GitHub 94adb4f

app.py

@@ -1,18 +1,24 @@
 # -*- coding:utf-8 -*-
-import gradio as gr
 import os
 import logging
 import sys
-import argparse
+
+import gradio as gr
+
 from utils import *
 from presets import *
+from overwrites import *
+from chat_func import *
 
-logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] [%(filename)s:%(lineno)d] %(message)s")
+logging.basicConfig(
+    level=logging.DEBUG,
+    format="%(asctime)s [%(levelname)s] [%(filename)s:%(lineno)d] %(message)s",
+)
 
-my_api_key = ""    # 在这里输入你的 API 密钥
+my_api_key = ""  # 在这里输入你的 API 密钥
 
-#if we are running in Docker
-if os.environ.get('dockerrun') == 'yes':
+# if we are running in Docker
+if os.environ.get("dockerrun") == "yes":
     dockerflag = True
 else:
     dockerflag = False
@@ -20,17 +26,21 @@ else:
 authflag = False
 
 if dockerflag:
-    my_api_key = os.environ.get('my_api_key')
+    my_api_key = os.environ.get("my_api_key")
     if my_api_key == "empty":
         logging.error("Please give a api key!")
         sys.exit(1)
-    #auth
-    username = os.environ.get('USERNAME')
-    password = os.environ.get('PASSWORD')
+    # auth
+    username = os.environ.get("USERNAME")
+    password = os.environ.get("PASSWORD")
     if not (isinstance(username, type(None)) or isinstance(password, type(None))):
         authflag = True
 else:
-    if not my_api_key and os.path.exists("api_key.txt") and os.path.getsize("api_key.txt"):
+    if (
+        not my_api_key
+        and os.path.exists("api_key.txt")
+        and os.path.getsize("api_key.txt")
+    ):
         with open("api_key.txt", "r") as f:
             my_api_key = f.read().strip()
     if os.path.exists("auth.json"):
@@ -42,11 +52,78 @@ else:
                 authflag = True
 
 gr.Chatbot.postprocess = postprocess
+PromptHelper.compact_text_chunks = compact_text_chunks
+
+with open("custom.css", "r", encoding="utf-8") as f:
+    customCSS = f.read()
 
-with gr.Blocks(css=customCSS) as demo:
+with gr.Blocks(
+    css=customCSS,
+    theme=gr.themes.Soft(
+        primary_hue=gr.themes.Color(
+            c50="#02C160",
+            c100="rgba(2, 193, 96, 0.2)",
+            c200="#02C160",
+            c300="rgba(2, 193, 96, 0.32)",
+            c400="rgba(2, 193, 96, 0.32)",
+            c500="rgba(2, 193, 96, 1.0)",
+            c600="rgba(2, 193, 96, 1.0)",
+            c700="rgba(2, 193, 96, 0.32)",
+            c800="rgba(2, 193, 96, 0.32)",
+            c900="#02C160",
+            c950="#02C160",
+        ),
+        secondary_hue=gr.themes.Color(
+            c50="#576b95",
+            c100="#576b95",
+            c200="#576b95",
+            c300="#576b95",
+            c400="#576b95",
+            c500="#576b95",
+            c600="#576b95",
+            c700="#576b95",
+            c800="#576b95",
+            c900="#576b95",
+            c950="#576b95",
+        ),
+        neutral_hue=gr.themes.Color(
+            name="gray",
+            c50="#f9fafb",
+            c100="#f3f4f6",
+            c200="#e5e7eb",
+            c300="#d1d5db",
+            c400="#B2B2B2",
+            c500="#808080",
+            c600="#636363",
+            c700="#515151",
+            c800="#393939",
+            c900="#272727",
+            c950="#171717",
+        ),
+        radius_size=gr.themes.sizes.radius_sm,
+    ).set(
+        button_primary_background_fill="#06AE56",
+        button_primary_background_fill_dark="#06AE56",
+        button_primary_background_fill_hover="#07C863",
+        button_primary_border_color="#06AE56",
+        button_primary_border_color_dark="#06AE56",
+        button_primary_text_color="#FFFFFF",
+        button_primary_text_color_dark="#FFFFFF",
+        button_secondary_background_fill="#F2F2F2",
+        button_secondary_background_fill_dark="#2B2B2B",
+        button_secondary_text_color="#393939",
+        button_secondary_text_color_dark="#FFFFFF",
+        # background_fill_primary="#F7F7F7",
+        # background_fill_primary_dark="#1F1F1F",
+        block_title_text_color="*primary_500",
+        block_title_background_fill = "*primary_100",
+        input_background_fill="#F6F6F6",
+    ),
+) as demo:
     history = gr.State([])
     token_count = gr.State([])
     promptTemplates = gr.State(load_template(get_template_names(plain=True)[0], mode=2))
+    user_api_key = gr.State(my_api_key)
     TRUECOMSTANT = gr.State(True)
     FALSECONSTANT = gr.State(False)
     topic = gr.State("未命名对话历史记录")
@@ -58,114 +135,318 @@ with gr.Blocks(css=customCSS) as demo:
     with gr.Row(scale=1).style(equal_height=True):
         with gr.Column(scale=5):
             with gr.Row(scale=1):
-                chatbot = gr.Chatbot().style(height=600)  # .style(color_map=("#1D51EE", "#585A5B"))
+                chatbot = gr.Chatbot(elem_id="chuanhu_chatbot").style(height="100%")
             with gr.Row(scale=1):
                 with gr.Column(scale=12):
-                    user_input = gr.Textbox(show_label=False, placeholder="在这里输入").style(
-                        container=False)
-                with gr.Column(min_width=50, scale=1):
-                    submitBtn = gr.Button("🚀", variant="primary")
+                    user_input = gr.Textbox(
+                        show_label=False, placeholder="在这里输入"
+                    ).style(container=False)
+                with gr.Column(min_width=70, scale=1):
+                    submitBtn = gr.Button("发送", variant="primary")
             with gr.Row(scale=1):
-                emptyBtn = gr.Button("🧹 新的对话",)
+                emptyBtn = gr.Button(
+                    "🧹 新的对话",
+                )
                 retryBtn = gr.Button("🔄 重新生成")
                 delLastBtn = gr.Button("🗑️ 删除一条对话")
                 reduceTokenBtn = gr.Button("♻️ 总结对话")
 
         with gr.Column():
-            with gr.Column(min_width=50,scale=1):
+            with gr.Column(min_width=50, scale=1):
                 with gr.Tab(label="ChatGPT"):
-                    keyTxt = gr.Textbox(show_label=True, placeholder=f"OpenAI API-key...",value=my_api_key, type="password", visible=not HIDE_MY_KEY, label="API-Key")
-                    model_select_dropdown = gr.Dropdown(label="选择模型", choices=MODELS, multiselect=False, value=MODELS[0])
-                    with gr.Accordion("参数", open=False):
-                        top_p = gr.Slider(minimum=-0, maximum=1.0, value=1.0, step=0.05,
-                                        interactive=True, label="Top-p (nucleus sampling)",)
-                        temperature = gr.Slider(minimum=-0, maximum=2.0, value=1.0,
-                                                step=0.1, interactive=True, label="Temperature",)
-                    use_streaming_checkbox = gr.Checkbox(label="实时传输回答", value=True, visible=enable_streaming_option)
+                    keyTxt = gr.Textbox(
+                        show_label=True,
+                        placeholder=f"OpenAI API-key...",
+                        value=hide_middle_chars(my_api_key),
+                        type="password",
+                        visible=not HIDE_MY_KEY,
+                        label="API-Key(按Enter提交)",
+                    )
+                    model_select_dropdown = gr.Dropdown(
+                        label="选择模型", choices=MODELS, multiselect=False, value=MODELS[0]
+                    )
+                    use_streaming_checkbox = gr.Checkbox(
+                        label="实时传输回答", value=True, visible=enable_streaming_option
+                    )
                     use_websearch_checkbox = gr.Checkbox(label="使用在线搜索", value=False)
+                    index_files = gr.Files(label="上传索引文件", type="file", multiple=True)
 
