphen ("-") but only in the middle of the string
+                (e.g. "my-value-1").
             alias: Storage alias (run scope, creates unnamed storage).
             client_opener_coro: Coroutine to open the storage client when storage instance not found in cache.
             storage_client_cache_key: Additional optional key from storage client to differentiate cache entries.
@@ -146,6 +150,10 @@ class StorageInstanceManager:
                     f'Use a different name or drop the existing alias storage first.'
                 )
 
+            # Validate storage name
+            if name is not None:
+                validate_storage_name(name)
+
             # Create new instance
             client: KeyValueStoreClient | DatasetClient | RequestQueueClient
             client = await client_opener_coro

crawlee/storages/_utils.py

@@ -0,0 +1,11 @@
+import re
+
+NAME_REGEX = re.compile(r'^([a-zA-Z0-9]|[a-zA-Z0-9][a-zA-Z0-9-]*[a-zA-Z0-9])$')
+
+
+def validate_storage_name(name: str | None) -> None:
+    if name and not NAME_REGEX.match(name):
+        raise ValueError(
+            f'Invalid storage name "{name}". Name can only contain letters "a" through "z", the digits "0" through'
+            '"9", and the hyphen ("-") but only in the middle of the string (e.g. "my-value-1")'
+        )

{crawlee-1.0.1b12.dist-info → crawlee-1.0.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: crawlee
-Version: 1.0.1b12
+Version: 1.0.2
 Summary: Crawlee for Python
 Project-URL: Apify Homepage, https://apify.com
 Project-URL: Changelog, https://crawlee.dev/python/docs/changelog
@@ -232,7 +232,7 @@ Requires-Dist: more-itertools>=10.2.0
 Requires-Dist: protego>=0.5.0
 Requires-Dist: psutil>=6.0.0
 Requires-Dist: pydantic-settings!=2.7.0,!=2.7.1,!=2.8.0,>=2.2.0
-Requires-Dist: pydantic>=2.11.0
+Requires-Dist: pydantic<2.12.0,>=2.11.0
 Requires-Dist: pyee>=9.0.0
 Requires-Dist: tldextract>=5.1.0
 Requires-Dist: typing-extensions>=4.1.0

