e-data 2.0.0.dev126__tar.gz → 2.0.1.dev128__tar.gz

{e_data-2.0.0.dev126/e_data.egg-info → e_data-2.0.1.dev128}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: e-data
-Version: 2.0.0.dev126
+Version: 2.0.1.dev128
 Summary: Python library for managing spanish energy data from various web providers
 Author-email: VMG <vmayorg@outlook.es>
 License:                     GNU GENERAL PUBLIC LICENSE
@@ -696,7 +696,7 @@ Requires-Dist: Jinja2<4,>=3.1
 Requires-Dist: pydantic<3,>=2.10
 Requires-Dist: python_dateutil<3,>=2.8
 Requires-Dist: Requests<3,>=2.31
-Requires-Dist: SQLAlchemy<3,>=2.0
+Requires-Dist: SQLAlchemy[asyncio]<3,>=2.0
 Requires-Dist: sqlmodel<0.1,>=0.0.22
 Requires-Dist: typer<1,>=0.12
 Requires-Dist: aiosqlite<1,>=0.21

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128/e_data.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: e-data
-Version: 2.0.0.dev126
+Version: 2.0.1.dev128
 Summary: Python library for managing spanish energy data from various web providers
 Author-email: VMG <vmayorg@outlook.es>
 License:                     GNU GENERAL PUBLIC LICENSE
@@ -696,7 +696,7 @@ Requires-Dist: Jinja2<4,>=3.1
 Requires-Dist: pydantic<3,>=2.10
 Requires-Dist: python_dateutil<3,>=2.8
 Requires-Dist: Requests<3,>=2.31
-Requires-Dist: SQLAlchemy<3,>=2.0
+Requires-Dist: SQLAlchemy[asyncio]<3,>=2.0
 Requires-Dist: sqlmodel<0.1,>=0.0.22
 Requires-Dist: typer<1,>=0.12
 Requires-Dist: aiosqlite<1,>=0.21

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128}/e_data.egg-info/requires.txt

@@ -7,7 +7,7 @@ Jinja2<4,>=3.1
 pydantic<3,>=2.10
 python_dateutil<3,>=2.8
 Requests<3,>=2.31
-SQLAlchemy<3,>=2.0
+SQLAlchemy[asyncio]<3,>=2.0
 sqlmodel<0.1,>=0.0.22
 typer<1,>=0.12
 aiosqlite<1,>=0.21

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128}/edata/core/utils.py

@@ -2,6 +2,7 @@
 
 import logging
 from datetime import datetime
+from functools import lru_cache
 from pathlib import Path
 
 import holidays
@@ -19,10 +20,22 @@ TARIFF_BY_WEEKDAY = [[], [], [5, 6]]
 _LOGGER = logging.getLogger(__name__)
 
 
+@lru_cache(maxsize=None)
+def _es_holidays(year: int) -> holidays.HolidayBase:
+    """Return (and cache) Spanish holidays for a single year.
+
+    Building the holidays calendar is expensive; get_tariff is called once per
+    hourly record, so without this cache a full-history rebuild spends minutes
+    reconstructing the same handful of yearly calendars.
+    """
+
+    return holidays.country_holidays("ES", years=year)
+
+
 def get_tariff(dt: datetime) -> int:
     """Return the tariff for the selected datetime."""
 
-    hdays = holidays.country_holidays("ES")
+    hdays = _es_holidays(dt.year)
     hour = dt.hour
     weekday = dt.weekday()
 

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128}/edata/database/controller.py

@@ -3,7 +3,7 @@ import os
 import typing
 from datetime import datetime
 
-from sqlalchemy import Select, insert
+from sqlalchemy import Select, event, insert
 from sqlalchemy.exc import IntegrityError
 from sqlalchemy.ext.asyncio import AsyncEngine, create_async_engine
 from sqlmodel import SQLModel
@@ -28,6 +28,25 @@ _LOGGER = logging.getLogger(__name__)
 T = typing.TypeVar("T", bound=SQLModel)
 
