atdata 0.2.2b1__py3-none-any.whl → 0.2.3b1__py3-none-any.whl

atdata/atmosphere/schema.py

@@ -17,7 +17,6 @@ from ._types import (
     LEXICON_NAMESPACE,
 )
 from .._type_utils import (
-    numpy_dtype_to_string,
     unwrap_optional,
     is_ndarray_type,
     extract_ndarray_dtype,
@@ -25,6 +24,7 @@ from .._type_utils import (
 
 # Import for type checking only to avoid circular imports
 from typing import TYPE_CHECKING
+
 if TYPE_CHECKING:
     from ..dataset import PackableSample
 
@@ -37,21 +37,19 @@ class SchemaPublisher:
     This class introspects a PackableSample class to extract its field
     definitions and publishes them as an ATProto schema record.
 
-    Example:
-        ::
-
-            >>> @atdata.packable
-            ... class MySample:
-            ...     image: NDArray
-            ...     label: str
-            ...
-            >>> client = AtmosphereClient()
-            >>> client.login("handle", "password")
-            >>>
-            >>> publisher = SchemaPublisher(client)
-            >>> uri = publisher.publish(MySample, version="1.0.0")
-            >>> print(uri)
-            at://did:plc:.../ac.foundation.dataset.sampleSchema/...
+    Examples:
+        >>> @atdata.packable
+        ... class MySample:
+        ...     image: NDArray
+        ...     label: str
+        ...
+        >>> client = AtmosphereClient()
+        >>> client.login("handle", "password")
+        >>>
+        >>> publisher = SchemaPublisher(client)
+        >>> uri = publisher.publish(MySample, version="1.0.0")
+        >>> print(uri)
+        at://did:plc:.../ac.foundation.dataset.sampleSchema/...
     """
 
     def __init__(self, client: AtmosphereClient):
@@ -90,7 +88,9 @@ class SchemaPublisher:
             TypeError: If a field type is not supported.
         """
         if not is_dataclass(sample_type):
-            raise ValueError(f"{sample_type.__name__} must be a dataclass (use @packable)")
+            raise ValueError(
+                f"{sample_type.__name__} must be a dataclass (use @packable)"
+            )
 
         # Build the schema record
         schema_record = self._build_schema_record(
@@ -155,12 +155,18 @@ class SchemaPublisher:
             return FieldType(kind="primitive", primitive="bytes")
 
         if is_ndarray_type(python_type):
-            return FieldType(kind="ndarray", dtype=extract_ndarray_dtype(python_type), shape=None)
+            return FieldType(
+                kind="ndarray", dtype=extract_ndarray_dtype(python_type), shape=None
+            )
 
         origin = get_origin(python_type)
         if origin is list:
             args = get_args(python_type)
-            items = self._python_type_to_field_type(args[0]) if args else FieldType(kind="primitive", primitive="str")
+            items = (
+                self._python_type_to_field_type(args[0])
+                if args
+                else FieldType(kind="primitive", primitive="str")
+            )
             return FieldType(kind="array", items=items)
 
         if is_dataclass(python_type):
@@ -178,16 +184,14 @@ class SchemaLoader:
     This class fetches schema records from ATProto and can list available
     schemas from a repository.
 
-    Example:
-        ::
-
-            >>> client = AtmosphereClient()
-            >>> client.login("handle", "password")
-            >>>
-            >>> loader = SchemaLoader(client)
-            >>> schema = loader.get("at://did:plc:.../ac.foundation.dataset.sampleSchema/...")
-            >>> print(schema["name"])
-            'MySample'
+    Examples:
+        >>> client = AtmosphereClient()
+        >>> client.login("handle", "password")
+        >>>
+        >>> loader = SchemaLoader(client)
+        >>> schema = loader.get("at://did:plc:.../ac.foundation.dataset.sampleSchema/...")
+        >>> print(schema["name"])
+        'MySample'
     """
 
     def __init__(self, client: AtmosphereClient):

atdata/atmosphere/store.py

@@ -6,23 +6,20 @@ protocol that stores dataset shards as ATProto blobs in a Personal Data Server.
 This enables fully decentralized dataset storage where both metadata (records)
 and data (blobs) live on the AT Protocol network.
 
