sportsball 0.3.16__tar.gz → 0.3.17__tar.gz

{sportsball-0.3.16/sportsball.egg-info → sportsball-0.3.17}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: sportsball
-Version: 0.3.16
+Version: 0.3.17
 Summary: A library for pulling in and normalising sports stats.
 Home-page: https://github.com/8W9aG/sportsball
 Author: Will Sackfield
@@ -244,7 +244,7 @@ Social media posts one day out from the game.
 This library uses very aggressive caching due to the large data requirements. If the requests are about a recent game (generally in the last 7 days) the caching is bypassed. The caching is as follows:
 
 1. A joblib disk cache that caches calls to pydantic model creation functions. This changes on every version update to keep the models in sync. This is the fastest cache.
-2. A requests cache backed by sqlite that caches requests for 1 year.
+2. A requests cache backed by sqlite that caches requests forever.
 3. An attempt to find the response is made to the wayback machine, and used if found.
 
 It's very recommended that the user uses proxies defined in the `PROXIES` environment variable. The more proxies the easier it is to collect data.

{sportsball-0.3.16 → sportsball-0.3.17}/README.md

@@ -191,7 +191,7 @@ Social media posts one day out from the game.
 This library uses very aggressive caching due to the large data requirements. If the requests are about a recent game (generally in the last 7 days) the caching is bypassed. The caching is as follows:
 
 1. A joblib disk cache that caches calls to pydantic model creation functions. This changes on every version update to keep the models in sync. This is the fastest cache.
-2. A requests cache backed by sqlite that caches requests for 1 year.
+2. A requests cache backed by sqlite that caches requests forever.
 3. An attempt to find the response is made to the wayback machine, and used if found.
 
 It's very recommended that the user uses proxies defined in the `PROXIES` environment variable. The more proxies the easier it is to collect data.

{sportsball-0.3.16 → sportsball-0.3.17}/setup.py

@@ -23,7 +23,7 @@ def install_requires() -> typing.List[str]:
 
 setup(
     name='sportsball',
-    version='0.3.16',
+    version='0.3.17',
     description='A library for pulling in and normalising sports stats.',
     long_description=long_description,
     long_description_content_type='text/markdown',

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/__init__.py

@@ -1,3 +1,3 @@
 """The main module for sportsball."""
 
-__VERSION__ = "0.3.16"
+__VERSION__ = "0.3.17"

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/afl/afltables/afl_afltables_league_model.py

@@ -1,5 +1,7 @@
 """AFL AFLTables league model."""
 
+# pylint: disable=too-many-statements
+import datetime
 import os
 import urllib.parse
 from typing import Iterator
@@ -34,7 +36,12 @@ class AFLAFLTablesLeagueModel(LeagueModel):
         filename, _ = os.path.splitext(last_component)
         year = int(filename)
 
-        response = self.session.get(season_url)
+        if year < datetime.datetime.now().year - 1:
+            with self.session.cache_disabled():
+                response = self.session.get(season_url)
+        else:
+            response = self.session.get(season_url)
+
         soup = BeautifulSoup(response.text, "html.parser")
         in_finals = False
         game_number = 0

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/afl/afltables/afl_afltables_venue_model.py

@@ -4,6 +4,7 @@ import datetime
 import os
 from urllib.parse import urlparse
 
+import pytest_is_running
 import requests_cache
 from bs4 import BeautifulSoup
 
@@ -12,11 +13,9 @@ from ...google.google_address_model import create_google_address_model
 from ...venue_model import VenueModel
 
 
-@MEMORY.cache(ignore=["session"])
-def create_afl_afltables_venue_model(
+def _create_afl_afltables_venue_model(
     url: str, session: requests_cache.CachedSession, dt: datetime.datetime
 ) -> VenueModel:
-    """Create a venue model from AFL tables."""
     o = urlparse(url)
     last_component = o.path.split("/")[-1]
     identifier, _ = os.path.splitext(last_component)
@@ -35,3 +34,23 @@ def create_afl_afltables_venue_model(
         is_grass=None,
         is_indoor=None,
     )
+
+
+@MEMORY.cache(ignore=["session"])
+def _cached_create_afl_afltables_venue_model(
+    url: str, session: requests_cache.CachedSession, dt: datetime.datetime
+) -> VenueModel:
+    return _create_afl_afltables_venue_model(url, session, dt)
+
+
+def create_afl_afltables_venue_model(
+    url: str, session: requests_cache.CachedSession, dt: datetime.datetime
+) -> VenueModel:
+    """Create a venue model from AFL tables."""
+    if (
+        not pytest_is_running.is_running()
+        and dt < datetime.datetime.now() - datetime.timedelta(days=7)
+    ):
+        return _cached_create_afl_afltables_venue_model(url, session, dt)
+    with session.cache_disabled():
+        return _create_afl_afltables_venue_model(url, session, dt)

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/aussportsbetting/aussportsbetting_league_model.py

@@ -88,7 +88,8 @@ class AusSportsBettingLeagueModel(LeagueModel):
 
     @property
     def games(self) -> Iterator[GameModel]:
-        response = self.session.get(self._spreadsheet_url)
+        with self.session.cache_disabled():
+            response = self.session.get(self._spreadsheet_url)
         response.raise_for_status()
         workbook = load_workbook(filename=BytesIO(response.content))
         ws = workbook.active

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/combined/combined_league_model.py

@@ -1,5 +1,6 @@
 """Combined league model."""
 
