From 1231ed3b735789ae1a4f41ad4b5cba2e3003bdb4 Mon Sep 17 00:00:00 2001
From: Sean Kim <33474168+seankim658@users.noreply.github.com>
Date: Mon, 4 Nov 2024 13:20:58 -0500
Subject: [PATCH 1/3] Refactor to parse title date

---
 kenpompy/FanMatch.py | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/kenpompy/FanMatch.py b/kenpompy/FanMatch.py
index 799fdbb..16572ef 100644
--- a/kenpompy/FanMatch.py
+++ b/kenpompy/FanMatch.py
@@ -5,6 +5,7 @@
 import pandas as pd
 from io import StringIO
 import re
+from datetime import datetime
 from cloudscraper import CloudScraper
 from bs4 import BeautifulSoup
 from typing import Optional
@@ -57,13 +58,15 @@ def __init__(self, browser: CloudScraper, date: Optional[str]=None):
         fm = BeautifulSoup(get_html(browser, self.url), "html.parser")
         if "Sorry, no games today." in fm.text:
             return
-        time_header = fm.find('th', string=re.compile(r"Time \(ET\)"))
-        if time_header and time_header.find("a"):
-            href = time_header.find("a")["href"]
-            date_match = re.search(r'd=(\d{4}-\d{2}-\d{2})', href)
+        if date is not None:
+            date_text = fm.find("div", class_="lh12").get_text()
+            date_match = re.search(r"for \w+, (\w+ \d{1,2}[a-z]{2})", date_text)
             if date_match:
-                found_date = date_match.group(1)
-                if found_date != date:
+                extracted_date_str = re.sub(r"(st|nd|rd|th)", "", date_match.group(1))
+                extracted_date = datetime.strptime(extracted_date_str, "%B %d")
+                extracted_mmdd = extracted_date.strftime("%m-%d")
+                user_mmdd = datetime.strptime(date, "%Y-%m-%d").strftime("%m-%d")
+                if extracted_mmdd != user_mmdd:
                     return
         table = fm.find_all("table")[0]
         fm_df = pd.read_html(StringIO(str(table)))

From a2e1751f06fa9cc1c20d18aa30b644a5ee135543 Mon Sep 17 00:00:00 2001
From: Sean Kim <33474168+seankim658@users.noreply.github.com>
Date: Mon, 4 Nov 2024 14:03:33 -0500
Subject: [PATCH 2/3] Handle predicted posessions

---
 kenpompy/FanMatch.py | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)

diff --git a/kenpompy/FanMatch.py b/kenpompy/FanMatch.py
index 16572ef..553edba 100644
--- a/kenpompy/FanMatch.py
+++ b/kenpompy/FanMatch.py
@@ -130,13 +130,11 @@ def __init__(self, browser: CloudScraper, date: Optional[str]=None):
         pos = fm_df.Game.str.split(r" \[").str[1]
         fm_df["Game"], fm_df["Possessions"] = fm_df.Game.str.split(r" \[").str[0], pos.astype("str")
         fm_df.Possessions = fm_df.Possessions.str.rstrip(r"\] ")
-        predict_info = fm_df.Prediction.str.split()
-        pred_winner = fm_df.Prediction.astype("str").str.split().str[0:-2].tolist()
-        pred_winner = [" ".join(i) if not any(pd.isnull(i)) else float("nan") for i in pred_winner]
-        fm_df["PredictedWinner"] = pred_winner
-        fm_df["PredictedScore"] = fm_df.Prediction.str.split().str[-2]
-        fm_df["WinProbability"] = fm_df.Prediction.str.split().str[-1]
-        fm_df.WinProbability = fm_df.WinProbability.str.strip("()")
+        fm_df["PredictedWinner"] = fm_df["Prediction"].str.extract(r"^(.+?) \d+-\d+")[0]
+        fm_df["PredictedScore"] = fm_df["Prediction"].str.extract(r" (\d+-\d+)")[0]
+        fm_df["WinProbability"] = fm_df["Prediction"].str.extract(r"\((\d+%)\)")[0]
+        fm_df["PredictedPossessions"] = fm_df["Prediction"].str.extract(r"\[(\d+)\]")[0].astype(float)
+        fm_df["Possessions"] = fm_df["Possessions"].where(fm_df["Possessions"] != "", fm_df["PredictedPossessions"])
 
         fm_df["PredictedMOV"] = [(int(x[0]) - int(x[1])) if len(x) > 1 else float("nan") for x 
                                  in fm_df.PredictedScore.astype("str").str.split("-")]
@@ -146,11 +144,12 @@ def __init__(self, browser: CloudScraper, date: Optional[str]=None):
         # Parse predicted loser.
         teams = fm_df.Game.str.split(", ").tolist()
         teams_np = fm_df.Game.str.split(" at ").tolist()
+        pred_winner = fm_df["PredictedWinner"].tolist()
         
         i = 0
         pred_loser = []
-        for x in teams:
-            if not len(x) == 2:
+        for i, x in enumerate(teams):
+            if len(x) != 2:
                 x = teams_np[i]
                 
                 # Account for neutral games.

From dba64a01f864865bb66ec538625403648e50333b Mon Sep 17 00:00:00 2001
From: Sean Kim <33474168+seankim658@users.noreply.github.com>
Date: Mon, 4 Nov 2024 18:27:02 -0500
Subject: [PATCH 3/3] Update test for predicted possessions field

---
 tests/test_fanmatch.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tests/test_fanmatch.py b/tests/test_fanmatch.py
index fa1a885..9c693d5 100644
--- a/tests/test_fanmatch.py
+++ b/tests/test_fanmatch.py
@@ -17,6 +17,7 @@ def test_fanmatch(browser):
 				"Marquette",
 				"73-72",
 				"51%",
+                "nan",
 				"1",
 				"Xavier",
 				"2OT",