                 with gr.Tab(label="Prompt"):
-                    systemPromptTxt = gr.Textbox(show_label=True, placeholder=f"在这里输入System Prompt...", label="System prompt", value=initial_prompt).style(container=True)
+                    systemPromptTxt = gr.Textbox(
+                        show_label=True,
+                        placeholder=f"在这里输入System Prompt...",
+                        label="System prompt",
+                        value=initial_prompt,
+                        lines=10,
+                    ).style(container=True)
                     with gr.Accordion(label="加载Prompt模板", open=True):
                         with gr.Column():
                             with gr.Row():
                                 with gr.Column(scale=6):
-                                    templateFileSelectDropdown = gr.Dropdown(label="选择Prompt模板集合文件", choices=get_template_names(plain=True), multiselect=False, value=get_template_names(plain=True)[0])
+                                    templateFileSelectDropdown = gr.Dropdown(
+                                        label="选择Prompt模板集合文件",
+                                        choices=get_template_names(plain=True),
+                                        multiselect=False,
+                                        value=get_template_names(plain=True)[0],
+                                    )
                                 with gr.Column(scale=1):
                                     templateRefreshBtn = gr.Button("🔄 刷新")
                             with gr.Row():
                                 with gr.Column():
-                                    templateSelectDropdown = gr.Dropdown(label="从Prompt模板中加载", choices=load_template(get_template_names(plain=True)[0], mode=1), multiselect=False, value=load_template(get_template_names(plain=True)[0], mode=1)[0])
+                                    templateSelectDropdown = gr.Dropdown(
+                                        label="从Prompt模板中加载",
+                                        choices=load_template(
+                                            get_template_names(plain=True)[0], mode=1
+                                        ),
+                                        multiselect=False,
+                                        value=load_template(
+                                            get_template_names(plain=True)[0], mode=1
+                                        )[0],
+                                    )
 
                 with gr.Tab(label="保存/加载"):
                     with gr.Accordion(label="保存/加载对话历史记录", open=True):
-                        gr.Markdown("对话历史默认保存在history文件夹中。")
                         with gr.Column():
                             with gr.Row():
                                 with gr.Column(scale=6):
-                                    saveFileName = gr.Textbox(
-                                        show_label=True, placeholder=f"设置文件名: 默认为.json，可选为.md", label="设置保存文件名", value="对话历史记录").style(container=True)
+                                    historyFileSelectDropdown = gr.Dropdown(
+                                        label="从列表中加载对话",
+                                        choices=get_history_names(plain=True),
+                                        multiselect=False,
+                                        value=get_history_names(plain=True)[0],
+                                    )
                                 with gr.Column(scale=1):
-                                    saveHistoryBtn = gr.Button("💾 保存对话")
-                                    exportMarkdownBtn = gr.Button("📝 导出为Markdown")
+                                    historyRefreshBtn = gr.Button("🔄 刷新")
                             with gr.Row():
                                 with gr.Column(scale=6):
-                                    historyFileSelectDropdown = gr.Dropdown(label="从列表中加载对话", choices=get_history_names(plain=True), multiselect=False, value=get_history_names(plain=True)[0])
+                                    saveFileName = gr.Textbox(
+                                        show_label=True,
+                                        placeholder=f"设置文件名: 默认为.json，可选为.md",
+                                        label="设置保存文件名",
+                                        value="对话历史记录",
+                                    ).style(container=True)
                                 with gr.Column(scale=1):
-                                    historyRefreshBtn = gr.Button("🔄 刷新")
+                                    saveHistoryBtn = gr.Button("💾 保存对话")
+                                    exportMarkdownBtn = gr.Button("📝 导出为Markdown")
+                                    gr.Markdown("默认保存于history文件夹")
                             with gr.Row():
                                 with gr.Column():
                                     downloadFile = gr.File(interactive=True)
 
+                with gr.Tab(label="高级"):
+                    default_btn = gr.Button("🔙 恢复默认设置")
+                    gr.Markdown("# ⚠️ 务必谨慎更改 ⚠️\n\n如果无法使用请恢复默认设置")
+
+                    with gr.Accordion("参数", open=False):
+                        top_p = gr.Slider(
+                            minimum=-0,
+                            maximum=1.0,
+                            value=1.0,
+                            step=0.05,
+                            interactive=True,
+                            label="Top-p (nucleus sampling)",
+                        )
+                        temperature = gr.Slider(
+                            minimum=-0,
+                            maximum=2.0,
+                            value=1.0,
+                            step=0.1,
+                            interactive=True,
+                            label="Temperature",
+                        )
+
+                    apiurlTxt = gr.Textbox(
+                        show_label=True,
+                        placeholder=f"在这里输入API地址...",
+                        label="API地址",
+                        value="https://api.openai.com/v1/chat/completions",
+                        lines=2,
+                    )
+                    changeAPIURLBtn = gr.Button("🔄 切换API地址")
+                    proxyTxt = gr.Textbox(
+                        show_label=True,
+                        placeholder=f"在这里输入代理地址...",
+                        label="代理地址（示例：http://127.0.0.1:10809）",
+                        value="",
+                        lines=2,
+                    )
+                    changeProxyBtn = gr.Button("🔄 设置代理地址")
+
     gr.Markdown(description)
 
+    keyTxt.submit(submit_key, keyTxt, [user_api_key, status_display])
     # Chatbot
-    user_input.submit(predict, [keyTxt, systemPromptTxt, history, user_input, chatbot, token_count, top_p, temperature, use_streaming_checkbox, model_select_dropdown, use_websearch_checkbox], [chatbot, history, status_display, token_count], show_progress=True)
+    user_input.submit(
+        predict,
+        [
+            user_api_key,
+            systemPromptTxt,
+            history,
+            user_input,
+            chatbot,
+            token_count,
+            top_p,
+            temperature,
+            use_streaming_checkbox,
+            model_select_dropdown,
+            use_websearch_checkbox,
+            index_files
+        ],
+        [chatbot, history, status_display, token_count],
+        show_progress=True,
+    )
     user_input.submit(reset_textbox, [], [user_input])
 
-    submitBtn.click(predict, [keyTxt, systemPromptTxt, history, user_input, chatbot, token_count, top_p, temperature, use_streaming_checkbox, model_select_dropdown, use_websearch_checkbox], [chatbot, history, status_display, token_count], show_progress=True)
+    submitBtn.click(
+        predict,
+        [
+            user_api_key,
+            systemPromptTxt,
+            history,
+            user_input,
+            chatbot,
+            token_count,
+            top_p,
+            temperature,
+            use_streaming_checkbox,
+            model_select_dropdown,
+            use_websearch_checkbox,
+            index_files
+        ],
+        [chatbot, history, status_display, token_count],
+        show_progress=True,
+    )
     submitBtn.click(reset_textbox, [], [user_input])
 
-    emptyBtn.click(reset_state, outputs=[chatbot, history, token_count, status_display], show_progress=True)
+    emptyBtn.click(
+        reset_state,
+        outputs=[chatbot, history, token_count, status_display],
+        show_progress=True,
+    )
 
-    retryBtn.click(retry, [keyTxt, systemPromptTxt, history, chatbot, token_count, top_p, temperature, use_streaming_checkbox, model_select_dropdown], [chatbot, history, status_display, token_count], show_progress=True)
+    retryBtn.click(
+        retry,
+        [
+            user_api_key,
+            systemPromptTxt,
+            history,
+            chatbot,
+            token_count,
+            top_p,
+            temperature,
+            use_streaming_checkbox,
+            model_select_dropdown,
+        ],
+        [chatbot, history, status_display, token_count],
+        show_progress=True,
+    )
 
-    delLastBtn.click(delete_last_conversation, [chatbot, history, token_count], [
-                     chatbot, history, token_count, status_display], show_progress=True)
+    delLastBtn.click(
+        delete_last_conversation,
+        [chatbot, history, token_count],
+        [chatbot, history, token_count, status_display],
+        show_progress=True,
+    )
 
