Python嵌套JSON展开打平教程：扁平化嵌套字段

把深层嵌套的JSON数据用json_normalize或递归函数打平成扁平表，点号路径作为列名，方便转CSV和做分析，附完整代码。

场景痛点

接口返回的JSON一层套一层：user.profile.address.city、order.items[0].sku，直接转CSV只能把嵌套字段塞成一坨字符串。想把 city、sku 单独拉成列做筛选统计，就得先把嵌套结构"打平"。这篇用pandas的 json_normalize 和递归两种方式解决。

用到的库

pip install pandas

完整代码

# 嵌套JSON展开打平
import json
import pandas as pd

def flatten_with_json_normalize(json_path):
"""
用pandas自带的json_normalize，自动把嵌套dict展开成点号列名。
record_path 指定数组路径，meta 指定要带上的父级字段。
"""
with open(json_path, "r", encoding="utf-8") as f:
data = json.load(f)

# 情况1：根就是list[dict]，直接打平
if isinstance(data, list):
df = pd.json_normalize(data)
else:
# 情况2：根是dict，里面有个数组需要展开
# 比如 {"orders": [{"user": {...}, "items": [...]}]}
# record_path 指向要展开的数组路径
df = pd.json_normalize(
data,
record_path="orders",          # 展开 orders 数组
meta=["user_id", "shop_id"],   # 把外层字段也带上
errors="ignore",
)

print("打平后列名：")
for c in df.columns:
print("  ", c)
print(f"\n形状：{df.shape}")
return df

def flatten_recursive(nested_dict, parent_key="", sep="."):
"""
递归手动打平，适定制灵活。
把 {"a": {"b": 1}} 变成 {"a.b": 1}
"""
items = {}
for k, v in nested_dict.items():
new_key = f"{parent_key}{sep}{k}" if parent_key else k
if isinstance(v, dict):
items.update(flatten_recursive(v, new_key, sep=sep))
elif isinstance(v, list):
# 列表类型：转成JSON字符串存着，或展开成多列
items[new_key] = json.dumps(v, ensure_ascii=False)
else:
items[new_key] = v
return items

def flatten_list_records(records):
"""对list[dict]批量递归打平。"""
flat = [flatten_recursive(r) for r in records]
return pd.DataFrame(flat)

def main():
json_file = "nested.json"

print("=== json_normalize 方式 ===")
df1 = flatten_with_json_normalize(json_file)
df1.to_csv("flatten_normalize.csv", index=False, encoding="utf-8-sig")

print("\n=== 递归方式（示例数据）===")
sample = [
{
"user": {"id": 1, "name": "张三", "address": {"city": "北京", "zip": "100000"}},
"order": {"id": 100, "amount": 299.5, "items": [{"sku": "A1"}, {"sku": "A2"}]},
},
{
"user": {"id": 2, "name": "李四", "address": {"city": "上海"}},
"order": {"id": 101, "amount": 599.0},
},
]
df2 = flatten_list_records(sample)
print(df2.T)
df2.to_csv("flatten_recursive.csv", index=False, encoding="utf-8-sig")

if __name__ == "__main__":
main()