excel_tool.py

import pandas as pd
import requests
from io import BytesIO
import numpy as np

def analyze_excel(path_or_url: str) -> str:
    """Analysera Excel-fil och returnera detaljerad information"""
    try:
        if path_or_url.startswith("http"):
            response = requests.get(path_or_url, timeout=10)
            response.raise_for_status()
            df = pd.read_excel(BytesIO(response.content))
        else:
            df = pd.read_excel(path_or_url)
    
    except Exception as e:
        return f"Error reading Excel file: {e}"
    
    # Detaljerad analys
    analysis = []
    analysis.append(f"Rows: {df.shape[0]}, Columns: {df.shape[1]}")
    analysis.append(f"Column names: {', '.join(df.columns)}")
    
    # Datatyper
    numeric_cols = df.select_dtypes(include=[np.number]).columns.tolist()
    if numeric_cols:
        analysis.append(f"Numeric columns: {', '.join(numeric_cols)}")
        for col in numeric_cols:
            analysis.append(f"{col} - Sum: {df[col].sum():.2f}, Mean: {df[col].mean():.2f}, Max: {df[col].max():.2f}, Min: {df[col].min():.2f}")
    
    text_cols = df.select_dtypes(include=['object']).columns.tolist()
    if text_cols:
        analysis.append(f"Text columns: {', '.join(text_cols)}")
        for col in text_cols:
            unique_values = df[col].nunique()
            analysis.append(f"{col} - Unique values: {unique_values}")
            if unique_values <= 10:
                analysis.append(f"{col} values: {', '.join(map(str, df[col].unique()))}")
    
    # Första 3 raderna som exempel
    analysis.append("First 3 rows:")
    analysis.append(df.head(3).to_string())
    
    return "\n".join(analysis)