-    reduceTokenBtn.click(reduce_token_size, [keyTxt, systemPromptTxt, history, chatbot, token_count, top_p, temperature, use_streaming_checkbox, model_select_dropdown], [chatbot, history, status_display, token_count], show_progress=True)
+    reduceTokenBtn.click(
+        reduce_token_size,
+        [
+            user_api_key,
+            systemPromptTxt,
+            history,
+            chatbot,
+            token_count,
+            top_p,
+            temperature,
+            use_streaming_checkbox,
+            model_select_dropdown,
+        ],
+        [chatbot, history, status_display, token_count],
+        show_progress=True,
+    )
 
     # Template
     templateRefreshBtn.click(get_template_names, None, [templateFileSelectDropdown])
-    templateFileSelectDropdown.change(load_template, [templateFileSelectDropdown],  [promptTemplates, templateSelectDropdown], show_progress=True)
-    templateSelectDropdown.change(get_template_content, [promptTemplates, templateSelectDropdown, systemPromptTxt],  [systemPromptTxt], show_progress=True)
+    templateFileSelectDropdown.change(
+        load_template,
+        [templateFileSelectDropdown],
+        [promptTemplates, templateSelectDropdown],
+        show_progress=True,
+    )
+    templateSelectDropdown.change(
+        get_template_content,
+        [promptTemplates, templateSelectDropdown, systemPromptTxt],
+        [systemPromptTxt],
+        show_progress=True,
+    )
 
     # S&L
-    saveHistoryBtn.click(save_chat_history, [saveFileName, systemPromptTxt, history, chatbot], downloadFile, show_progress=True)
+    saveHistoryBtn.click(
+        save_chat_history,
+        [saveFileName, systemPromptTxt, history, chatbot],
+        downloadFile,
+        show_progress=True,
+    )
     saveHistoryBtn.click(get_history_names, None, [historyFileSelectDropdown])
-    exportMarkdownBtn.click(export_markdown, [saveFileName, systemPromptTxt, history, chatbot], downloadFile, show_progress=True)
+    exportMarkdownBtn.click(
+        export_markdown,
+        [saveFileName, systemPromptTxt, history, chatbot],
+        downloadFile,
+        show_progress=True,
+    )
     historyRefreshBtn.click(get_history_names, None, [historyFileSelectDropdown])
-    historyFileSelectDropdown.change(load_chat_history, [historyFileSelectDropdown, systemPromptTxt, history, chatbot],  [saveFileName, systemPromptTxt, history, chatbot], show_progress=True)
-    downloadFile.change(load_chat_history, [downloadFile, systemPromptTxt, history, chatbot],  [saveFileName, systemPromptTxt, history, chatbot])
+    historyFileSelectDropdown.change(
+        load_chat_history,
+        [historyFileSelectDropdown, systemPromptTxt, history, chatbot],
+        [saveFileName, systemPromptTxt, history, chatbot],
+        show_progress=True,
+    )
+    downloadFile.change(
+        load_chat_history,
+        [downloadFile, systemPromptTxt, history, chatbot],
+        [saveFileName, systemPromptTxt, history, chatbot],
+    )
 
+    # Advanced
+    default_btn.click(
+        reset_default, [], [apiurlTxt, proxyTxt, status_display], show_progress=True
+    )
+    changeAPIURLBtn.click(
+        change_api_url,
+        [apiurlTxt],
+        [status_display],
+        show_progress=True,
+    )
+    changeProxyBtn.click(
+        change_proxy,
+        [proxyTxt],
+        [status_display],
+        show_progress=True,
+    )
 
-logging.info(colorama.Back.GREEN + "\n川虎的温馨提示：访问 http://localhost:7860 查看界面" + colorama.Style.RESET_ALL)
+logging.info(
+    colorama.Back.GREEN
+    + "\n川虎的温馨提示：访问 http://localhost:7860 查看界面"
+    + colorama.Style.RESET_ALL
+)
 # 默认开启本地服务器，默认可以直接从IP访问，默认不创建公开分享链接
 demo.title = "川虎ChatGPT 🚀"
 
 if __name__ == "__main__":
-    #if running in Docker
+    # if running in Docker
     if dockerflag:
         if authflag:
-            demo.queue().launch(server_name="0.0.0.0", server_port=7860,auth=(username, password))
+            demo.queue().launch(
+                server_name="0.0.0.0", server_port=7860, auth=(username, password)
+            )
         else:
             demo.queue().launch(server_name="0.0.0.0", server_port=7860, share=False)
-    #if not running in Docker
+    # if not running in Docker
     else:
         if authflag:
             demo.queue().launch(share=False, auth=(username, password))
         else:
-            demo.queue().launch(share=False) # 改为 share=True 可以创建公开分享链接
-        #demo.queue().launch(server_name="0.0.0.0", server_port=7860, share=False) # 可自定义端口
-        #demo.queue().launch(server_name="0.0.0.0", server_port=7860,auth=("在这里填写用户名", "在这里填写密码")) # 可设置用户名与密码
-        #demo.queue().launch(auth=("在这里填写用户名", "在这里填写密码")) # 适合Nginx反向代理
+            demo.queue().launch(share=False)  # 改为 share=True 可以创建公开分享链接
+        # demo.queue().launch(server_name="0.0.0.0", server_port=7860, share=False) # 可自定义端口
+        # demo.queue().launch(server_name="0.0.0.0", server_port=7860,auth=("在这里填写用户名", "在这里填写密码")) # 可设置用户名与密码
+        # demo.queue().launch(auth=("在这里填写用户名", "在这里填写密码")) # 适合Nginx反向代理

