Compare commits
No commits in common. "24db487104ef0ef9c944eb18769d7c10a80574c0" and "567c0dbf87aa10c3d49fcf9d431720d035baea35" have entirely different histories.
24db487104
...
567c0dbf87
@ -1,7 +1,7 @@
|
||||
[metadata]
|
||||
# replace with your username:
|
||||
name = guan
|
||||
version = 0.1.138
|
||||
version = 0.1.129
|
||||
author = guanjihuan
|
||||
author_email = guanjihuan@163.com
|
||||
description = An open source python package
|
||||
|
||||
@ -1,6 +1,6 @@
|
||||
Metadata-Version: 2.1
|
||||
Name: guan
|
||||
Version: 0.1.138
|
||||
Version: 0.1.129
|
||||
Summary: An open source python package
|
||||
Home-page: https://py.guanjihuan.com
|
||||
Author: guanjihuan
|
||||
|
||||
@ -1,49 +1,25 @@
|
||||
# Module: data_processing
|
||||
|
||||
# AI 对话
|
||||
def chat(prompt='你好', stream=1, model=1, top_p=0.8, temperature=0.85):
|
||||
# 模型对话
|
||||
def chat(prompt='你好', model=1, stream=0, top_p=0.8, temperature=0.85):
|
||||
import socket
|
||||
import json
|
||||
import time
|
||||
import guan
|
||||
with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as client_socket:
|
||||
client_socket.settimeout(30)
|
||||
client_socket.connect(('socket.guanjihuan.com', 12345))
|
||||
split_text_list = guan.split_text(prompt, width=100)
|
||||
message_times = len(split_text_list)
|
||||
if message_times == 1 or message_times == 0:
|
||||
message = {
|
||||
'server': "chat.guanjihuan.com",
|
||||
'prompt': prompt,
|
||||
'model': model,
|
||||
'top_p': top_p,
|
||||
'temperature': temperature,
|
||||
}
|
||||
send_message = json.dumps(message)
|
||||
client_socket.send(send_message.encode('utf-8'))
|
||||
else:
|
||||
end_message = 0
|
||||
for i0 in range(message_times):
|
||||
if i0 == message_times-1:
|
||||
end_message = 1
|
||||
prompt_0 = split_text_list[i0]
|
||||
message = {
|
||||
'server': "chat.guanjihuan.com",
|
||||
'prompt': prompt_0,
|
||||
'model': model,
|
||||
'top_p': top_p,
|
||||
'temperature': temperature,
|
||||
'end_message': end_message,
|
||||
}
|
||||
send_message = json.dumps(message)
|
||||
client_socket.send(send_message.encode('utf-8'))
|
||||
time.sleep(0.2)
|
||||
message = {
|
||||
'server': "chat.guanjihuan.com",
|
||||
'prompt': prompt,
|
||||
'model': model,
|
||||
'top_p': top_p,
|
||||
'temperature': temperature,
|
||||
}
|
||||
send_message = json.dumps(message)
|
||||
client_socket.send(send_message.encode('utf-8'))
|
||||
if stream == 1:
|
||||
print('\n--- Begin Chat Stream Message ---\n')
|
||||
print('\n--- Begin Stream Message ---\n')
|
||||
response = ''
|
||||
while True:
|
||||
if prompt == '':
|
||||
break
|
||||
try:
|
||||
data = client_socket.recv(1024)
|
||||
if data != b'':
|
||||
@ -56,109 +32,48 @@ def chat(prompt='你好', stream=1, model=1, top_p=0.8, temperature=0.85):
|
||||
break
|
||||
else:
|
||||
if stream == 1:
|
||||
print(stream_response, end='', flush=True)
|
||||
print(stream_response)
|
||||
except:
|
||||
break
|
||||
client_socket.close()
|
||||
if stream == 1:
|
||||
print('\n\n--- End Chat Stream Message ---\n')
|
||||
print('\n--- End Stream Message ---\n')
|
||||
return response
|
||||
|
||||
# 加上函数代码的 AI 对话
|
||||
def chat_with_function_code(function_name, prompt='', stream=1, model=1, top_p=0.8, temperature=0.85):
|
||||
import guan
|
||||
function_source = guan.get_source(function_name)
|
||||
if prompt == '':
|
||||
response = guan.chat(prompt=function_source, stream=stream, model=model, top_p=top_p, temperature=temperature)
|
||||
else:
|
||||