-Example:
-    ::
-
-        >>> from atdata.atmosphere import AtmosphereClient, PDSBlobStore
-        >>>
-        >>> client = AtmosphereClient()
-        >>> client.login("handle.bsky.social", "app-password")
-        >>>
-        >>> store = PDSBlobStore(client)
-        >>> urls = store.write_shards(dataset, prefix="mnist/v1")
-        >>> print(urls)
-        ['at://did:plc:.../blob/bafyrei...', ...]
+Examples:
+    >>> from atdata.atmosphere import AtmosphereClient, PDSBlobStore
+    >>>
+    >>> client = AtmosphereClient()
+    >>> client.login("handle.bsky.social", "app-password")
+    >>>
+    >>> store = PDSBlobStore(client)
+    >>> urls = store.write_shards(dataset, prefix="mnist/v1")
+    >>> print(urls)
+    ['at://did:plc:.../blob/bafyrei...', ...]
 """
 
 from __future__ import annotations
 
-import io
 import tempfile
 from dataclasses import dataclass
 from typing import TYPE_CHECKING, Any
@@ -31,6 +28,7 @@ import webdataset as wds
 
 if TYPE_CHECKING:
     from ..dataset import Dataset
+    from .._sources import BlobSource
     from .client import AtmosphereClient
 
 
@@ -48,13 +46,11 @@ class PDSBlobStore:
     Attributes:
         client: Authenticated AtmosphereClient instance.
 
-    Example:
-        ::
-
-            >>> store = PDSBlobStore(client)
-            >>> urls = store.write_shards(dataset, prefix="training/v1")
-            >>> # Returns AT URIs like:
-            >>> # ['at://did:plc:abc/blob/bafyrei...', ...]
+    Examples:
+        >>> store = PDSBlobStore(client)
+        >>> urls = store.write_shards(dataset, prefix="training/v1")
+        >>> # Returns AT URIs like:
+        >>> # ['at://did:plc:abc/blob/bafyrei...', ...]
     """
 
     client: "AtmosphereClient"

atdata/cli/__init__.py

@@ -42,7 +42,8 @@ def main(argv: Sequence[str] | None = None) -> int:
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
     parser.add_argument(
-        "--version", "-v",
+        "--version",
+        "-v",
         action="store_true",
         help="Show version information",
     )
@@ -83,7 +84,8 @@ def main(argv: Sequence[str] | None = None) -> int:
         help="MinIO console port (default: 9001)",
     )
     up_parser.add_argument(
-        "--detach", "-d",
+        "--detach",
+        "-d",
         action="store_true",
         default=True,
         help="Run containers in detached mode (default: True)",
@@ -95,7 +97,8 @@ def main(argv: Sequence[str] | None = None) -> int:
         help="Stop local development containers",
     )
     down_parser.add_argument(
-        "--volumes", "-v",
+        "--volumes",
+        "-v",
         action="store_true",
         help="Also remove volumes (deletes all data)",
     )
@@ -165,10 +168,12 @@ def _cmd_version() -> int:
     """Show version information."""
     try:
         from atdata import __version__
+
         version = __version__
     except ImportError:
         # Fallback to package metadata
         from importlib.metadata import version as pkg_version
+
         version = pkg_version("atdata")
 
     print(f"atdata {version}")
@@ -183,6 +188,7 @@ def _cmd_local_up(
 ) -> int:
     """Start local development infrastructure."""
     from .local import local_up
+
     return local_up(
         redis_port=redis_port,
         minio_port=minio_port,
@@ -194,18 +200,21 @@ def _cmd_local_up(
 def _cmd_local_down(remove_volumes: bool) -> int:
     """Stop local development infrastructure."""
     from .local import local_down
+
     return local_down(remove_volumes=remove_volumes)
 
 
 def _cmd_local_status() -> int:
     """Show status of local infrastructure."""
     from .local import local_status
+
     return local_status()
 
 
 def _cmd_diagnose(host: str, port: int) -> int:
     """Diagnose Redis configuration."""
     from .diagnose import diagnose_redis
+
     return diagnose_redis(host=host, port=port)
 