+# pylint: disable=raise-missing-from
 import logging
 import multiprocessing
 from multiprocessing import Pool

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/weather/gribstream/gribstream_weather_model.py

@@ -7,21 +7,21 @@ import json
 import os
 
 import pandas as pd
+import pytest_is_running
 import pytz
 import requests
+import requests_cache
 
 from ....cache import MEMORY
 from ...weather_model import WeatherModel
 
 
-@MEMORY.cache(ignore=["session"])
-def create_gribstream_weather_model(
-    session: requests.Session,
+def _create_gribstream_weather_model(
+    session: requests_cache.CachedSession,
     latitude: float,
     longitude: float,
     dt: datetime.datetime,
 ) -> WeatherModel | None:
-    """Create a weather model from gribstream."""
     api_key = os.environ.get("GRIBSTREAM_API_KEY")
     if api_key is None:
         return None
@@ -58,3 +58,28 @@ def create_gribstream_weather_model(
     temperature = df.iloc[idx]["TMP|2 m above ground|"]  # type: ignore
     relative_humidity = df.iloc[idx]["RH|2 m above ground|"]  # type: ignore
     return WeatherModel(temperature=temperature, relative_humidity=relative_humidity)
+
+
+@MEMORY.cache(ignore=["session"])
+def _cached_create_gribstream_weather_model(
+    session: requests_cache.CachedSession,
+    latitude: float,
+    longitude: float,
+    dt: datetime.datetime,
+) -> WeatherModel | None:
+    return _create_gribstream_weather_model(session, latitude, longitude, dt)
+
+
+def create_gribstream_weather_model(
+    session: requests_cache.CachedSession,
+    latitude: float,
+    longitude: float,
+    dt: datetime.datetime,
+) -> WeatherModel | None:
+    """Create a weather model from gribstream."""
+    if not pytest_is_running.is_running() and dt < datetime.datetime.now().replace(
+        tzinfo=dt.tzinfo
+    ) - datetime.timedelta(days=3):
+        return _cached_create_gribstream_weather_model(session, latitude, longitude, dt)
+    with session.cache_disabled():
+        return _create_gribstream_weather_model(session, latitude, longitude, dt)

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/weather/multi_weather_model.py

@@ -2,7 +2,7 @@
 
 import datetime
 
-import requests
+import requests_cache
 
 from ...cache import MEMORY
 from ..weather_model import WeatherModel
@@ -13,7 +13,7 @@ from .openmeteo.openmeteo_weather_model import create_openmeteo_weather_model
 
 @MEMORY.cache(ignore=["session"])
 def create_mutli_weather_model(
-    session: requests.Session,
+    session: requests_cache.CachedSession,
     latitude: float,
     longitude: float,
     dt: datetime.datetime,

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/data/weather/openmeteo/openmeteo_weather_model.py

@@ -4,8 +4,10 @@ import datetime
 
 import openmeteo_requests  # type: ignore
 import pandas as pd
+import pytest_is_running
 import pytz
 import requests
+import requests_cache
 from openmeteo_requests.Client import OpenMeteoRequestsError  # type: ignore
 from openmeteo_requests.Client import WeatherApiResponse
 
@@ -54,15 +56,13 @@ def _parse_openmeteo(
     return WeatherModel(temperature=temperature, relative_humidity=relative_humidity)
 
