sibi-dst 2025.9.13__py3-none-any.whl → 2025.9.14__py3-none-any.whl

sibi_dst/df_helper/_df_helper.py

@@ -137,7 +137,7 @@ class DfHelper(ManagedResource):
     def __init__(self, backend="sqlalchemy", **kwargs):
         self.default_config = self.default_config or {}
         kwargs = {**self.default_config.copy(), **kwargs}
-        kwargs.setdefault("auto_sse", True)
+        kwargs.setdefault("auto_sse", False)
         super().__init__(**kwargs)
         self.backend = backend
 

sibi_dst/utils/base.py

@@ -14,7 +14,18 @@ from sibi_dst.utils import Logger
 
 # --------- Minimal built-in SSE sink (used when auto_sse=True) ----------
 class _QueueSSE:
-    """Async queue–backed SSE sink: async send/put, async iterator, graceful close."""
+    """
+    Handles asynchronous streaming of events with structured data.
+
+    This class provides the ability to manage an asynchronous queue for handling
+    streamed Server-Sent Events (SSE). It supports operations like sending events
+    with associated data, manually enqueuing items, and iterating over items in an
+    asynchronous loop. The class also includes mechanisms for clean closure of the
+    stream.
+
+    :ivar q: An asynchronous queue used to store events and data.
+    :type q: asyncio.Queue
+    """
     __slots__ = ("q", "_closed")
 
     def __init__(self) -> None:
@@ -46,11 +57,31 @@ class _QueueSSE:
 # ------------------------------ Base class ------------------------------
 class ManagedResource(abc.ABC):
     """
-    Owns a logger and optional fsspec filesystem. Can emit SSE events via:
-    - an async emitter callable:  await emitter(event, data)
-    - a sink exposing async send(event, data) or async put(item)
-    If neither is provided and auto_sse=True, a queue-backed sink is created eagerly.
-    Thread-safe lifecycle, CM support, GC finalizer.
+    Management of shared resources with configurable verbosity, logging,
+    and support for external file systems and server-sent events (SSE).
+
+    This class is designed to assist in managing resources such as logging,
+    file systems, and SSE within an asynchronous or synchronous environment.
+    It provides facilities for handling resource lifecycle, introspection,
+    and cleanup while ensuring resources are appropriately managed. The class
+    also supports lazy initialization of external dependencies via factories.
+
+    :ivar verbose: Controls verbosity of logging or operations. If set to True,
+        more detailed logging/output will be generated.
+    :type verbose: bool
+    :ivar debug: Enables debug-level logging and internal diagnostics when True.
+        Typically used for troubleshooting purposes.
+    :type debug: bool
+    :ivar logger: The logger instance used for this resource. If left unset,
+        a default logger will be created.
+    :type logger: Optional[Logger]
+    :ivar fs: The file system interface being used. Typically an instance of
+        `fsspec.AbstractFileSystem`. If not provided, it may be created lazily
+        using a supplied factory function.
+    :type fs: Optional[fsspec.AbstractFileSystem]
+    :ivar emitter: A callable, potentially asynchronous, function for emitting
+        events. Events are sent as a combination of event names and payload data.
+    :type emitter: Optional[Callable[[str, Dict[str, Any]], Awaitable[None]]]
     """
 
     __slots__ = (
@@ -74,28 +105,23 @@ class ManagedResource(abc.ABC):
         debug: bool = False,
         log_cleanup_errors: bool = True,
         logger: Optional[Logger] = None,
-        # filesystem
         fs: Optional[fsspec.AbstractFileSystem] = None,
         fs_factory: Optional[Callable[[], fsspec.AbstractFileSystem]] = None,
-        # SSE
         emitter: Optional[Callable[[str, Dict[str, Any]], Awaitable[None]]] = None,
         emitter_factory: Optional[Callable[[], Callable[[str, Dict[str, Any]], Awaitable[None]]]] = None,
         sse: Optional[object] = None,
         sse_factory: Optional[Callable[[], object]] = None,
-        auto_sse: bool = False,           # eager auto-create if no emitter/sink is supplied
+        auto_sse: bool = False,
         **_: object,
     ) -> None:
-        # flags
         self.verbose = verbose
         self.debug = debug
         self._log_cleanup_errors = log_cleanup_errors
 
-        # lifecycle
         self._is_closed = False
         self._closing = False
         self._close_lock = threading.RLock()
 
-        # logger
         if logger is None:
             self.logger = Logger.default_logger(logger_name=self.__class__.__name__)
             self._owns_logger = True
@@ -105,7 +131,6 @@ class ManagedResource(abc.ABC):
             self.logger = logger
             self._owns_logger = False
 