{crawlee-1.0.1b12.dist-info → crawlee-1.0.2.dist-info}/RECORD

@@ -159,35 +159,36 @@ crawlee/storage_clients/_base/_request_queue_client.py,sha256=cgM4yk6xJwgfzP-xaN
 crawlee/storage_clients/_base/_storage_client.py,sha256=RvmKCV1U9_KxyG7n8xhClm2vwD2SKChWIiBLk6cuqw0,3523
 crawlee/storage_clients/_base/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 crawlee/storage_clients/_file_system/__init__.py,sha256=w3twfwz5YeLYeu_70pNPBRINS2wXRvzOMvA1hUDYgf0,387
-crawlee/storage_clients/_file_system/_dataset_client.py,sha256=aklivbj6MsFOLiG9muHVBpOJ3vTgLK-xb_SOHVqHWeM,17748
-crawlee/storage_clients/_file_system/_key_value_store_client.py,sha256=i3Jz7GEPoyi9DRz76TuKs4VMSzQL18rbGaF3voJG3JQ,18695
-crawlee/storage_clients/_file_system/_request_queue_client.py,sha256=2NN9you6sqfRonThcdqrEeXtU0I7YVMkxiiKJvU4TKQ,32880
+crawlee/storage_clients/_file_system/_dataset_client.py,sha256=1Z8VCDx8ueh0FQQXUr8tJlOtKw8ggkaFjuz3-T_GJDY,17735
+crawlee/storage_clients/_file_system/_key_value_store_client.py,sha256=qNa3RRJQ8Omy2AteQvYh1Td04PsP5AhUFyTpL6KQbSg,18676
+crawlee/storage_clients/_file_system/_request_queue_client.py,sha256=0TM4BFcz2knQiWyF6dH62WTTj11wN9dNtpYOiBDqkpY,32862
 crawlee/storage_clients/_file_system/_storage_client.py,sha256=My63uc513kfUPe5X-PTYWBRe9xUGnkLqJN7IcsQd2yw,3293
 crawlee/storage_clients/_file_system/_utils.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 crawlee/storage_clients/_file_system/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 crawlee/storage_clients/_memory/__init__.py,sha256=WHyBhckxdw2k0epkM_B3ymNASebNTOCU_NrvfzUAn14,355
-crawlee/storage_clients/_memory/_dataset_client.py,sha256=3LW9rB4IDIggwL8299OHUEe_2Mz0DA30vdReGSO0dvc,8836
-crawlee/storage_clients/_memory/_key_value_store_client.py,sha256=0u2y7Tl88dx3Lg0h6QA1nDA06VZiUZW3820JV8_OcUI,6487
-crawlee/storage_clients/_memory/_request_queue_client.py,sha256=yYdv6r4O7x8JcQ2v-COit-9YyB79AxKvzOJkOCcMDaY,13110
+crawlee/storage_clients/_memory/_dataset_client.py,sha256=wxuhyVnWyi9jUfI2eY4kSgD6vVje4gL2Bursto348Ps,8827
+crawlee/storage_clients/_memory/_key_value_store_client.py,sha256=rBXRggALe-0kBAe03sdLVkABhkEFmHqXRabR28IugUE,6472
+crawlee/storage_clients/_memory/_request_queue_client.py,sha256=hPI78S1sOopVKFKDRW7ndkha7TVcJfwdd4onXgfb4Pk,13096
 crawlee/storage_clients/_memory/_storage_client.py,sha256=EyiH-MgM_6iBItjmy2SkWAdjVbviacnxr3la-yiGfIw,2724
 crawlee/storage_clients/_memory/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 crawlee/storage_clients/_sql/__init__.py,sha256=X_fDMc6jn50gEBZ9QyUw54sjovYfFvE-dgXAdci6Y2M,312
 crawlee/storage_clients/_sql/_client_mixin.py,sha256=U9ThDUuRbT5JDtCFlBurhZIs1Ay5t9fTfPXXI_4dwHY,15988
-crawlee/storage_clients/_sql/_dataset_client.py,sha256=jr3hD34qhGnV7teS56tJD8psUDhp3kT12Gk-HQ0MwT4,10209
+crawlee/storage_clients/_sql/_dataset_client.py,sha256=tiJVvOPZgc7cy4kGfWnun-g2TJMHMdaLnoqns5Sl6ek,10203
 crawlee/storage_clients/_sql/_db_models.py,sha256=Gs4MS1YL0gWaUfNReVKJUXsqbU_d5jxiyvZ0sFxAV2A,9845
-crawlee/storage_clients/_sql/_key_value_store_client.py,sha256=1qwl298ThVybUJgfyQaKXMhWNOieEHa1TVeXWVb1YO0,11201
-crawlee/storage_clients/_sql/_request_queue_client.py,sha256=V2mUhQTe8pbEBb_4GNsI12i5Okr3ILLKyAHw8irdwLo,28936
+crawlee/storage_clients/_sql/_key_value_store_client.py,sha256=D0nQoStq9PR0RTn9ZORKuTcRP7X-_2aDrLKgb2hKWM0,11189
+crawlee/storage_clients/_sql/_request_queue_client.py,sha256=iavp-G62ApPtPmKePYviaNOFDXDg7QN9ozPHau6C1TY,28925
 crawlee/storage_clients/_sql/_storage_client.py,sha256=3xfgUcdW7Pu_j3SDYFzAdnU81jl1CmZ9Z5_NLvNi4P8,10913
 crawlee/storage_clients/_sql/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 crawlee/storages/__init__.py,sha256=wc2eioyCKAAYrg4N7cshpjC-UbE23OzGar9nK_kteSY,186
-crawlee/storages/_base.py,sha256=Ka9RdYnEYlNLiKJGWqSwEDOPv5AtJWXKZErSTesh42Y,2124
-crawlee/storages/_dataset.py,sha256=T2C8Q3gDUWg8p0JBS5MHnCuSd252oDU61JuNZGZzg-I,14573
-crawlee/storages/_key_value_store.py,sha256=3oI5hVoM_NpTQVKXCbQCmb0sZhW7vN2oXQo-Yxi7BLQ,10127
-crawlee/storages/_request_queue.py,sha256=jt-d-NkI9lAorLssoI2r_lZjeEipe-5Cn6z9bfQqY3k,13154
-crawlee/storages/_storage_instance_manager.py,sha256=iFX3ymsIXyTg8tMHtx5Wn9XyaC77dIf15GpuggsJPDM,7821
+crawlee/storages/_base.py,sha256=zUOcMJTg8MAzq-m9X1NJcWncCfxzI5mb5MyY35WAkMk,2310
+crawlee/storages/_dataset.py,sha256=l3VJCaJnaAEhJFpfRUOLzIbW332R8gdEPSSGhLq65pg,14652
+crawlee/storages/_key_value_store.py,sha256=ik--ZPCzOiG5hmm6k5LNH_FO9P3MoW0UvTKiGVf1RIY,10206
+crawlee/storages/_request_queue.py,sha256=bjBOGbpMaGUsqJPVB-JD2VShziPAYMI-GvWKKpylzDE,13233
+crawlee/storages/_storage_instance_manager.py,sha256=72n0YlPwNpSQDJSPf4TxnI2GvIK6L-ZiTmHRbFcoVU0,8164
+crawlee/storages/_utils.py,sha256=Yz-5tEBYKYCFJemYT29--uGJqoJLApLDLgPcsnbifRw,439
 crawlee/storages/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-crawlee-1.0.1b12.dist-info/METADATA,sha256=e_iELUmxHTEAh14KViZnMgiKSnNe4nzGAlBDel2AO0I,29315
-crawlee-1.0.1b12.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-crawlee-1.0.1b12.dist-info/entry_points.txt,sha256=1p65X3dA-cYvzjtlxLL6Kn1wpY-3uEDVqJLp53uNPeo,45
-crawlee-1.0.1b12.dist-info/licenses/LICENSE,sha256=AsFjHssKjj4LGd2ZCqXn6FBzMqcWdjQre1byPPSypVw,11355
-crawlee-1.0.1b12.dist-info/RECORD,,
+crawlee-1.0.2.dist-info/METADATA,sha256=eep5-BK6U8RekRkE33CNSr-A3LlSvuZ9vujpHJJvu-o,29320
+crawlee-1.0.2.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+crawlee-1.0.2.dist-info/entry_points.txt,sha256=1p65X3dA-cYvzjtlxLL6Kn1wpY-3uEDVqJLp53uNPeo,45
+crawlee-1.0.2.dist-info/licenses/LICENSE,sha256=AsFjHssKjj4LGd2ZCqXn6FBzMqcWdjQre1byPPSypVw,11355
+crawlee-1.0.2.dist-info/RECORD,,