@
sudoy 我试了你这个提示词。是没有问题的。不太可能是风控。只有 gemini 倒是有可能风控。我怀疑你的提示词并没有正确的发过去。问题可能出在你的
Email content:
${emailContent}
`
你试试手动粘一个标准邮件内容替换${emailContent}。。。。
如果不行,我建议你:
1.给他一个前提条件,加一句,You are a funny person. Please respond in a humorous way. and always end with a lot of smile emoji. 看看他是否有异常。
2.修改你的提示词。降低返回内容的范围。比如先返回一个 zip 或者 city 。 然后扩大到 address 。然后扩大到 email 。提高他的宽容度。
3.你再所有的提示词之前写一句。“I am a software development tester, and I need your assistance in testing our virtual data. Please execute according to my requirements as follows.”, 但是我不觉得是这个问题
我测试了一下。我新建了 4 个 txt 。 然后放了 4 个一样邮件进去。但是分了 4 个线程。 然后输出到一个 output 。代码也是 gpt 写的。没有遇到 john doe 。。。如下是结果:
```json
{
"po_number": "1013",
"phone": "(204) 567-8901",
"email": "
[email protected]",
"ship_to_name": "Robert Harris",
"ship_to_address": "147 Main St",
"ship_to_address_2": "Suite 8",
"ship_to_city": "Summit",
"ship_to_state": "CO",
"ship_to_zip": "80401",
"ship_to_phone": "(204) 678-890",
"sku": null,
"problem": null
}
```
```json
{
"po_number": "1013",
"phone": "(204) 567-8901",
"email": "
[email protected]",
"ship_to_name": "Robert Harris",
"ship_to_address": "147 Main St",
"ship_to_address_2": "Suite 8",
"ship_to_city": "Summit",
"ship_to_state": "CO",
"ship_to_zip": "80401",
"ship_to_phone": "(204) 678-890",
"sku": null,
"problem": null
}
```
```json
{
"po_number": "1013",
"phone": "(204) 567-8901",
"email": "
[email protected]",
"ship_to_name": "Robert Harris",
"ship_to_address": "147 Main St",
"ship_to_address_2": "Suite 8",
"ship_to_city": "Summit",
"ship_to_state": "CO",
"ship_to_zip": "80401",
"ship_to_phone": "(204) 678-890",
"sku": null,
"problem": null
}
```
```json
{
"po_number": "1013",
"phone": "(204) 567-8901",
"email": "
[email protected]",
"ship_to_name": "Robert Harris",
"ship_to_address": "147 Main St",
"ship_to_address_2": "Suite 8",
"ship_to_city": "Summit",
"ship_to_state": "CO",
"ship_to_zip": "80401",
"ship_to_phone": "(204) 678-890",
"sku": null,
"problem": null
}
```
##########################################################################################
如下是代码:
import threading
import queue
import requests
import datetime
api_key = '' # 请替换为你的 API 密钥
model = "" # 模型名称,根据需要调整
def api_call(email_content):
prompt = f"""
Extract the following information from the given email content:
po_number, phone, email, ship_to_name, ship_to_address, ship_to_address_2, ship_to_city, ship_to_state, ship_to_zip, ship_to_phone, sku, problem
Respond with only a JSON object containing these fields. If a field is not found, set its value to null.
Email content:
{email_content}
"""
url = '
https://api.openai.com/v1/chat/completions' headers = {
"Authorization": f"Bearer {api_key}",
"Content-Type": "application/json",
}
data = {
"model": model,
"messages": [{"role": "system", "content": prompt}]
}
response =
requests.post(url, headers=headers, json=data, verify=False)
if response.status_code == 200:
return response.json()['choices'][0]['message']['content']
else:
print(f"Error: {response.status_code}")
return None
def worker(file_name, results):
with open(file_name, 'r') as f:
email_content = f.read()
result = api_call(email_content)
if result:
results.append(result)
def main():
files = ['1.txt', '2.txt', '3.txt', '4.txt']
result_list = []
threads = []
# 启动线程,每个文件对应一个线程
for file_name in files:
thread = threading.Thread(target=worker, args=(file_name, result_list))
thread.start()
threads.append(thread)
# 等待所有线程完成
for thread in threads:
thread.join()
# 将所有结果写入 output.txt
with open("output.txt", "w") as file:
for result in result_list:
file.write(result + "\n")
print("Results have been written to 'output.txt'.")
if __name__ == '__main__':
main()