response = guan.chat(prompt=function_source+'\n\n'+prompt, stream=stream, model=model, top_p=top_p, temperature=temperature)
|
||||
return response
|
||||
|
||||
# 机器人自动对话
|
||||
def auto_chat(prompt='你好', round=2):
|
||||
import guan
|
||||
response0 = prompt
|
||||
for i0 in range(round):
|
||||
print(f'【对话第 {i0+1} 轮】\n')
|
||||
print('机器人 1: ')
|
||||
response1 = guan.chat(prompt=response0, stream=1)
|
||||
print('机器人 2: ')
|
||||
response0 = guan.chat(prompt=response1, stream=1)
|
||||
|
||||
# 机器人自动对话（引导对话）
|
||||
def auto_chat_with_guide(prompt='你好', guide_message='（回答字数少于30个字，最后反问我一个问题）', round=5):
|
||||
import guan
|
||||
response0 = prompt
|
||||
for i0 in range(round):
|
||||
print(f'【对话第 {i0+1} 轮】\n')
|
||||
print('机器人 1: ')
|
||||
response1 = guan.chat(prompt=response0+guide_message, stream=1)
|
||||
print('机器人 2: ')
|
||||
response0 = guan.chat(prompt=response1+guide_message, stream=1)
|
||||
|
||||
# 在云端服务器上运行函数（需要函数是独立可运行的代码）
|
||||
def run(function_name, *args, **kwargs):
|
||||
import socket
|
||||
import json
|
||||
import pickle
|
||||
import base64
|
||||
import time
|
||||
import guan
|
||||
with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as client_socket:
|
||||
client_socket.connect(('run.guanjihuan.com', 12345))
|
||||
client_socket.connect(('socket.guanjihuan.com', 12345))
|
||||
function_source = guan.get_source(function_name)
|
||||
split_text_list = guan.split_text(function_source, width=100)
|
||||
message_times = len(split_text_list)
|
||||
if message_times == 1 or message_times == 0:
|
||||
message = {
|
||||
'server': "run.guanjihuan.com",
|
||||
'function_name': function_name.__name__,
|
||||
'function_source': function_source,
|
||||
'args': str(args),
|
||||
'kwargs': str(kwargs)
|
||||
}
|
||||
send_message = json.dumps(message)
|
||||
client_socket.send(send_message.encode())
|
||||
else:
|
||||
end_message = 0
|
||||
for i0 in range(message_times):
|
||||
if i0 == message_times-1:
|
||||
end_message = 1
|
||||
source_0 = split_text_list[i0]
|
||||
message = {
|
||||
'server': "run",
|
||||
'function_name': function_name.__name__,
|
||||
'function_source': source_0,
|
||||
'args': str(args),
|
||||
'kwargs': str(kwargs),
|
||||
'end_message': end_message,
|
||||
}
|
||||
send_message = json.dumps(message)
|
||||
client_socket.send(send_message.encode())
|
||||
time.sleep(0.2)
|
||||
print('\nguan.run: 云端服务器正在计算，请等待返回结果。\n')
|
||||
return_data = ''
|
||||
print_data = ''
|
||||
message = {
|
||||
'server': "run",
|
||||
'function_name': function_name.__name__,
|
||||
'function_source': function_source,
|
||||
'args': str(args),
|
||||
'kwargs': str(kwargs)
|
||||
}
|
||||
send_message = json.dumps(message)
|
||||
client_socket.send(send_message.encode())
|
||||
return_data = None
|
||||
while True:
|
||||
try:
|
||||
data = client_socket.recv(1024)
|
||||
return_text = data.decode()
|
||||
return_dict = json.loads(return_text)
|
||||
return_data += return_dict['return_data']
|
||||
print_data += return_dict['print_data']
|
||||
return_data = return_dict['return_data']
|
||||
print_data = return_dict['print_data']
|
||||
end_message = return_dict['end_message']
|
||||
if end_message == 1:
|
||||
if print_data != '':
|
||||
print('--- Start Print ---\n')
|
||||
print(print_data)
|
||||
print('--- End Print ---\n')
|
||||
if end_message == 1 or return_text == '':
|
||||
break
|
||||
except:
|
||||
break
|
||||
if print_data != '':
|
||||
print('--- Start Print ---\n')
|
||||
print(print_data)
|
||||
print('--- End Print ---\n')
