Update app.py

app.py

@@ -4,7 +4,7 @@ import pytz
 import yaml
 import os
 from typing import Dict, Union
-
+from bs4 import BeautifulSoup
 from smolagents import CodeAgent, DuckDuckGoSearchTool, HfApiModel, load_tool, tool
 from tools.final_answer import FinalAnswerTool
 from Gradio_UI import GradioUI
@@ -17,162 +17,103 @@ BASE_URL = "https://api.cricketdata.org/v1"
 # Custom Tool: Analyze Cricketer Form
 @tool
 def analyze_cricketer_form(player_name: str, role: str) -> str:
-    """A tool to analyze a cricketer's recent form based on their role and name.
-    
+    """Analyzes the recent form of a cricketer based on their role.
     Args:
-        player_name: The name of the cricketer (e.g., 'Virat Kohli').
-        role: The player's role ('batter', 'bowler', 'wicket_keeper', 'fielder').
-    
-    Returns:
-        A string summary of the player's career stats and recent form.
+        player_name: The name of the cricketer.
+        role: The role of the player (batter, bowler, wicketkeeper, fielder).
     """
-    valid_roles = ['batter', 'bowler', 'wicket_keeper', 'fielder']
-    if role.lower() not in valid_roles:
-        return f"Error: Invalid role '{role}'. Choose from {valid_roles}."
+    try:
+        # Fetch player ID from API
+        player_id = get_player_id(player_name)
+        if not player_id:
+            return f"Player '{player_name}' not found."
+        
+        # Fetch basic player details
+        player_details = get_player_details(player_id)
+        
+        # Scrape detailed stats
+        detailed_stats = scrape_cricketer_stats(player_name, role)
+        
+        # Generate form analysis
+        return format_analysis(player_name, role, player_details, detailed_stats)
+    
+    except Exception as e:
+        return f"Error analyzing cricketer's form: {str(e)}"
 
-    player_id = get_player_id(player_name)
-    if not player_id:
-        return f"Error: Player '{player_name}' not found."
 
-    career_stats = get_career_stats(player_id)
-    if not career_stats:
-        return f"Error: Could not fetch career stats for '{player_name}'."
+def get_player_id(player_name: str):
+    response = requests.get(f"{BASE_URL}/players?name={player_name}", headers={"Authorization": f"Bearer {API_KEY}"})
+    if response.status_code == 200:
+        data = response.json()
+        return data.get("id")
+    return None
 
-    recent_matches = get_recent_matches(player_id)
-    
-    response = f"Analysis of {player_name} ({role.capitalize()}):\n\n"
-    response += "Career Stats:\n" + format_career_stats(career_stats, role) + "\n\n"
-    response += "Recent Form (Last 5 Matches):\n" + analyze_recent_form(recent_matches, role, player_name)
-    
-    return response
 
-def get_player_id(player_name: str) -> Union[str, None]:
-    url = f"{BASE_URL}/players/search?query={player_name}&apikey={API_KEY}"
-    try:
-        response = requests.get(url)
-        response.raise_for_status()
-        data = response.json()
-        if data.get("data") and len(data["data"]) > 0:
-            return data["data"][0]["id"]
-        return None
-    except requests.RequestException as e:
-        print(f"Error fetching player ID: {e}")
-        return None
-
-def get_career_stats(player_id: str) -> Union[Dict, None]:
-    url = f"{BASE_URL}/players/{player_id}/stats?apikey={API_KEY}"
-    try:
-        response = requests.get(url)
-        response.raise_for_status()
+def get_player_details(player_id: str):
+    response = requests.get(f"{BASE_URL}/player/{player_id}", headers={"Authorization": f"Bearer {API_KEY}"})
+    if response.status_code == 200:
         return response.json()
-    except requests.RequestException as e:
-        print(f"Error fetching career stats: {e}")
-        return None
+    return {}
 
-def get_recent_matches(player_id: str) -> list:
-    url = f"{BASE_URL}/players/{player_id}/matches?apikey={API_KEY}"
+
+def scrape_cricketer_stats(player_name: str, role: str):
+    """Scrapes recent performance stats from an online cricket stats website."""
     try:
-        response = requests.get(url)
-        response.raise_for_status()
-        matches = response.json().get("data", [])
-        return matches[:5]  # Limit to last 5 matches
-    except requests.RequestException as e:
-        print(f"Error fetching recent matches: {e}")
-        return []
-
-def format_career_stats(stats: Dict, role: str) -> str:
-    if not stats or "data" not in stats:
-        return "No stats available."
-    
-    data = stats["data"]
-    output = ""
-    
-    if role == "batter":
-        batting = data.get("batting", {})
-        output += f"Matches: {batting.get('matches', 'N/A')}\n"
-        output += f"Runs: {batting.get('runs', 'N/A')}\n"
-        output += f"Average: {batting.get('average', 'N/A')}\n"
-        output += f"Strike Rate: {batting.get('strike_rate', 'N/A')}\n"
+        search_url = f"https://stats.espncricinfo.com/ci/engine/player/{player_name.replace(' ', '_')}.html"
+        headers = {"User-Agent": "Mozilla/5.0"}
+        response = requests.get(search_url, headers=headers)
+        
+        if response.status_code != 200:
+            return {"error": "Failed to fetch data from ESPN Cricinfo"}
+        
+        soup = BeautifulSoup(response.text, "html.parser")
+        
+        if role.lower() == "batter":
+            stats = {
+                "last_50_scores": extract_batting_scores(soup),
+                "batting_avg": extract_batting_average(soup),
+                "strike_rate": extract_strike_rate(soup),
+            }
+        elif role.lower() == "bowler":
+            stats = {
+                "last_50_wickets": extract_wickets(soup),
+                "bowling_economy": extract_bowling_economy(soup),
+                "bowling_speed": extract_bowling_speed(soup),
+            }
+        elif role.lower() == "wicketkeeper":
+            stats = {
+                "catches": extract_catches(soup),
+                "stumpings": extract_stumpings(soup),
+                "missed_stumpings": extract_missed_stumpings(soup),
+            }
+        elif role.lower() == "fielder":
+            stats = {
+                "catches": extract_fielder_catches(soup),
+                "direct_runouts": extract_direct_runouts(soup),
+            }
+        else:
+            stats = {"error": "Invalid role specified"}
+        
+        return stats
     
