如何在Streamlit中为会话状态设置嵌套字典结构的值？

你当前的input_number函数只能将值存入扁平结构的st.session_state，但需要构建嵌套字典来实现分组汇总（如asset1小计、asset总计）。以下是具体的修改方案：

修改后的input_number函数 #

import streamlit as st

def input_number(labelnya, session_path, value=None, help=None, format='%g'):
    valuenya = st.number_input(labelnya, value=value, help=help, format=format)
    valuenya = 0 if valuenya is None else valuenya
    
    # 解析嵌套路径，将"asset.asset1.cash_on_hand"拆分为键列表
    keys = session_path.split('.')
    current_dict = st.session_state
    
    # 逐层创建嵌套字典（若不存在则初始化）
    for key in keys[:-1]:
        if key not in current_dict:
            current_dict[key] = {}
        current_dict = current_dict[key]
    
    # 给最终层级的键赋值
    current_dict[keys[-1]] = valuenya
    
    print("All Session:\n{}".format(st.session_state))
    return valuenya

函数调用示例 #

直接传入用.分隔的嵌套路径作为session_path参数，即可自动生成你需要的嵌套结构：

# 给asset.asset1.cash_on_hand赋值
input_number("Cash on Hand", session_path='asset.asset1.cash_on_hand', value=1)
# 给asset.asset1.account_receivable赋值
input_number("Account Receivable", session_path='asset.asset1.account_receivable', value=2)
# 给liabilities下的字段赋值
input_number("Short-term Loan", session_path='liabilities.liabilities1.short_term_loan', value=5)

执行后，st.session_state会生成目标嵌套结构：

{
    "asset": {
        "asset1": {
            "cash_on_hand": 1,
            "account_receivable": 2
        },
        "asset2": {}
    },
    "liabilities": {
        "liabilities1": {
            "short_term_loan": 5
        },
        "liabilities2": {}
    }
}

分组汇总实现 #

基于嵌套结构可以轻松完成分组计算：

# 计算asset1的总计
asset1_total = sum(st.session_state['asset']['asset1'].values())
st.write(f"Asset1 Total: {asset1_total}")

# 计算所有资产的总计
asset_total = sum(sum(group.values()) for group in st.session_state['asset'].values())
st.write(f"Total Assets: {asset_total}")

# 计算所有负债的总计
liabilities_total = sum(sum(group.values()) for group in st.session_state['liabilities'].values())
st.write(f"Total Liabilities: {liabilities_total}")

内容的提问来源于stack exchange，提问作者Cignitor