chat_func.py

@@ -0,0 +1,423 @@
+# -*- coding:utf-8 -*-
+from __future__ import annotations
+from typing import TYPE_CHECKING, List
+
+import logging
+import json
+import os
+import requests
+
+from tqdm import tqdm
+import colorama
+from duckduckgo_search import ddg
+
+from presets import *
+from llama_func import *
+from utils import *
+
+# logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] [%(filename)s:%(lineno)d] %(message)s")
+
+if TYPE_CHECKING:
+    from typing import TypedDict
+
+    class DataframeData(TypedDict):
+        headers: List[str]
+        data: List[List[str | int | bool]]
+
+
+initial_prompt = "You are a helpful assistant."
+API_URL = "https://api.openai.com/v1/chat/completions"
+HISTORY_DIR = "history"
+TEMPLATES_DIR = "templates"
+
+def get_response(
+    openai_api_key, system_prompt, history, temperature, top_p, stream, selected_model
+):
+    headers = {
+        "Content-Type": "application/json",
+        "Authorization": f"Bearer {openai_api_key}",
+    }
+
+    history = [construct_system(system_prompt), *history]
+
+    payload = {
+        "model": selected_model,
+        "messages": history,  # [{"role": "user", "content": f"{inputs}"}],
+        "temperature": temperature,  # 1.0,
+        "top_p": top_p,  # 1.0,
+        "n": 1,
+        "stream": stream,
+        "presence_penalty": 0,
+        "frequency_penalty": 0,
+    }
+    if stream:
+        timeout = timeout_streaming
+    else:
+        timeout = timeout_all
+
+    # 获取环境变量中的代理设置
+    http_proxy = os.environ.get("HTTP_PROXY") or os.environ.get("http_proxy")
+    https_proxy = os.environ.get("HTTPS_PROXY") or os.environ.get("https_proxy")
+
+    # 如果存在代理设置，使用它们
+    proxies = {}
+    if http_proxy:
+        logging.info(f"Using HTTP proxy: {http_proxy}")
+        proxies["http"] = http_proxy
+    if https_proxy:
+        logging.info(f"Using HTTPS proxy: {https_proxy}")
+        proxies["https"] = https_proxy
+
+    # 如果有代理，使用代理发送请求，否则使用默认设置发送请求
+    if proxies:
+        response = requests.post(
+            API_URL,
+            headers=headers,
+            json=payload,
+            stream=True,
+            timeout=timeout,
+            proxies=proxies,
+        )
+    else:
+        response = requests.post(
+            API_URL,
+            headers=headers,
+            json=payload,
+            stream=True,
+            timeout=timeout,
+        )
+    return response
+
+
+def stream_predict(
+    openai_api_key,
+    system_prompt,
+    history,
+    inputs,
+    chatbot,
+    all_token_counts,
+    top_p,
+    temperature,
+    selected_model,
+):
+    def get_return_value():
+        return chatbot, history, status_text, all_token_counts
+
+    logging.info("实时回答模式")
+    partial_words = ""
+    counter = 0
+    status_text = "开始实时传输回答……"
+    history.append(construct_user(inputs))
+    history.append(construct_assistant(""))
+    chatbot.append((parse_text(inputs), ""))
+    user_token_count = 0
+    if len(all_token_counts) == 0:
+        system_prompt_token_count = count_token(construct_system(system_prompt))
+        user_token_count = (
+            count_token(construct_user(inputs)) + system_prompt_token_count
+        )
+    else:
+        user_token_count = count_token(construct_user(inputs))
+    all_token_counts.append(user_token_count)
+    logging.info(f"输入token计数: {user_token_count}")
+    yield get_return_value()
+    try:
+        response = get_response(
+            openai_api_key,
+            system_prompt,
+            history,
+            temperature,
+            top_p,
+            True,
+            selected_model,
+        )
+    except requests.exceptions.ConnectTimeout:
+        status_text = (
+            standard_error_msg + connection_timeout_prompt + error_retrieve_prompt
+        )
+        yield get_return_value()
+        return
+    except requests.exceptions.ReadTimeout:
+        status_text = standard_error_msg + read_timeout_prompt + error_retrieve_prompt
+        yield get_return_value()
+        return
+
+    yield get_return_value()
+    error_json_str = ""
+
+    for chunk in tqdm(response.iter_lines()):
+        if counter == 0:
+            counter += 1
+            continue
+        counter += 1
+        # check whether each line is non-empty
+        if chunk:
+            chunk = chunk.decode()
+            chunklength = len(chunk)
+            try:
+                chunk = json.loads(chunk[6:])
+            except json.JSONDecodeError:
+                logging.info(chunk)
+                error_json_str += chunk
+                status_text = f"JSON解析错误。请重置对话。收到的内容: {error_json_str}"
+                yield get_return_value()
+                continue
+            # decode each line as response data is in bytes
+            if chunklength > 6 and "delta" in chunk["choices"][0]:
+                finish_reason = chunk["choices"][0]["finish_reason"]
+                status_text = construct_token_message(
+                    sum(all_token_counts), stream=True
+                )
+                if finish_reason == "stop":
+                    yield get_return_value()
+                    break
+                try:
+                    partial_words = (
+                        partial_words + chunk["choices"][0]["delta"]["content"]
+                    )
+                except KeyError:
+                    status_text = (
+                        standard_error_msg
+                        + "API回复中找不到内容。很可能是Token计数达到上限了。请重置对话。当前Token计数: "
+                        + str(sum(all_token_counts))
+                    )
+                    yield get_return_value()
+                    break
+                history[-1] = construct_assistant(partial_words)
+                chatbot[-1] = (parse_text(inputs), parse_text(partial_words))
+                all_token_counts[-1] += 1
+                yield get_return_value()
+
+
+def predict_all(
+    openai_api_key,
+    system_prompt,
+    history,
+    inputs,
+    chatbot,
+    all_token_counts,
+    top_p,
+    temperature,
+    selected_model,
+):
+    logging.info("一次性回答模式")
+    history.append(construct_user(inputs))
+    history.append(construct_assistant(""))
+    chatbot.append((parse_text(inputs), ""))
+    all_token_counts.append(count_token(construct_user(inputs)))
+    try:
+        response = get_response(
+            openai_api_key,
+            system_prompt,
+            history,
+            temperature,
+            top_p,
+            False,
+            selected_model,
+        )
+    except requests.exceptions.ConnectTimeout:
+        status_text = (
+            standard_error_msg + connection_timeout_prompt + error_retrieve_prompt
+        )
+        return chatbot, history, status_text, all_token_counts
+    except requests.exceptions.ProxyError:
+        status_text = standard_error_msg + proxy_error_prompt + error_retrieve_prompt
+        return chatbot, history, status_text, all_token_counts
+    except requests.exceptions.SSLError:
+        status_text = standard_error_msg + ssl_error_prompt + error_retrieve_prompt
+        return chatbot, history, status_text, all_token_counts
+    response = json.loads(response.text)
+    content = response["choices"][0]["message"]["content"]
+    history[-1] = construct_assistant(content)
+    chatbot[-1] = (parse_text(inputs), parse_text(content))
+    total_token_count = response["usage"]["total_tokens"]
+    all_token_counts[-1] = total_token_count - sum(all_token_counts)
+    status_text = construct_token_message(total_token_count)
+    return chatbot, history, status_text, all_token_counts
+
+
+def predict(
+    openai_api_key,
+    system_prompt,
+    history,
+    inputs,
+    chatbot,
+    all_token_counts,
+    top_p,
+    temperature,
+    stream=False,
+    selected_model=MODELS[0],
+    use_websearch_checkbox=False,
+    files = None,
+    should_check_token_count=True,
+):  # repetition_penalty, top_k
+    logging.info("输入为：" + colorama.Fore.BLUE + f"{inputs}" + colorama.Style.RESET_ALL)
+    if files:
+        msg = "构建索引中……（这可能需要比较久的时间）"
+        logging.info(msg)
+        yield chatbot, history, msg, all_token_counts
+        index = construct_index(openai_api_key, file_src=files)
+        msg = "索引构建完成，获取回答中……"
+        yield chatbot, history, msg, all_token_counts
+        history, chatbot, status_text = chat_ai(openai_api_key, index, inputs, history, chatbot)
+        yield chatbot, history, status_text, all_token_counts
+        return
+    if use_websearch_checkbox:
+        results = ddg(inputs, max_results=3)
+        web_results = []
+        for idx, result in enumerate(results):
+            logging.info(f"搜索结果{idx + 1}：{result}")
+            web_results.append(f'[{idx+1}]"{result["body"]}"\nURL: {result["href"]}')
+        web_results = "\n\n".join(web_results)
+        inputs = (
+            replace_today(WEBSEARCH_PTOMPT_TEMPLATE)
+            .replace("{query}", inputs)
+            .replace("{web_results}", web_results)
+        )
+    if len(openai_api_key) != 51:
+        status_text = standard_error_msg + no_apikey_msg
+        logging.info(status_text)
+        chatbot.append((parse_text(inputs), ""))
+        if len(history) == 0:
+            history.append(construct_user(inputs))
+            history.append("")
+            all_token_counts.append(0)
+        else:
+            history[-2] = construct_user(inputs)
+        yield chatbot, history, status_text, all_token_counts
+        return
+    if stream:
+        yield chatbot, history, "开始生成回答……", all_token_counts
+    if stream:
+        logging.info("使用流式传输")
+        iter = stream_predict(
+            openai_api_key,
+            system_prompt,
+            history,
+            inputs,
+            chatbot,
+            all_token_counts,
+            top_p,
+            temperature,
+            selected_model,
+        )
+        for chatbot, history, status_text, all_token_counts in iter:
+            yield chatbot, history, status_text, all_token_counts
+    else:
+        logging.info("不使用流式传输")
+        chatbot, history, status_text, all_token_counts = predict_all(
+            openai_api_key,
+            system_prompt,
+            history,
+            inputs,
+            chatbot,
+            all_token_counts,
+            top_p,
+            temperature,
+            selected_model,
+        )
+        yield chatbot, history, status_text, all_token_counts
+    logging.info(f"传输完毕。当前token计数为{all_token_counts}")
+    if len(history) > 1 and history[-1]["content"] != inputs:
+        logging.info(
+            "回答为："
+            + colorama.Fore.BLUE
+            + f"{history[-1]['content']}"
+            + colorama.Style.RESET_ALL
+        )
+    if stream:
+        max_token = max_token_streaming
+    else:
+        max_token = max_token_all
+    if sum(all_token_counts) > max_token and should_check_token_count:
+        status_text = f"精简token中{all_token_counts}/{max_token}"
+        logging.info(status_text)
+        yield chatbot, history, status_text, all_token_counts
+        iter = reduce_token_size(
+            openai_api_key,
+            system_prompt,
+            history,
+            chatbot,
+            all_token_counts,
+            top_p,
+            temperature,
+            stream=False,
+            selected_model=selected_model,
+            hidden=True,
+        )
+        for chatbot, history, status_text, all_token_counts in iter:
+            status_text = f"Token 达到上限，已自动降低Token计数至 {status_text}"
+            yield chatbot, history, status_text, all_token_counts
+
+
+def retry(
+    openai_api_key,
+    system_prompt,
+    history,
+    chatbot,
+    token_count,
+    top_p,
+    temperature,
+    stream=False,
+    selected_model=MODELS[0],
+):
+    logging.info("重试中……")
+    if len(history) == 0:
+        yield chatbot, history, f"{standard_error_msg}上下文是空的", token_count
+        return
+    history.pop()
+    inputs = history.pop()["content"]
+    token_count.pop()
+    iter = predict(
+        openai_api_key,
+        system_prompt,
+        history,
+        inputs,
+        chatbot,
+        token_count,
+        top_p,
+        temperature,
+        stream=stream,
+        selected_model=selected_model,
+    )
+    logging.info("重试完毕")
+    for x in iter:
+        yield x
+
+
+def reduce_token_size(
+    openai_api_key,
+    system_prompt,
+    history,
+    chatbot,
+    token_count,
+    top_p,
+    temperature,
+    stream=False,
+    selected_model=MODELS[0],
+    hidden=False,
+):
+    logging.info("开始减少token数量……")
+    iter = predict(
+        openai_api_key,
+        system_prompt,
+        history,
+        summarize_prompt,
+        chatbot,
+        token_count,
+        top_p,
+        temperature,
+        stream=stream,
+        selected_model=selected_model,
+        should_check_token_count=False,
+    )
+    logging.info(f"chatbot: {chatbot}")
+    for chatbot, history, status_text, previous_token_count in iter:
+        history = history[-2:]
+        token_count = previous_token_count[-1:]
+        if hidden:
+            chatbot.pop()
+        yield chatbot, history, construct_token_message(
+            sum(token_count), stream=stream
+        ), token_count
+    logging.info("减少token数量完毕")