 

atdata/cli/diagnose.py

@@ -5,7 +5,6 @@ and other infrastructure components.
 """
 
 import sys
-from typing import Any
 
 
 def _print_status(label: str, ok: bool, detail: str = "") -> None:
@@ -41,6 +40,7 @@ def diagnose_redis(host: str = "localhost", port: int = 6379) -> int:
     # Try to connect
     try:
         from redis import Redis
+
         redis = Redis(host=host, port=port, socket_connect_timeout=5)
         redis.ping()
         _print_status("Connection", True, "connected")
@@ -70,7 +70,7 @@ def diagnose_redis(host: str = "localhost", port: int = 6379) -> int:
         _print_status(
             "AOF Persistence",
             aof_ok,
-            "enabled" if aof_ok else "DISABLED - data may be lost on restart!"
+            "enabled" if aof_ok else "DISABLED - data may be lost on restart!",
         )
         if not aof_ok:
             issues_found = True
@@ -85,7 +85,7 @@ def diagnose_redis(host: str = "localhost", port: int = 6379) -> int:
         _print_status(
             "RDB Persistence",
             rdb_ok,
-            f"configured ({save_config})" if rdb_ok else "DISABLED"
+            f"configured ({save_config})" if rdb_ok else "DISABLED",
         )
         # RDB disabled is only a warning if AOF is enabled
     except Exception as e:
@@ -95,7 +95,13 @@ def diagnose_redis(host: str = "localhost", port: int = 6379) -> int:
     try:
         policy = redis.config_get("maxmemory-policy").get("maxmemory-policy", "unknown")
         # Safe policies that won't evict index data
-        safe_policies = {"noeviction", "volatile-lru", "volatile-lfu", "volatile-ttl", "volatile-random"}
+        safe_policies = {
+            "noeviction",
+            "volatile-lru",
+            "volatile-lfu",
+            "volatile-ttl",
+            "volatile-random",
+        }
         policy_ok = policy in safe_policies
 
         if policy_ok:
@@ -104,7 +110,7 @@ def diagnose_redis(host: str = "localhost", port: int = 6379) -> int:
             _print_status(
                 "Memory Policy",
                 False,
-                f"{policy} - may evict index data! Use 'noeviction' or 'volatile-*'"
+                f"{policy} - may evict index data! Use 'noeviction' or 'volatile-*'",
             )
             issues_found = True
     except Exception as e:
@@ -141,9 +147,7 @@ def diagnose_redis(host: str = "localhost", port: int = 6379) -> int:
         for key in redis.scan_iter(match="LocalSchema:*", count=100):
             schema_count += 1
         _print_status(
-            "atdata Keys",
-            True,
-            f"{dataset_count} datasets, {schema_count} schemas"
+            "atdata Keys", True, f"{dataset_count} datasets, {schema_count} schemas"
         )
     except Exception as e:
         _print_status("atdata Keys", False, f"check failed: {e}")

atdata/cli/local.py

@@ -144,7 +144,9 @@ def _run_compose(
         elif shutil.which("docker-compose"):
             base_cmd = ["docker-compose"]
         else:
-            raise RuntimeError("Neither 'docker compose' nor 'docker-compose' available")
+            raise RuntimeError(
+                "Neither 'docker compose' nor 'docker-compose' available"
+            )
     else:
         raise RuntimeError("Docker not found")
 
@@ -195,6 +197,7 @@ def local_up(
 
     # Wait a moment for containers to be healthy
     import time
+
     time.sleep(2)
 
     # Show status