 
-@MEMORY.cache(ignore=["session"])
-def create_openmeteo_weather_model(
-    session: requests.Session,
+def _create_openmeteo_weather_model(
+    session: requests_cache.CachedSession,
     latitude: float,
     longitude: float,
     dt: datetime.datetime,
     tz: str,
 ) -> WeatherModel | None:
-    """Create a weather model from openmeteo."""
     # pylint: disable=broad-exception-caught
     client = openmeteo_requests.Client(session=session)
     try:
@@ -138,3 +138,32 @@ def create_openmeteo_weather_model(
         if "Parameter 'start_date' is out of allowed range from" in e_text:
             return None
         raise e
+
+
+@MEMORY.cache(ignore=["session"])
+def _cached_create_openmeteo_weather_model(
+    session: requests_cache.CachedSession,
+    latitude: float,
+    longitude: float,
+    dt: datetime.datetime,
+    tz: str,
+) -> WeatherModel | None:
+    return _create_openmeteo_weather_model(session, latitude, longitude, dt, tz)
+
+
+def create_openmeteo_weather_model(
+    session: requests_cache.CachedSession,
+    latitude: float,
+    longitude: float,
+    dt: datetime.datetime,
+    tz: str,
+) -> WeatherModel | None:
+    """Create a weather model from openmeteo."""
+    if not pytest_is_running.is_running() and dt < datetime.datetime.now().replace(
+        tzinfo=dt.tzinfo
+    ) - datetime.timedelta(days=3):
+        return _cached_create_openmeteo_weather_model(
+            session, latitude, longitude, dt, tz
+        )
+    with session.cache_disabled():
+        return _create_openmeteo_weather_model(session, latitude, longitude, dt, tz)

{sportsball-0.3.16 → sportsball-0.3.17}/sportsball/proxy_session.py

@@ -86,47 +86,6 @@ class ProxySession(requests_cache.CachedSession):
     def _perform_request(self, *args, **kwargs) -> requests.Response:
         return super().request(*args, **kwargs)
 
-    @retry(
-        stop=stop_after_attempt(50),
-        after=after_log(logging.getLogger(__name__), logging.DEBUG),
-        before=before_log(logging.getLogger(__name__), logging.DEBUG),
-        retry=retry_if_exception_type(FunctionTimedOut)
-        | retry_if_exception_type(requests.exceptions.ProxyError)
-        | retry_if_exception_type(requests.exceptions.ConnectionError)
-        | retry_if_exception_type(requests.exceptions.ChunkedEncodingError)
-        | retry_if_exception_type(ValueError)
-        | retry_if_exception_type(requests.exceptions.HTTPError)
-        | retry_if_exception_type(requests.exceptions.ReadTimeout),
-        reraise=True,
-    )
-    def _perform_proxy_request(self, *args, **kwargs) -> requests.Response:
-        proxy = self._suggest_proxy()
-        if proxy:
-            logging.debug("Using proxy: %s", proxy)
-            kwargs.setdefault(
-                "proxies",
-                {
-                    "http": proxy,
-                    "https": proxy,
-                },
-            )
-        kwargs.setdefault("timeout", DEFAULT_TIMEOUT)
-        kwargs.setdefault("verify", False)
-
-        try:
-            response = self._perform_request(*args, **kwargs)
-            if not _is_fast_fail_url(response.url):
-                response.raise_for_status()
-            if not response.from_cache:  # pyright: ignore
-                self._proxies.remove(proxy)
-                self._proxies.insert(0, proxy)
-            return response
-        except Exception as e:
-            logging.debug("Burning proxy: %s due to %s", proxy, str(e))
-            self._proxies.remove(proxy)
-            self._proxies.append(proxy)
-            raise e
-
     def _wayback_machine_request(
         self, method, url, **kwargs
     ) -> requests.Response | None:
@@ -205,6 +164,7 @@ class ProxySession(requests_cache.CachedSession):
         | retry_if_exception_type(ValueError)
         | retry_if_exception_type(requests.exceptions.HTTPError)
         | retry_if_exception_type(requests.exceptions.ReadTimeout),
+        reraise=True,
     )
     def _perform_retry_send(self, request: requests.PreparedRequest, **kwargs) -> Any:
         return self._perform_timeout_send(request, **kwargs)
@@ -289,5 +249,5 @@ def create_proxy_session() -> ProxySession:
     """Creates a standard proxy session."""
     return ProxySession(
         "sportsball",
-        expire_after=datetime.timedelta(days=365),
+        expire_after=requests_cache.NEVER_EXPIRE,
     )

{sportsball-0.3.16 → sportsball-0.3.17/sportsball.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: sportsball
-Version: 0.3.16
+Version: 0.3.17
 Summary: A library for pulling in and normalising sports stats.
 Home-page: https://github.com/8W9aG/sportsball
 Author: Will Sackfield
@@ -244,7 +244,7 @@ Social media posts one day out from the game.
 This library uses very aggressive caching due to the large data requirements. If the requests are about a recent game (generally in the last 7 days) the caching is bypassed. The caching is as follows:
 
 1. A joblib disk cache that caches calls to pydantic model creation functions. This changes on every version update to keep the models in sync. This is the fastest cache.
-2. A requests cache backed by sqlite that caches requests for 1 year.
+2. A requests cache backed by sqlite that caches requests forever.
 3. An attempt to find the response is made to the wayback machine, and used if found.
 
 It's very recommended that the user uses proxies defined in the `PROXIES` environment variable. The more proxies the easier it is to collect data.