custom.css

@@ -0,0 +1,188 @@
+/* status_display */
+#status_display {
+    display: flex;
+    min-height: 2.5em;
+    align-items: flex-end;
+    justify-content: flex-end;
+}
+#status_display p {
+    font-size: .85em;
+    font-family: monospace;
+    color: var(--text-color-subdued) !important;
+}
+/* chatbot */
+:root {
+    --bg-color-light: #F3F3F3;
+    --bg-color-dark: #121111;
+  }
+
+@media (prefers-color-scheme: light) {
+    #chuanhu_chatbot {
+        background-color: var(--bg-color-light) !important;
+    }
+    [data-testid = "bot"] {
+        background-color: #FFFFFF !important;
+    }
+    [data-testid = "user"] {
+        background-color: #95EC69 !important;
+    }
+}
+
+@media (prefers-color-scheme: dark) {
+    #chuanhu_chatbot {
+        background-color: var(--bg-color-dark) !important;
+    }
+    [data-testid = "bot"] {
+        background-color: #2C2C2C !important;
+    }
+    [data-testid = "user"] {
+        background-color: #26B561 !important;
+    }
+}
+
+/* 屏幕宽度大于等于500px的设备 */
+@media (min-width: 500px) {
+    #chuanhu_chatbot {
+        height: calc(100vh - 200px);
+    }
+    #chuanhu_chatbot .wrap {
+        max-height: calc(100vh - 200px - var(--line-sm)*1rem - 2*var(--block-label-margin) );
+    }
+}
+/* 屏幕宽度小于500px的设备 */
+@media (max-width: 499px) {
+    #chuanhu_chatbot {
+        height: calc(100vh - 140px);
+    }
+    #chuanhu_chatbot .wrap {
+        max-height: calc(100vh - 140 - var(--line-sm)*1rem - 2*var(--block-label-margin) );
+    }
+}
+/* 对话气泡 */
+[class *= "message"] {
+    border-radius: var(--radius-xl) !important;
+    border: none;
+    padding: var(--spacing-xl) !important;
+    font-size: var(--text-md) !important;
+    line-height: var(--line-md) !important;
+}
+[data-testid = "bot"] {
+    max-width: 85%;
+    border-bottom-left-radius: 0 !important;
+}
+[data-testid = "user"] {
+    max-width: 85%;
+    width: auto !important;
+    border-bottom-right-radius: 0 !important;
+}
+/* 表格 */
+table {
+    margin: 1em 0;
+    border-collapse: collapse;
+    empty-cells: show;
+}
+td,th {
+    border: 1.2px solid var(--color-border-primary) !important;
+    padding: 0.2em;
+}
+thead {
+    background-color: rgba(175,184,193,0.2);
+}
+thead th {
+    padding: .5em .2em;
+}
+/* 行内代码 */
+code {
+    display: inline;
+    white-space: break-spaces;
+    border-radius: 6px;
+    margin: 0 2px 0 2px;
+    padding: .2em .4em .1em .4em;
+    background-color: rgba(175,184,193,0.2);
+}
+/* 代码块 */
+pre code {
+    display: block;
+    white-space: pre;
+    background-color: hsla(0, 0%, 0%, 80%)!important;
+    border-radius: 10px;
+    padding: 1rem 1.2rem 1rem;
+    margin: 1.2em 2em 1.2em 0.5em;
+    color: #FFF;
+    box-shadow: 6px 6px 16px hsla(0, 0%, 0%, 0.2);
+}
+/* 代码高亮样式 */
+.codehilite .hll { background-color: #49483e }
+.codehilite .c { color: #75715e } /* Comment */
+.codehilite .err { color: #960050; background-color: #1e0010 } /* Error */
+.codehilite .k { color: #66d9ef } /* Keyword */
+.codehilite .l { color: #ae81ff } /* Literal */
+.codehilite .n { color: #f8f8f2 } /* Name */
+.codehilite .o { color: #f92672 } /* Operator */
+.codehilite .p { color: #f8f8f2 } /* Punctuation */
+.codehilite .ch { color: #75715e } /* Comment.Hashbang */
+.codehilite .cm { color: #75715e } /* Comment.Multiline */
+.codehilite .cp { color: #75715e } /* Comment.Preproc */
+.codehilite .cpf { color: #75715e } /* Comment.PreprocFile */
+.codehilite .c1 { color: #75715e } /* Comment.Single */
+.codehilite .cs { color: #75715e } /* Comment.Special */
+.codehilite .gd { color: #f92672 } /* Generic.Deleted */
+.codehilite .ge { font-style: italic } /* Generic.Emph */
+.codehilite .gi { color: #a6e22e } /* Generic.Inserted */
+.codehilite .gs { font-weight: bold } /* Generic.Strong */
+.codehilite .gu { color: #75715e } /* Generic.Subheading */
+.codehilite .kc { color: #66d9ef } /* Keyword.Constant */
+.codehilite .kd { color: #66d9ef } /* Keyword.Declaration */
+.codehilite .kn { color: #f92672 } /* Keyword.Namespace */
+.codehilite .kp { color: #66d9ef } /* Keyword.Pseudo */
+.codehilite .kr { color: #66d9ef } /* Keyword.Reserved */
+.codehilite .kt { color: #66d9ef } /* Keyword.Type */
+.codehilite .ld { color: #e6db74 } /* Literal.Date */
+.codehilite .m { color: #ae81ff } /* Literal.Number */
+.codehilite .s { color: #e6db74 } /* Literal.String */
+.codehilite .na { color: #a6e22e } /* Name.Attribute */
+.codehilite .nb { color: #f8f8f2 } /* Name.Builtin */
+.codehilite .nc { color: #a6e22e } /* Name.Class */
+.codehilite .no { color: #66d9ef } /* Name.Constant */
+.codehilite .nd { color: #a6e22e } /* Name.Decorator */
+.codehilite .ni { color: #f8f8f2 } /* Name.Entity */
+.codehilite .ne { color: #a6e22e } /* Name.Exception */
+.codehilite .nf { color: #a6e22e } /* Name.Function */
+.codehilite .nl { color: #f8f8f2 } /* Name.Label */
+.codehilite .nn { color: #f8f8f2 } /* Name.Namespace */
+.codehilite .nx { color: #a6e22e } /* Name.Other */
+.codehilite .py { color: #f8f8f2 } /* Name.Property */
+.codehilite .nt { color: #f92672 } /* Name.Tag */
+.codehilite .nv { color: #f8f8f2 } /* Name.Variable */
+.codehilite .ow { color: #f92672 } /* Operator.Word */
+.codehilite .w { color: #f8f8f2 } /* Text.Whitespace */
+.codehilite .mb { color: #ae81ff } /* Literal.Number.Bin */
+.codehilite .mf { color: #ae81ff } /* Literal.Number.Float */
+.codehilite .mh { color: #ae81ff } /* Literal.Number.Hex */
+.codehilite .mi { color: #ae81ff } /* Literal.Number.Integer */
+.codehilite .mo { color: #ae81ff } /* Literal.Number.Oct */
+.codehilite .sa { color: #e6db74 } /* Literal.String.Affix */
+.codehilite .sb { color: #e6db74 } /* Literal.String.Backtick */
+.codehilite .sc { color: #e6db74 } /* Literal.String.Char */
+.codehilite .dl { color: #e6db74 } /* Literal.String.Delimiter */
+.codehilite .sd { color: #e6db74 } /* Literal.String.Doc */
+.codehilite .s2 { color: #e6db74 } /* Literal.String.Double */
+.codehilite .se { color: #ae81ff } /* Literal.String.Escape */
+.codehilite .sh { color: #e6db74 } /* Literal.String.Heredoc */
+.codehilite .si { color: #e6db74 } /* Literal.String.Interpol */
+.codehilite .sx { color: #e6db74 } /* Literal.String.Other */
+.codehilite .sr { color: #e6db74 } /* Literal.String.Regex */
+.codehilite .s1 { color: #e6db74 } /* Literal.String.Single */
+.codehilite .ss { color: #e6db74 } /* Literal.String.Symbol */
+.codehilite .bp { color: #f8f8f2 } /* Name.Builtin.Pseudo */
+.codehilite .fm { color: #a6e22e } /* Name.Function.Magic */
+.codehilite .vc { color: #f8f8f2 } /* Name.Variable.Class */
+.codehilite .vg { color: #f8f8f2 } /* Name.Variable.Global */
+.codehilite .vi { color: #f8f8f2 } /* Name.Variable.Instance */
+.codehilite .vm { color: #f8f8f2 } /* Name.Variable.Magic */
+.codehilite .il { color: #ae81ff } /* Literal.Number.Integer.Long */
+
+/* 全局元素 */
+* {
+    transition: all 0.6s;
+}