-        # fs
         self.fs: Optional[fsspec.AbstractFileSystem] = None
         self._fs_factory = None
         self._owns_fs = False
@@ -119,7 +144,6 @@ class ManagedResource(abc.ABC):
             self._fs_factory = fs_factory
             self._owns_fs = True
 
-        # sse / emitter
         self._sse: Optional[object] = None
         self._sse_factory: Optional[Callable[[], object]] = None
         self._owns_sse = False
@@ -140,16 +164,15 @@ class ManagedResource(abc.ABC):
             self._sse_factory = sse_factory
             self._owns_sse = True
 
-        # EAGER auto-SSE: create sink+emitter now if none supplied
         if self._auto_sse and self._sse is None and self._emitter is None and self._sse_factory is None:
             self._create_auto_sse()
 
-        # GC finalizer
+        # Garbage Collector finaliser
         self._finalizer = weakref.finalize(self, self._finalize_static, weakref.ref(self))
 
         if self.debug:
             with contextlib.suppress(Exception):
-                self.logger.debug("Initialized %s %s", self.__class__.__name__, repr(self))
+                self.logger.debug("Initialised %s %s", self.__class__.__name__, repr(self))
 
     # ---------- Introspection ----------
     @property
@@ -222,7 +245,6 @@ class ManagedResource(abc.ABC):
 
     # ---------- SSE ----------
     def _create_auto_sse(self) -> None:
-        # internal helper: create queue sink + emitter, mark as owned
         sink = _QueueSSE()
         self._sse = sink
         self._owns_sse = True

sibi_dst/utils/dask_utils.py

@@ -5,7 +5,6 @@ import logging
 from typing import List, Any, Dict
 
 import dask
-# dask.config.set({"distributed.worker.daemon": False})
 import dask.dataframe as dd
 
 def _to_int_safe(x) -> int:
@@ -158,43 +157,3 @@ async def shared_dask_session(**kwargs):
         yield mixin.dask_client
     finally:
         mixin._close_dask_client()
-
-# from contextlib import suppress
-# from dask.distributed import Client, get_client
-#
-# class DaskClientMixin:
-#     """
-#     Provides shared Dask client lifecycle management.
-#     Ensures reuse of existing client when available, otherwise creates a lightweight local one.
-#     """
-#
-#     def _init_dask_client(self, dask_client=None, logger=None):
-#         self.dask_client = dask_client
-#         self.own_dask_client = False
-#         self.logger = logger
-#
-#         if self.dask_client is None:
-#             with suppress(ValueError, RuntimeError):
-#                 # Try to attach to an existing active client if running inside a Dask context
-#                 self.dask_client = get_client()
-#
-#         if self.dask_client is None:
-#             # Start a local in-process scheduler for fallback
-#             self.dask_client = Client(processes=False)
-#             self.own_dask_client = True
-#             if self.logger:
-#                 self.logger.info(f"Started local Dask client: {self.dask_client.dashboard_link}")
-#         else:
-#             if self.logger:
-#                 self.logger.debug(f"Using existing Dask client: {self.dask_client.dashboard_link}")
-#
-#     def _close_dask_client(self):
-#         """Close client only if this instance created it."""
-#         if getattr(self, "own_dask_client", False) and self.dask_client is not None:
-#             try:
-#                 self.dask_client.close()
-#                 if self.logger:
-#                     self.logger.info("Closed local Dask client.")
-#             except Exception as e:
-#                 if self.logger:
-#                     self.logger.warning(f"Error while closing Dask client: {e}")

{sibi_dst-2025.9.13.dist-info → sibi_dst-2025.9.14.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sibi-dst
-Version: 2025.9.13
+Version: 2025.9.14
 Summary: A data science toolkit for scalable data processing and analysis.
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown

