Refactor pt2 [untested]

.github/workflows/python-package.yml

@@ -16,7 +16,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.10", "3.11", "3.12"]
+        python-version: [ "3.12" ]
 
     steps:
     - uses: actions/checkout@v4
@@ -27,8 +27,9 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        python -m pip install flake8 pytest build
-        if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
+        python -m pip install flake8 pytest build poetry
+        poetry install
+        poetry run post-install
     - name: Lint with flake8
       run: |
         # stop the build if there are Python syntax errors or undefined names
@@ -37,7 +38,7 @@ jobs:
         flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
     - name: Test with pytest
       run: |
-        if [ -d tests ]; then pytest; fi
+        if [ -d tests ]; then poetry run pytest; fi
     - name: Build package
       run: python -m build
     - name: Upload build artifacts

pyball/__init__.py

@@ -1,13 +1,12 @@
-name = "pyball"
-from .playerid_lookup import *
-from .utils import *
-from .batting_stats import *
-from .pitching_stats import *
-from .team_batting_stats import *
-from .team_pitching_stats import *
-from .savant import *
+"""
+This is the pyball module.
+"""
 
 import subprocess
 
+
 def post_install():
-    subprocess.run(["playwright", "install"], check=True)
+    """
+    Run the playwright install command.
+    """
+    subprocess.run(["playwright", "install"], check=True)

pyball/baseball_reference_player.py

@@ -0,0 +1,137 @@
+# File: baseball_reference_player.py
+# Author: Gabriel DiFiore <difioregabe@gmail.com>
+# (c) 2022-2024
+#
+# Description: File containing functions to obtain player stats from Baseball-Reference
+
+from typing import Optional
+import logging
+import pandas as pd
+from bs4 import BeautifulSoup
+
+from pyball.utils import read_url, is_bbref_player_url
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+
+class BaseballReferencePlayerStatsScraper:
+    """
+    A class for scraping player statistics from Baseball-Reference.
+
+    This class provides methods to extract batting and pitching statistics
+    for baseball players from their Baseball-Reference profile pages.
+
+    Methods:
+    --------
+    batting_stats() -> Optional[pd.DataFrame]:
+        Retrieves the batting statistics for the player.
+
+    pitching_stats() -> Optional[pd.DataFrame]:
+        Retrieves the pitching statistics for the player.
+    """
+
+    TABLE_IDS = {
+        'batting': 'batting_standard',
+        'pitching': 'pitching_standard'
+    }
+
+    def __init__(self, url: str):
+        """
+        Initializes a new instance of the BaseballReferencePlayerStatsScraper class.
+
+        Parameters:
+        -----------
+        url : str
+            The URL of the Baseball-Reference profile page for the player.
+
+        Raises:
+        -------
+        ValueError:
+            If the provided URL is invalid.
+        """
+        if not is_bbref_player_url(url):
+            raise ValueError(f"Invalid player URL: {url}")
+        self.url = url
+        self.soup = self._get_soup()
+
+    def _get_soup(self) -> Optional[BeautifulSoup]:
+        """
+        Retrieves the BeautifulSoup object for the player's profile page.
+
+        Returns:
+        --------
+        Optional[BeautifulSoup]:
+            The BeautifulSoup object representing the player's profile page, or None if retrieval failed.
+        """
+        soup = read_url(self.url)
+        if soup is None:
+            logger.warning("Failed to retrieve content from URL: %s", self.url)
+        return soup
+
+    def _find_table(self, table_id: str) -> Optional[BeautifulSoup]:
+        """
+        Finds the HTML table element with the specified ID.
+
+        Parameters:
+        -----------
+        table_id : str
+            The ID of the table to find.
+
+        Returns:
+        --------
+        Optional[BeautifulSoup]:
+            The BeautifulSoup object representing the found table, or None if not found.
+        """
+        return self.soup.find("table", id=self.TABLE_IDS[table_id])
+
+    def _get_dataframe(self, table_id: str) -> Optional[pd.DataFrame]:
+        """
+        Parses the HTML table and returns it as a pandas DataFrame.
+
+        Parameters:
+        -----------
+        table_id : str
+            The ID of the table to parse.
+
+        Returns:
+        --------
+        Optional[pd.DataFrame]:
+            The parsed table as a pandas DataFrame, or None if parsing failed.
+        """
+        table = self._find_table(table_id)
+        if table is None:
+            logger.warning("%s stats table not found for URL: %s", table_id.capitalize(), self.url)
+            return None
+
+        try:
+            df = pd.read_html(str(table))[0]
+            return df.dropna(how="all")
+        except ValueError as e:
+            logger.error("Error parsing %s stats table (no tables found): %s", table_id, str(e))
+            return None
+        except Exception as e:
+            logger.error("Error parsing %s stats table: %s", table_id, str(e))
+            return None
+
+    def batting_stats(self) -> Optional[pd.DataFrame]:
+        """
+        Retrieves the batting statistics for the player.
+
+        Returns:
+        --------
+        Optional[pd.DataFrame]:
+            The batting statistics for the player as a pandas DataFrame, or None if not available.
+        """
+        return self._get_dataframe('batting')
+
+    def pitching_stats(self) -> Optional[pd.DataFrame]:
+        """
+        Retrieves the pitching statistics for the player.
+
+        Returns:
+        --------
+        Optional[pd.DataFrame]:
+            The pitching statistics for the player as a pandas DataFrame, or None if not available.
+        """
+        return self._get_dataframe('pitching')