llama_func.py

@@ -0,0 +1,192 @@
+import os
+import logging
+
+from llama_index import GPTSimpleVectorIndex
+from llama_index import download_loader
+from llama_index import (
+    Document,
+    LLMPredictor,
+    PromptHelper,
+    QuestionAnswerPrompt,
+    RefinePrompt,
+)
+from langchain.llms import OpenAI
+import colorama
+
+
+from presets import *
+from utils import *
+
+
+def get_documents(file_src):
+    documents = []
+    index_name = ""
+    logging.debug("Loading documents...")
+    logging.debug(f"file_src: {file_src}")
+    for file in file_src:
+        logging.debug(f"file: {file.name}")
+        index_name += file.name
+        if os.path.splitext(file.name)[1] == ".pdf":
+            logging.debug("Loading PDF...")
+            CJKPDFReader = download_loader("CJKPDFReader")
+            loader = CJKPDFReader()
+            documents += loader.load_data(file=file.name)
+        elif os.path.splitext(file.name)[1] == ".docx":
+            logging.debug("Loading DOCX...")
+            DocxReader = download_loader("DocxReader")
+            loader = DocxReader()
+            documents += loader.load_data(file=file.name)
+        elif os.path.splitext(file.name)[1] == ".epub":
+            logging.debug("Loading EPUB...")
+            EpubReader = download_loader("EpubReader")
+            loader = EpubReader()
+            documents += loader.load_data(file=file.name)
+        else:
+            logging.debug("Loading text file...")
+            with open(file.name, "r", encoding="utf-8") as f:
+                text = add_space(f.read())
+                documents += [Document(text)]
+    index_name = sha1sum(index_name)
+    return documents, index_name
+
+
+def construct_index(
+    api_key,
+    file_src,
+    max_input_size=4096,
+    num_outputs=1,
+    max_chunk_overlap=20,
+    chunk_size_limit=600,
+    embedding_limit=None,
+    separator=" ",
+    num_children=10,
+    max_keywords_per_chunk=10,
+):
+    os.environ["OPENAI_API_KEY"] = api_key
+    chunk_size_limit = None if chunk_size_limit == 0 else chunk_size_limit
+    embedding_limit = None if embedding_limit == 0 else embedding_limit
+    separator = " " if separator == "" else separator
+
+    llm_predictor = LLMPredictor(
+        llm=OpenAI(model_name="gpt-3.5-turbo-0301", openai_api_key=api_key)
+    )
+    prompt_helper = PromptHelper(
+        max_input_size,
+        num_outputs,
+        max_chunk_overlap,
+        embedding_limit,
+        chunk_size_limit,
+        separator=separator,
+    )
+    documents, index_name = get_documents(file_src)
+    if os.path.exists(f"./index/{index_name}.json"):
+        logging.info("找到了缓存的索引文件，加载中……")
+        return GPTSimpleVectorIndex.load_from_disk(f"./index/{index_name}.json")
+    else:
+        try:
+            logging.debug("构建索引中……")
+            index = GPTSimpleVectorIndex(
+                documents, llm_predictor=llm_predictor, prompt_helper=prompt_helper
+            )
+            os.makedirs("./index", exist_ok=True)
+            index.save_to_disk(f"./index/{index_name}.json")
+            return index
+        except Exception as e:
+            print(e)
+            return None
+
+
+def chat_ai(
+    api_key,
+    index,
+    question,
+    context,
+    chatbot,
+):
+    os.environ["OPENAI_API_KEY"] = api_key
+
+    logging.info(f"Question: {question}")
+
+    response, chatbot_display, status_text = ask_ai(
+        api_key,
+        index,
+        question,
+        replace_today(PROMPT_TEMPLATE),
+        REFINE_TEMPLATE,
+        SIM_K,
+        INDEX_QUERY_TEMPRATURE,
+        context,
+    )
+    if response is None:
+        status_text = "查询失败，请换个问法试试"
+        return context, chatbot
+    response = response
+
+    context.append({"role": "user", "content": question})
+    context.append({"role": "assistant", "content": response})
+    chatbot.append((question, chatbot_display))
+
+    os.environ["OPENAI_API_KEY"] = ""
+    return context, chatbot, status_text
+
+
+def ask_ai(
+    api_key,
+    index,
+    question,
+    prompt_tmpl,
+    refine_tmpl,
+    sim_k=1,
+    temprature=0,
+    prefix_messages=[],
+):
+    os.environ["OPENAI_API_KEY"] = api_key
+
+    logging.debug("Index file found")
+    logging.debug("Querying index...")
+    llm_predictor = LLMPredictor(
+        llm=OpenAI(
+            temperature=temprature,
+            model_name="gpt-3.5-turbo-0301",
+            prefix_messages=prefix_messages,
+        )
+    )
+
+    response = None  # Initialize response variable to avoid UnboundLocalError
+    qa_prompt = QuestionAnswerPrompt(prompt_tmpl)
+    rf_prompt = RefinePrompt(refine_tmpl)
+    response = index.query(
+        question,
+        llm_predictor=llm_predictor,
+        similarity_top_k=sim_k,
+        text_qa_template=qa_prompt,
+        refine_template=rf_prompt,
+        response_mode="compact",
+    )
+
+    if response is not None:
+        logging.info(f"Response: {response}")
+        ret_text = response.response
+        nodes = []
+        for index, node in enumerate(response.source_nodes):
+            brief = node.source_text[:25].replace("\n", "")
+            nodes.append(
+                f"<details><summary>[{index+1}]\t{brief}...</summary><p>{node.source_text}</p></details>"
+            )
+        new_response = ret_text + "\n----------\n" + "\n\n".join(nodes)
+        logging.info(
+            f"Response: {colorama.Fore.BLUE}{ret_text}{colorama.Style.RESET_ALL}"
+        )
+        os.environ["OPENAI_API_KEY"] = ""
+        return ret_text, new_response, f"查询消耗了{llm_predictor.last_token_usage} tokens"
+    else:
+        logging.warning("No response found, returning None")
+        os.environ["OPENAI_API_KEY"] = ""
+        return None
+
+
+def add_space(text):
+    punctuations = {"，": "， ", "。": "。 ", "？": "？ ", "！": "！ ", "：": "： ", "；": "； "}
+    for cn_punc, en_punc in punctuations.items():
+        text = text.replace(cn_punc, en_punc)
+    return text