|
||||
print('guan.run: 云端服务器计算结束，以上是打印结果。\n')
|
||||
else:
|
||||
print('guan.run: 云端服务器计算结束。\n')
|
||||
return_data = pickle.loads(base64.b64decode(return_data))
|
||||
client_socket.close()
|
||||
return return_data
|
||||
|
||||
@ -388,23 +303,18 @@ def print_array_with_index(array, show_index=1, index_type=0):
|
||||
index += 1
|
||||
print(index, i0)
|
||||
|
||||
# 根据一定的字符长度来分割文本
|
||||
def split_text(text, width=100):
|
||||
split_text_list = [text[i:i+width] for i in range(0, len(text), width)]
|
||||
return split_text_list
|
||||
|
||||
# 使用textwrap根据一定的字符长度来分割文本（会自动微小调节宽度，但存在换行符和空格丢失的问题）
|
||||
def split_text_with_textwrap(text, width=100):
|
||||
import textwrap
|
||||
split_text_list = textwrap.wrap(text, width)
|
||||
return split_text_list
|
||||
|
||||
# 使用jieba软件包进行分词
|
||||
def divide_text_into_words(text):
|
||||
import jieba
|
||||
words = jieba.lcut(text)
|
||||
return words
|
||||
|
||||
# 根据一定的字符长度来分割文本
|
||||
def split_text(text, wrap_width=3000):
|
||||
import textwrap
|
||||
split_text_list = textwrap.wrap(text, wrap_width)
|
||||
return split_text_list
|
||||
|
||||
# 判断某个字符是中文还是英文或其他
|
||||
def check_Chinese_or_English(a):
|
||||
if '\u4e00' <= a <= '\u9fff' :
|
||||
@ -1287,28 +1197,25 @@ def convert_wordpress_xml_to_markdown(xml_file='./a.xml', convert_content=1, rep
|
||||
title = item.find('title').text
|
||||
content = item.find('.//content:encoded', namespaces={'content': 'http://purl.org/rss/1.0/modules/content/'}).text
|
||||
if convert_content == 1:
|
||||
try:
|
||||
content = re.sub(r'<!--.*?-->', '', content)
|
||||
content = content.replace('<p>', '')
|
||||
content = content.replace('</p>', '')
|
||||
content = content.replace('<ol>', '')
|
||||
content = content.replace('</ol>', '')
|
||||
content = content.replace('<ul>', '')
|
||||
content = content.replace('</ul>', '')
|
||||
content = content.replace('<strong>', '')
|
||||
content = content.replace('</strong>', '')
|
||||
content = content.replace('</li>', '')
|
||||
content = content.replace('<li>', '+ ')
|
||||
content = content.replace('</h3>', '')
|
||||
content = re.sub(r'<h2.*?>', '## ', content)
|
||||
content = re.sub(r'<h3.*?>', '### ', content)
|
||||
content = re.sub(r'<h4.*?>', '#### ', content)
|
||||
for replace_item in replace_more:
|
||||
content = content.replace(replace_item, '')
|
||||
for _ in range(100):
|
||||
content = content.replace('\n\n\n', '\n\n')
|
||||
except:
|
||||
print(f'提示：字符串替换出现问题！出现问题的内容为：{content}')
|
||||
content = re.sub(r'<!--.*?-->', '', content)
|
||||
content = content.replace('<p>', '')
|
||||
content = content.replace('</p>', '')
|
||||
content = content.replace('<ol>', '')
|
||||
content = content.replace('</ol>', '')
|
||||
content = content.replace('<ul>', '')
|
||||
content = content.replace('</ul>', '')
|
||||
content = content.replace('<strong>', '')
|
||||
content = content.replace('</strong>', '')
|
||||
content = content.replace('</li>', '')
|
||||
content = content.replace('<li>', '+ ')
|
||||
content = content.replace('</h3>', '')
|
||||
content = re.sub(r'<h2.*?>', '## ', content)
|
||||
content = re.sub(r'<h3.*?>', '### ', content)
|
||||
content = re.sub(r'<h4.*?>', '#### ', content)
|
||||
for replace_item in replace_more:
|
||||
content = content.replace(replace_item, '')
|
||||
for _ in range(100):
|
||||
content = content.replace('\n\n\n', '\n\n')
|
||||
else:
|
||||
pass
|
||||
markdown_content = f"# {title}\n{content}"
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user