用ChatGPT生成测试数据

原创

Crossin先生

发布于 2023-06-11 10:56:06

3290

发布于 2023-06-11 10:56:06

文章被收录于专栏：Crossin的编程教室Crossin的编程教室

大家好，欢迎来到 Crossin的编程教室！

在之前的文章用ChatGPT写一个数据采集程序?中，我们演示了如何用 ChatGPT 辅助编写代码。

除了直接让ChatGPT写代码，我们也可以让它生成一些开发中使用的测试数据。

比如在开发和测试时，常会用到 JSON?格式的数据：

[
  {
    "id": 1,
    "name": "John Smith",
    "age": 28,
    "gender": "male",
    "email": "john.smith@example.com",
    "address": "123 Main Street, City",
    "phone": "123-456-7890"
  },
  {
    "id": 2,
    "name": "Jane Doe",
    "age": 32,
    "gender": "female",
    "email": "jane.doe@example.com",
    "address": "456 Oak Avenue, Town",
    "phone": "987-654-3210"
  },
  {
    "id": 3,
    "name": "David Johnson",
    "age": 45,
    "gender": "male",
    "email": "david.johnson@example.com",
    "address": "789 Elm Street, Village",
    "phone": "555-123-4567"
  },
  {
    "id": 4,
    "name": "Emily Williams",
    "age": 21,
    "gender": "female",
    "email": "emily.williams@example.com",
    "address": "321 Pine Avenue, Town",
    "phone": "777-888-9999"
  }
]

如果对数据内容或结构不满意，可以让ChatGPT做进一步修改，比如增加一个存在重复数据的字段：

[
  {
    "id": 1,
    "name": "John Smith",
    "age": 28,
    "gender": "male",
    "email": "john.smith@example.com",
    "address": "123 Main Street, City",
    "phone": "123-456-7890",
    "occupation": "engineer"
  },
  {
    "id": 2,
    "name": "Jane Doe",
    "age": 32,
    "gender": "female",
    "email": "jane.doe@example.com",
    "address": "456 Oak Avenue, Town",
    "phone": "987-654-3210",
    "occupation": "doctor"
  },
  {
    "id": 3,
    "name": "David Johnson",
    "age": 45,
    "gender": "male",
    "email": "david.johnson@example.com",
    "address": "789 Elm Street, Village",
    "phone": "555-123-4567",
    "occupation": "teacher"
  },
  {
    "id": 4,
    "name": "Emily Williams",
    "age": 21,
    "gender": "female",
    "email": "emily.williams@example.com",
    "address": "321 Pine Avenue, Town",
    "phone": "777-888-9999",
    "occupation": "engineer"
  }
]

数据的格式也可以很方便地调整，比如从JSON换成XML：

<users>
  <user>
    <id>1</id>
    <name>John Smith</name>
    <age>28</age>
    <gender>male</gender>
    <email>john.smith@example.com</email>
    <address>123 Main Street, City</address>
    <phone>123-456-7890</phone>
    <occupation>engineer</occupation>
  </user>
  <user>
    <id>2</id>
    <name>Jane Doe</name>
    <age>32</age>
    <gender>female</gender>
    <email>jane.doe@example.com</email>
    <address>456 Oak Avenue, Town</address>
    <phone>987-654-3210</phone>
    <occupation>doctor</occupation>
  </user>
  <user>
    <id>3</id>
    <name>David Johnson</name>
    <age>45</age>
    <gender>male</gender>
    <email>david.johnson@example.com</email>
    <address>789 Elm Street, Village</address>
    <phone>555-123-4567</phone>
    <occupation>teacher</occupation>
  </user>
  <user>
    <id>4</id>
    <name>Emily Williams</name>
    <age>21</age>
    <gender>female</gender>
    <email>emily.williams@example.com</email>
    <address>321 Pine Avenue, Town</address>
    <phone>777-888-9999</phone>
    <occupation>engineer</occupation>
  </user>
</users>

有同学要说了，可以用?faker 模块来生成哎！

是的，但那样毕竟也要自己动手写代码，还可能时不时要调整。现在有ChatGPT的辅助，可以秒出数据，岂不是更轻松？

当然你也可以将二者结合，比如这样：

import random
from faker import Faker
import json

faker = Faker()

def generate_user_data(num_users):
    users = []
    for _ in range(num_users):
        user = {
            "id": faker.random_number(digits=4),
            "name": faker.name(),
            "age": random.randint(18, 60),
            "gender": random.choice(["male", "female"]),
            "email": faker.email(),
            "address": faker.address(),
            "phone": faker.phone_number()
        }
        users.append(user)
    return users

num_users = 4
user_data = generate_user_data(num_users)

json_data = json.dumps(user_data, indent=4)
print(json_data)

以往在开发中，如果需要类似的测试数据，手动编写是非常耗时和低效的。现在，类似的很多编程辅助工作都可交由ChatGPT来处理，从而让开发者把时间花在更重要的事情之上。

以上内容节选自Crossin的新书《码上行动：零基础学会PYTHON编程（CHATGPT版）》。

原创声明：本文系作者授权腾讯云开发者社区发表，未经许可，不得转载。

如有侵权，请联系 cloudcommunity@tencent.com 删除。

python