pyball/baseball_reference_team.py

@@ -0,0 +1,90 @@
+# File: baseball_reference_team.py
+# Author: Gabriel DiFiore <difioregabe@gmail.com>
+# (c) 2022-2024
+#
+# Description: File containing functions to obtain team stats from Baseball-Reference
+
+from typing import Optional
+import logging
+import pandas as pd
+from bs4 import BeautifulSoup
+
+from pyball.utils import read_url, is_bbref_team_url
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+
+class BaseballReferenceTeamStatsScraper:
+    """
+    A class for scraping team statistics from Baseball-Reference.
+
+    This class provides methods to extract batting and pitching statistics
+    for baseball teams from their Baseball-Reference pages.
+
+    Methods:
+    --------
+    batting_stats(self) -> Optional[pd.DataFrame]
+        Returns the batting stats for the team as a pandas DataFrame.
+
+    pitching_stats(self) -> Optional[pd.DataFrame]
+        Returns the pitching stats for the team as a pandas DataFrame.
+    """
+
+    TABLE_IDS = {
+        'batting': 'team_batting',
+        'pitching': 'team_pitching'
+    }
+
+    def __init__(self, url: str):
+        if not is_bbref_team_url(url):
+            raise ValueError(f"Invalid team URL: {url}")
+        self.url = url
+        self.soup = self._get_soup()
+
+    def _get_soup(self) -> Optional[BeautifulSoup]:
+        soup = read_url(self.url)
+        if soup is None:
+            logger.warning("Failed to retrieve content from URL: %s", self.url)
+        return soup
+
+    def _find_table(self, table_id: str) -> Optional[BeautifulSoup]:
+        return self.soup.find("table", id=self.TABLE_IDS[table_id])
+
+    def _get_dataframe(self, table_id: str) -> Optional[pd.DataFrame]:
+        table = self._find_table(table_id)
+        if table is None:
+            logger.warning("%s stats table not found for URL: %s", table_id.capitalize(), self.url)
+            return None
+
+        try:
+            df = pd.read_html(str(table))[0]
+            return df.dropna(how="all")
+        except ValueError as e:
+            logger.error("Error parsing %s stats table (no tables found): %s", table_id, str(e))
+            return None
+        except Exception as e:
+            logger.error("Error parsing %s stats table: %s", table_id, str(e))
+            return None
+
+    def batting_stats(self) -> Optional[pd.DataFrame]:
+        """
+        Return the (Baseball-Reference) batting stats for a team as a pandas dataframe
+
+        Returns:
+        --------
+        Optional[pd.DataFrame]
+            Contains the batting stats for the team, or None if not available
+        """
+        return self._get_dataframe('batting')
+
+    def pitching_stats(self) -> Optional[pd.DataFrame]:
+        """
+        Return the (Baseball-Reference) pitching stats for a team as a pandas dataframe
+
+        Returns:
+        --------
+        Optional[pd.DataFrame]
+            Contains the pitching stats for the team, or None if not available
+        """
+        return self._get_dataframe('pitching')