File size: 4,517 Bytes
55fdf91
 
87f4646
55fdf91
16d9c95
87f4646
 
cea8bca
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
55fdf91
 
 
16d9c95
cea8bca
 
16d9c95
55fdf91
 
 
cea8bca
16d9c95
cea8bca
16d9c95
 
cea8bca
 
 
 
 
 
 
55fdf91
 
16d9c95
55fdf91
16d9c95
55fdf91
 
 
cea8bca
 
 
 
 
55fdf91
 
cea8bca
55fdf91
 
 
 
 
8acdf8b
55fdf91
 
 
8acdf8b
55fdf91
 
 
 
 
 
 
 
 
 
87f4646
 
 
 
cea8bca
 
 
 
 
 
 
 
 
87f4646
 
 
 
cea8bca
87f4646
 
 
 
 
cea8bca
87f4646
 
 
 
 
 
 
 
 
 
 
cea8bca
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
import json
import os
import re
import uuid
import streamlit as st
import pandas as pd
from custom import *
import copy
import io


def get_history_chats(path: str) -> list:
    if "apikey" in st.secrets:
        if not os.path.exists(path):
            os.makedirs(path)
        files = [f for f in os.listdir(f'./{path}') if f.endswith('.json')]
        files_with_time = [(f, os.stat(f'./{path}/' + f).st_ctime) for f in files]
        sorted_files = sorted(files_with_time, key=lambda x: x[1], reverse=True)
        chat_names = [os.path.splitext(f[0])[0] for f in sorted_files]
        if len(chat_names) == 0:
            chat_names.append('New Chat_' + str(uuid.uuid4()))
    else:
        chat_names = ['New Chat_' + str(uuid.uuid4())]
    return chat_names


def save_data(path: str, file_name: str, history: list, paras: dict, contexts: dict, **kwargs):
    if not os.path.exists(path):
        os.makedirs(path)
    with open(f"./{path}/{file_name}.json", 'w', encoding='utf-8') as f:
        json.dump({"history": history, "paras": paras, "contexts": contexts, **kwargs}, f)


def remove_data(path: str, chat_name: str):
    try:
        os.remove(f"./{path}/{chat_name}.json")
    except FileNotFoundError:
        pass
    # 清除缓存
    try:
        st.session_state.pop('history' + chat_name)
        for item in ["context_select", "context_input", "context_level", *initial_content_all['paras']]:
            st.session_state.pop(item + chat_name + "value")
    except KeyError:
        pass


def load_data(path: str, file_name: str) -> dict:
    try:
        with open(f"./{path}/{file_name}.json", 'r', encoding='utf-8') as f:
            data = json.load(f)
            return data
    except FileNotFoundError:
        content = copy.deepcopy(initial_content_all)
        if "apikey" in st.secrets:
            with open(f"./{path}/{file_name}.json", 'w', encoding='utf-8') as f:
                f.write(json.dumps(content))
        return content


def show_each_message(message: str, role: str, area=None):
    if area is None:
        area = [st.markdown] * 2
    if role == 'user':
        icon = user_svg
        name = user_name
        background_color = user_background_color
    else:
        icon = gpt_svg
        name = gpt_name
        background_color = gpt_background_color
    area[0](f"\n<div class='avatar'>{icon}<h2>{name}:</h2></div>", unsafe_allow_html=True)
    area[1](f"""<div class='content-div' style='background-color: {background_color};'>\n\n{message}""",
            unsafe_allow_html=True)


def show_messages(messages: list):
    for each in messages:
        if (each["role"] == "user") or (each["role"] == "assistant"):
            show_each_message(each["content"], each["role"])
        if each["role"] == "assistant":
            st.write("---")


# 根据context_level提取history
def get_history_input(history: list, level: int) -> list:
    if level != 0:
        df_history = pd.DataFrame(history)
        df_system = df_history.query('role=="system"')
        df_input = df_history.query('role!="system"')
        df_input = df_input[-level * 2:]
        res = pd.concat([df_system, df_input], ignore_index=True).to_dict('records')
    else:
        res = []
    return res


# 去除#号右边的空格
def remove_hashtag_right__space(text: str) -> str:
    res = re.sub(r"(#+)\s*", r"\1", text)
    return res


# 提取文本
def extract_chars(text: str, num: int) -> str:
    char_num = 0
    chars = ''
    for char in text:
        # 汉字算两个字符
        if '\u4e00' <= char <= '\u9fff':
            char_num += 2
        else:
            char_num += 1
        chars += char
        if char_num >= num:
            break
    return chars


def download_history(history: list):
    md_text = ""
    for msg in history:
        if msg['role'] == 'user':
            md_text += f'## {user_name}:\n{msg["content"]}\n'
        elif msg['role'] == 'assistant':
            md_text += f'## {gpt_name}:\n{msg["content"]}\n'
    output = io.BytesIO()
    output.write(md_text.encode('utf-8'))
    output.seek(0)
    return output


def filename_correction(filename: str) -> str:
    pattern = r'[^\w\.-]'
    filename = re.sub(pattern, '', filename)
    return filename


def url_correction(text: str) -> str:
    pattern = r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+#]|[!*\(\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+'
    links = re.findall(pattern, text)
    for link in links:
        text = text.replace(link, " " + link + " ")
    return text