-    elif role == "bowler":
-        bowling = data.get("bowling", {})
-        output += f"Matches: {bowling.get('matches', 'N/A')}\n"
-        output += f"Wickets: {bowling.get('wickets', 'N/A')}\n"
-        output += f"Economy: {bowling.get('economy', 'N/A')}\n"
-        output += f"Best: {bowling.get('best', {}).get('figures', 'N/A')}\n"
-    
-    elif role == "wicket_keeper":
-        fielding = data.get("fielding", {})
-        output += f"Catches: {fielding.get('catches', 'N/A')}\n"
-        output += f"Stumpings: {fielding.get('stumpings', 'N/A')}\n"
-    
-    elif role == "fielder":
-        fielding = data.get("fielding", {})
-        output += f"Catches: {fielding.get('catches', 'N/A')}\n"
-        output += f"Run Outs: {fielding.get('run_outs', 'N/A')}\n"
-    
-    return output
+    except Exception as e:
+        return {"error": f"Error scraping data: {str(e)}"}
 
-def analyze_recent_form(matches: list, role: str, player_name: str) -> str:
-    if not matches:
-        return "No recent match data available."
-    
-    output = ""
-    if role == "batter":
-        total_runs = 0
-        games = len(matches)
-        for match in matches:
-            batting = match.get("batting", {})
-            runs = batting.get("runs", 0)
-            total_runs += runs
-            output += f"Match vs {match.get('opponent', 'Unknown')}: {runs} runs\n"
-        avg = total_runs / games if games > 0 else 0
-        output += f"Recent Average: {avg:.2f}\n"
-        output += f"Form Comment: {'Strong' if avg > 30 else 'Inconsistent'} batting form."
-    
-    elif role == "bowler":
-        total_wickets = 0
-        games = len(matches)
-        for match in matches:
-            bowling = match.get("bowling", {})
-            wickets = bowling.get("wickets", 0)
-            total_wickets += wickets
-            output += f"Match vs {match.get('opponent', 'Unknown')}: {wickets} wickets\n"
-        avg_wickets = total_wickets / games if games > 0 else 0
-        output += f"Recent Wickets per Game: {avg_wickets:.2f}\n"
-        output += f"Form Comment: {'Strong' if avg_wickets > 1 else 'Inconsistent'} bowling form."
-    
-    elif role == "wicket_keeper":
-        total_stumpings = sum(m.get("fielding", {}).get("stumpings", 0) for m in matches)
-        total_catches = sum(m.get("fielding", {}).get("catches", 0) for m in matches)
-        output += f"Stumpings: {total_stumpings}, Catches: {total_catches}\n"
-        output += f"Form Comment: {'Reliable' if total_stumpings + total_catches > 3 else 'Average'} keeping."
-    
-    elif role == "fielder":
-        total_catches = sum(m.get("fielding", {}).get("catches", 0) for m in matches)
-        total_run_outs = sum(m.get("fielding", {}).get("run_outs", 0) for m in matches)
-        output += f"Catches: {total_catches}, Run Outs: {total_run_outs}\n"
-        output += f"Form Comment: {'Sharp' if total_catches + total_run_outs > 3 else 'Average'} fielding."
-    
-    return output
 
-# Example tool from your template
-@tool
-def get_current_time_in_timezone(timezone: str) -> str:
-    """A tool that fetches the current local time in a specified timezone.
-    Args:
-        timezone: A string representing a valid timezone (e.g., 'America/New_York').
+def format_analysis(player_name, role, player_details, detailed_stats):
+    return f"""
+    Cricketer: {player_name}
+    Role: {role}
+    Country: {player_details.get('country', 'Unknown')}
+    Batting Style: {player_details.get('battingStyle', 'Unknown')}
+    Bowling Style: {player_details.get('bowlingStyle', 'Unknown')}
+    
+    Recent Performance:
+    {detailed_stats}
     """
-    try:
-        tz = pytz.timezone(timezone)
-        local_time = datetime.datetime.now(tz).strftime("%Y-%m-%d %H:%M:%S")
-        return f"The current local time in {timezone} is: {local_time}"
-    except Exception as e:
-        return f"Error fetching time for timezone '{timezone}': {str(e)}"
 
 # Initialize tools
 final_answer = FinalAnswerTool()
-image_generation_tool = load_tool("agents-course/text-to-image", trust_remote_code=True)
 
 # Model setup
 model = HfApiModel(
@@ -189,7 +130,7 @@ with open("prompts.yaml", 'r') as stream:
 # Initialize the agent
 agent = CodeAgent(
     model=model,
-    tools=[final_answer, analyze_cricketer_form, get_current_time_in_timezone, image_generation_tool],
+    tools=[final_answer, analyze_cricketer_form],
     max_steps=6,
     verbosity_level=1,
     grammar=None,