overwrites.py

@@ -0,0 +1,40 @@
+from __future__ import annotations
+import logging
+
+from llama_index import Prompt
+from typing import List, Tuple
+import mdtex2html
+
+from presets import *
+from llama_func import *
+
+
+def compact_text_chunks(self, prompt: Prompt, text_chunks: List[str]) -> List[str]:
+    logging.debug("Compacting text chunks...🚀🚀🚀")
+    combined_str = [c.strip() for c in text_chunks if c.strip()]
+    combined_str = [f"[{index+1}] {c}" for index, c in enumerate(combined_str)]
+    combined_str = "\n\n".join(combined_str)
+    # resplit based on self.max_chunk_overlap
+    text_splitter = self.get_text_splitter_given_prompt(prompt, 1, padding=1)
+    return text_splitter.split_text(combined_str)
+
+
+def postprocess(
+    self, y: List[Tuple[str | None, str | None]]
+) -> List[Tuple[str | None, str | None]]:
+    """
+    Parameters:
+        y: List of tuples representing the message and response pairs. Each message and response should be a string, which may be in Markdown format.
+    Returns:
+        List of tuples representing the message and response. Each message and response will be a string of HTML.
+    """
+    if y is None:
+        return []
+    for i, (message, response) in enumerate(y):
+        y[i] = (
+            # None if message is None else markdown.markdown(message),
+            # None if response is None else markdown.markdown(response),
+            None if message is None else message,
+            None if response is None else mdtex2html.convert(response, extensions=['fenced_code','codehilite','tables']),
+        )
+    return y

presets.py

@@ -1,6 +1,26 @@
 # -*- coding:utf-8 -*-
+# 错误信息
+standard_error_msg = "☹️发生了错误："  # 错误信息的标准前缀
+error_retrieve_prompt = "请检查网络连接，或者API-Key是否有效。"  # 获取对话时发生错误
+connection_timeout_prompt = "连接超时，无法获取对话。"  # 连接超时
+read_timeout_prompt = "读取超时，无法获取对话。"  # 读取超时
+proxy_error_prompt = "代理错误，无法获取对话。"  # 代理错误
+ssl_error_prompt = "SSL错误，无法获取对话。"  # SSL 错误
+no_apikey_msg = "API key长度不是51位，请检查是否输入正确。"  # API key 长度不足 51 位
+
+max_token_streaming = 3500  # 流式对话时的最大 token 数
+timeout_streaming = 30  # 流式对话时的超时时间
+max_token_all = 3500  # 非流式对话时的最大 token 数
+timeout_all = 200  # 非流式对话时的超时时间
+enable_streaming_option = True  # 是否启用选择选择是否实时显示回答的勾选框
+HIDE_MY_KEY = False  # 如果你想在UI中隐藏你的 API 密钥，将此值设置为 True
+
+SIM_K = 5
+INDEX_QUERY_TEMPRATURE = 1.0
+
 title = """<h1 align="left" style="min-width:200px; margin-top:0;">川虎ChatGPT 🚀</h1>"""
-description = """<div align="center" style="margin:16px 0">
+description = """\
+<div align="center" style="margin:16px 0">
 
 由Bilibili [土川虎虎虎](https://space.bilibili.com/29125536) 和 [明昭MZhao](https://space.bilibili.com/24807452)开发
 
@@ -9,62 +29,21 @@ description = """<div align="center" style="margin:16px 0">
 此App使用 `gpt-3.5-turbo` 大语言模型
 </div>
 """
-customCSS = """
-#status_display {
-    display: flex;
-    min-height: 2.5em;
-    align-items: flex-end;
-    justify-content: flex-end;
-}
-#status_display p {
-    font-size: .85em;
-    font-family: monospace;
-    color: var(--text-color-subdued) !important;
-}
-[class *= "message"] {
-    border-radius: var(--radius-xl) !important;
-    border: none;
-    padding: var(--spacing-xl) !important;
-    font-size: var(--text-md) !important;
-    line-height: var(--line-md) !important;
-}
-[data-testid = "bot"] {
-    max-width: 85%;
-    border-bottom-left-radius: 0 !important;
-}
-[data-testid = "user"] {
-    max-width: 85%;
-    width: auto !important;
-    border-bottom-right-radius: 0 !important;
-}
-code {
-    display: inline;
-    white-space: break-spaces;
-    border-radius: 6px;
-    margin: 0 2px 0 2px;
-    padding: .2em .4em .1em .4em;
-    background-color: rgba(175,184,193,0.2);
-}
-pre code {
-    display: block;
-    white-space: pre;
-    background-color: hsla(0, 0%, 0%, 72%);
-    border: solid 5px var(--color-border-primary) !important;
-    border-radius: 10px;
-    padding: 0 1.2rem 1.2rem;
-    margin-top: 1em !important;
-    color: #FFF;
-    box-shadow: inset 0px 8px 16px hsla(0, 0%, 0%, .2)
-}
 
-* {
-    transition: all 0.6s;
-}
-"""
+summarize_prompt = "你是谁？我们刚才聊了什么？"  # 总结对话时的 prompt
 
-summarize_prompt = "你是谁？我们刚才聊了什么？" # 总结对话时的 prompt
-MODELS = ["gpt-3.5-turbo", "gpt-3.5-turbo-0301", "gpt-4","gpt-4-0314", "gpt-4-32k", "gpt-4-32k-0314"] # 可选的模型
-websearch_prompt = """Web search results:
+MODELS = [
+    "gpt-3.5-turbo",
+    "gpt-3.5-turbo-0301",
+    "gpt-4",
+    "gpt-4-0314",
+    "gpt-4-32k",
+    "gpt-4-32k-0314",
+]  # 可选的模型
+
+
+WEBSEARCH_PTOMPT_TEMPLATE = """\
+Web search results:
 
 {web_results}
 Current date: {current_date}
@@ -73,18 +52,29 @@ Instructions: Using the provided web search results, write a comprehensive reply
 Query: {query}
 Reply in 中文"""
 
-# 错误信息
-standard_error_msg = "☹️发生了错误：" # 错误信息的标准前缀
-error_retrieve_prompt = "请检查网络连接，或者API-Key是否有效。" # 获取对话时发生错误
-connection_timeout_prompt = "连接超时，无法获取对话。" # 连接超时
-read_timeout_prompt = "读取超时，无法获取对话。" # 读取超时
-proxy_error_prompt = "代理错误，无法获取对话。" # 代理错误
-ssl_error_prompt = "SSL错误，无法获取对话。" # SSL 错误
-no_apikey_msg = "API key长度不是51位，请检查是否输入正确。" # API key 长度不足 51 位
+PROMPT_TEMPLATE = """\
+Context information is below.
+---------------------
+{context_str}
+---------------------
+Current date: {current_date}.
+Using the provided context information, write a comprehensive reply to the given query.
+Make sure to cite results using [number] notation after the reference.
+If the provided context information refer to multiple subjects with the same name, write separate answers for each subject.
+Use prior knowledge only if the given context didn't provide enough information.
+Answer the question: {query_str}
+Reply in 中文
+"""
 