{sibi_dst-2025.9.13.dist-info → sibi_dst-2025.9.14.dist-info}/RECORD

@@ -2,7 +2,7 @@ sibi_dst/__init__.py,sha256=QQVT3Xlj8iZN17sSMfRQFSb_DHr8A7giJP8hn02K2Oo,585
 sibi_dst/df_helper/__init__.py,sha256=7rUdMybgCNZhQL_J7IFTTHz_xtFin81xavi5-PUExkA,463
 sibi_dst/df_helper/_artifact_updater_async.py,sha256=AZp0vM3vji0tjiaScr8a9SUMH15NjPIKYPdRQ7SJe3Y,11372
 sibi_dst/df_helper/_artifact_updater_threaded.py,sha256=M5GNZismOqMmBrcyfolP1DPv87VILQf_P18is_epn50,7238
-sibi_dst/df_helper/_df_helper.py,sha256=rgVP4ggiCW6tTHmUz2UqUvLznwOtY5IyoVS3WSlg73U,17005
+sibi_dst/df_helper/_df_helper.py,sha256=30J6f7TwfOkCXAUbk3QS-K_5ouYx2tc8BGXrcvNwvC8,17006
 sibi_dst/df_helper/_parquet_artifact.py,sha256=UXkhDSAVRNKp9DykVhJd3agnryCZT0Sj2qhdhUZomuM,19421
 sibi_dst/df_helper/_parquet_reader.py,sha256=RI1e7S7u5RLqkdOD5apcuXel7KtefglS9bhzCO_TB_k,3259
 sibi_dst/df_helper/backends/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -38,11 +38,11 @@ sibi_dst/tests/test_baseclass.py,sha256=5huAwjWo_SOEZR2_0y5w9qUmw5G7pVdm8X1OTG87
 sibi_dst/tests/test_data_wrapper_class.py,sha256=6uFmZR2DxnxQz49L5jT2ehlKvlLnpUHMLFB_PqqUq7k,3336
 sibi_dst/utils/__init__.py,sha256=eoW7iROrCUVYjNT1owMgGvW6U7lolbNy3FrBb_wInPs,1304
 sibi_dst/utils/async_utils.py,sha256=53aywfgq1Q6-0OVr9qR1Sf6g7Qv3I9qunAAR4fjFXBE,351
-sibi_dst/utils/base.py,sha256=sFngliI7Ku8bZMz0YdVhppuaPNZ0dvqRwCsPe9XdF1A,16256
+sibi_dst/utils/base.py,sha256=ycaaXlXTH4tn3fM954jGx8-zKWEmdjJkJKArQ2rfAH0,17527
 sibi_dst/utils/business_days.py,sha256=DPZExTXTt7n3IbAaEuVacm-vZgbR_Ug2bJTPBUaoP3g,6694
 sibi_dst/utils/clickhouse_writer.py,sha256=8W_dTEOKQp4pXANznVSxRqFA2H5oD8UJifiBAONpXWY,17001
 sibi_dst/utils/credentials.py,sha256=cHJPPsmVyijqbUQIq7WWPe-lIallA-mI5RAy3YUuRME,1724
-sibi_dst/utils/dask_utils.py,sha256=FT85GPvHxaR7BDhxgeQT4mgQ_lFmNT5CVmiyhRv9Fns,7251
+sibi_dst/utils/dask_utils.py,sha256=t6pXmg-xDBhVi_CLg8_bw3pCy_CyM7bhbS4pbZG3Mbs,5500
 sibi_dst/utils/data_from_http_source.py,sha256=AcpKNsqTgN2ClNwuhgUpuNCx62r5_DdsAiKY8vcHEBA,1867
 sibi_dst/utils/data_utils.py,sha256=7bLidEjppieNoozDFb6OuRY0W995cxg4tiGAlkGfePI,7768
 sibi_dst/utils/data_wrapper.py,sha256=9HTuDXgvfhmFAOyNG_GEOaHuojxE3639yyzOoBt7Unc,18000
@@ -94,7 +94,7 @@ sibi_dst/v2/df_helper/core/_params_config.py,sha256=DYx2drDz3uF-lSPzizPkchhy-kxR
 sibi_dst/v2/df_helper/core/_query_config.py,sha256=Y8LVSyaKuVkrPluRDkQoOwuXHQxner1pFWG3HPfnDHM,441
 sibi_dst/v2/utils/__init__.py,sha256=6H4cvhqTiFufnFPETBF0f8beVVMpfJfvUs6Ne0TQZNY,58
 sibi_dst/v2/utils/log_utils.py,sha256=rfk5VsLAt-FKpv6aPTC1FToIPiyrnHAFFBAkHme24po,4123
-sibi_dst-2025.9.13.dist-info/METADATA,sha256=v8QE7j-yP0-1s6i4UcOhHIjsDpvSJGtBvzSkLy3f-v8,2413
-sibi_dst-2025.9.13.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-sibi_dst-2025.9.13.dist-info/top_level.txt,sha256=g3Cj4R-rciuNyJgcxuxNgw5nhN0n4TCB0ujcTEjZNiU,9
-sibi_dst-2025.9.13.dist-info/RECORD,,
+sibi_dst-2025.9.14.dist-info/METADATA,sha256=pKn_x6mtctOqGIy4MDZgy9kFkR9vsRTneFkwNFjJAj8,2413
+sibi_dst-2025.9.14.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+sibi_dst-2025.9.14.dist-info/top_level.txt,sha256=g3Cj4R-rciuNyJgcxuxNgw5nhN0n4TCB0ujcTEjZNiU,9
+sibi_dst-2025.9.14.dist-info/RECORD,,