import pymysql
import re
from collections import Counter

def get_db_connection():
    return pymysql.connect(
        host='localhost',
        user='root',
        password='45278434',
        database='pm_proto_test',
        charset='utf8mb4',
        cursorclass=pymysql.cursors.DictCursor
    )

def analyze_logs():
    conn = get_db_connection()
    try:
        with conn.cursor() as cursor:
            cursor.execute("SELECT DISTINCT recent_log FROM projects_history WHERE recent_log IS NOT NULL AND recent_log != ''")
            rows = cursor.fetchall()
            
            logs = [r['recent_log'] for r in rows]
            
            output = []
            output.append("[Raw Log Samples]")
            for log in logs[:20]:
                output.append(f"- {log}")

            patterns = []
            for log in logs:
                p = re.sub(r'\d{2,4}\.\d{2}\.\d{2}', '[DATE]', log)
                p = re.sub(r'\d+', '[NUM]', p)
                patterns.append(p)
            
            output.append("\n[Log Patterns Frequency]")
            pattern_counts = Counter(patterns).most_common(20)
            for p, count in pattern_counts:
                output.append(f"({count}) {p}")

            with open("log_analysis_result.txt", "w", encoding="utf-8") as f:
                f.write("\n".join(output))
            print("Analysis complete. Result saved to log_analysis_result.txt")

    finally:
        conn.close()

if __name__ == "__main__":
    analyze_logs()