 
+def _set_sqlite_pragmas(dbapi_connection, connection_record) -> None:
+    """Tune SQLite for this write-heavy, single-writer workload.
+
+    WAL lets readers proceed while a write is in flight and avoids an fsync per
+    statement; synchronous=NORMAL is durable under WAL (only a crash mid-checkpoint
+    could lose the last transaction, and the data is re-fetchable from Datadis);
+    busy_timeout avoids spurious "database is locked" errors under concurrency.
+    journal_mode is persisted in the file; the others are per-connection.
+    """
+
+    cursor = dbapi_connection.cursor()
+    try:
+        cursor.execute("PRAGMA journal_mode=WAL")
+        cursor.execute("PRAGMA synchronous=NORMAL")
+        cursor.execute("PRAGMA busy_timeout=5000")
+    finally:
+        cursor.close()
+
+
 class EdataDB:
 
     _instance = None
@@ -39,10 +58,11 @@ class EdataDB:
         if cls._instance is None:
             cls._instance = super().__new__(cls)
             cls._db_url = db_url
-            cls._engine = create_async_engine(db_url, future=True)
-            # Ensure parent directory exists
+            # Ensure parent directory exists before the first connection is opened.
             dir_path = os.path.dirname(os.path.abspath(sqlite_path))
             os.makedirs(dir_path, exist_ok=True)
+            cls._engine = create_async_engine(db_url, future=True)
+            event.listen(cls._engine.sync_engine, "connect", _set_sqlite_pragmas)
             cls._instance._tables_initialized = False
         elif db_url != cls._db_url:
             raise ValueError("EdataDB already initialized with a different db_url")

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128}/edata/services/bill_service.py

@@ -286,7 +286,7 @@ class BillService:
         missing = await self._find_missing_stats()
         for day in missing:
             _LOGGER.debug("%s updating daily statistics for date %s", self._scups, day)
-            end = day + relativedelta.relativedelta(day=1) - timedelta(minutes=1)
+            end = day + relativedelta.relativedelta(days=1) - timedelta(minutes=1)
             await self._update_daily_statistics(day, end)
 
         missing_months = list(set([get_month(x) for x in missing]))

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128}/edata/services/data_service.py

@@ -120,7 +120,7 @@ class DataService:
         missing = await self._find_missing_stats()
         for day in missing:
             _LOGGER.info("%s updating daily statistics for date %s", self._scups, day)
-            end = day + relativedelta.relativedelta(day=1) - timedelta(hours=1)
+            end = day + relativedelta.relativedelta(days=1) - timedelta(hours=1)
             await self._update_daily_statistics(day, end)
 
         missing_months = await self._find_missing_stats("month")
@@ -128,7 +128,7 @@ class DataService:
             _LOGGER.info(
                 "%s updating monthly statistics for date %s", self._scups, month
             )
-            end = month + relativedelta.relativedelta(month=1) - timedelta(hours=1)
+            end = month + relativedelta.relativedelta(months=1) - timedelta(hours=1)
             await self._update_monthly_statistics(month, end)
 
     async def update(

{e_data-2.0.0.dev126 → e_data-2.0.1.dev128}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "e-data"
-version = "2.0.0.dev126"
+version = "2.0.1.dev128"
 description = "Python library for managing spanish energy data from various web providers"
 readme = "README.md"
 authors = [
@@ -29,7 +29,7 @@ dependencies = [
     "pydantic>=2.10,<3",
     "python_dateutil>=2.8,<3",
     "Requests>=2.31,<3",
-    "SQLAlchemy>=2.0,<3",
+    "SQLAlchemy[asyncio]>=2.0,<3",
     "sqlmodel>=0.0.22,<0.1",
     "typer>=0.12,<1",
     "aiosqlite>=0.21,<1",