-max_token_streaming = 3500 # 流式对话时的最大 token 数
-timeout_streaming = 30 # 流式对话时的超时时间
-max_token_all = 3500 # 非流式对话时的最大 token 数
-timeout_all = 200 # 非流式对话时的超时时间
-enable_streaming_option = True  # 是否启用选择选择是否实时显示回答的勾选框
-HIDE_MY_KEY = False # 如果你想在UI中隐藏你的 API 密钥，将此值设置为 True
+REFINE_TEMPLATE = """\
+The original question is as follows: {query_str}
+We have provided an existing answer: {existing_answer}
+We have the opportunity to refine the existing answer
+(only if needed) with some more context below.
+------------
+{context_msg}
+------------
+Given the new context, refine the original answer to better
+Answer in the same language as the question, such as English, 中文, 日本語, Español, Français, or Deutsch.
+If the context isn't useful, return the original answer.
+"""

requirements.txt

@@ -6,3 +6,6 @@ socksio
 tqdm
 colorama
 duckduckgo_search
+Pygments
+llama_index
+langchain

utils.py

@@ -3,21 +3,16 @@ from __future__ import annotations
 from typing import TYPE_CHECKING, Any, Callable, Dict, List, Tuple, Type
 import logging
 import json
-import gradio as gr
-# import openai
 import os
-import traceback
-import requests
-# import markdown
+import datetime
+import hashlib
 import csv
-import mdtex2html
+
+import gradio as gr
 from pypinyin import lazy_pinyin
-from presets import *
 import tiktoken
-from tqdm import tqdm
-import colorama
-from duckduckgo_search import ddg
-import datetime
+
+from presets import *
 
 # logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] [%(filename)s:%(lineno)d] %(message)s")
 
@@ -28,30 +23,12 @@ if TYPE_CHECKING:
         headers: List[str]
         data: List[List[str | int | bool]]
 
+
 initial_prompt = "You are a helpful assistant."
 API_URL = "https://api.openai.com/v1/chat/completions"
 HISTORY_DIR = "history"
 TEMPLATES_DIR = "templates"
 
-def postprocess(
-        self, y: List[Tuple[str | None, str | None]]
-    ) -> List[Tuple[str | None, str | None]]:
-        """
-        Parameters:
-            y: List of tuples representing the message and response pairs. Each message and response should be a string, which may be in Markdown format.
-        Returns:
-            List of tuples representing the message and response. Each message and response will be a string of HTML.
-        """
-        if y is None:
-            return []
-        for i, (message, response) in enumerate(y):
-            y[i] = (
-                # None if message is None else markdown.markdown(message),
-                # None if response is None else markdown.markdown(response),
-                None if message is None else message,
-                None if response is None else mdtex2html.convert(response),
-            )
-        return y
 
 def count_token(message):
     encoding = tiktoken.get_encoding("cl100k_base")
@@ -59,251 +36,43 @@ def count_token(message):
     length = len(encoding.encode(input_str))
     return length
 
+
 def parse_text(text):
-    lines = text.split("\n")
-    lines = [line for line in lines if line != ""]
-    count = 0
-    for i, line in enumerate(lines):
-        if "```" in line:
-            count += 1
-            items = line.split('`')
-            if count % 2 == 1:
-                lines[i] = f'<pre><code class="language-{items[-1]}">'
-            else:
-                lines[i] = f'<br></code></pre>'
+    in_code_block = False
+    new_lines = []
+    for line in text.split("\n"):
+        if line.strip().startswith("```"):
+            in_code_block = not in_code_block
+        if in_code_block:
+            if line.strip() != "":
+                new_lines.append(line)
         else:
-            if i > 0:
-                if count % 2 == 1:
-                    line = line.replace("`", "\`")
-                    line = line.replace("<", "&lt;")
-                    line = line.replace(">", "&gt;")
-                    line = line.replace(" ", "&nbsp;")
-                    line = line.replace("*", "&ast;")
-                    line = line.replace("_", "&lowbar;")
-                    line = line.replace("-", "&#45;")
-                    line = line.replace(".", "&#46;")
-                    line = line.replace("!", "&#33;")
-                    line = line.replace("(", "&#40;")
-                    line = line.replace(")", "&#41;")
-                    line = line.replace("$", "&#36;")
-                lines[i] = "<br>"+line
-    text = "".join(lines)
+            new_lines.append(line)
+    if in_code_block:
+        new_lines.append("```")
+    text = "\n".join(new_lines)
     return text
 
+
 def construct_text(role, text):
     return {"role": role, "content": text}
 
+
 def construct_user(text):
     return construct_text("user", text)
 
+
 def construct_system(text):
     return construct_text("system", text)
 
+
 def construct_assistant(text):
     return construct_text("assistant", text)
 
+
 def construct_token_message(token, stream=False):
     return f"Token 计数: {token}"
 
-def get_response(openai_api_key, system_prompt, history, temperature, top_p, stream, selected_model):
-    headers = {
-        "Content-Type": "application/json",
-        "Authorization": f"Bearer {openai_api_key}"
-    }
-
-    history = [construct_system(system_prompt), *history]
-
-    payload = {
-        "model": selected_model,
-        "messages": history,  # [{"role": "user", "content": f"{inputs}"}],
-        "temperature": temperature,  # 1.0,
-        "top_p": top_p,  # 1.0,
-        "n": 1,
-        "stream": stream,
-        "presence_penalty": 0,
-        "frequency_penalty": 0,
-    }
-    if stream:
-        timeout = timeout_streaming
-    else:
-        timeout = timeout_all
-    response = requests.post(API_URL, headers=headers, json=payload, stream=True, timeout=timeout)
-    return response
-
-def stream_predict(openai_api_key, system_prompt, history, inputs, chatbot, all_token_counts, top_p, temperature, selected_model):
-    def get_return_value():
-        return chatbot, history, status_text, all_token_counts
-
-    logging.info("实时回答模式")
-    partial_words = ""
-    counter = 0
-    status_text = "开始实时传输回答……"
-    history.append(construct_user(inputs))
-    history.append(construct_assistant(""))
-    chatbot.append((parse_text(inputs), ""))
-    user_token_count = 0
-    if len(all_token_counts) == 0:
-        system_prompt_token_count = count_token(construct_system(system_prompt))
-        user_token_count = count_token(construct_user(inputs)) + system_prompt_token_count
-    else:
-        user_token_count = count_token(construct_user(inputs))
-    all_token_counts.append(user_token_count)
-    logging.info(f"输入token计数: {user_token_count}")
-    yield get_return_value()
-    try:
-        response = get_response(openai_api_key, system_prompt, history, temperature, top_p, True, selected_model)
-    except requests.exceptions.ConnectTimeout:
-        status_text = standard_error_msg + connection_timeout_prompt + error_retrieve_prompt
-        yield get_return_value()
-        return
-    except requests.exceptions.ReadTimeout:
-        status_text = standard_error_msg + read_timeout_prompt + error_retrieve_prompt
-        yield get_return_value()
-        return
-
-    yield get_return_value()
-    error_json_str = ""
-
-    for chunk in tqdm(response.iter_lines()):
-        if counter == 0:
-            counter += 1
-            continue
-        counter += 1
-        # check whether each line is non-empty
-        if chunk:
-            chunk = chunk.decode()
-            chunklength = len(chunk)
-            try:
-                chunk = json.loads(chunk[6:])
-            except json.JSONDecodeError:
-                logging.info(chunk)
-                error_json_str += chunk
-                status_text = f"JSON解析错误。请重置对话。收到的内容: {error_json_str}"
-                yield get_return_value()
-                continue
-            # decode each line as response data is in bytes
-            if chunklength > 6 and "delta" in chunk['choices'][0]:
-                finish_reason = chunk['choices'][0]['finish_reason']
-                status_text = construct_token_message(sum(all_token_counts), stream=True)
-                if finish_reason == "stop":
-                    yield get_return_value()
-                    break
-                try:
-                    partial_words = partial_words + chunk['choices'][0]["delta"]["content"]
-                except KeyError:
-                    status_text = standard_error_msg + "API回复中找不到内容。很可能是Token计数达到上限了。请重置对话。当前Token计数: " + str(sum(all_token_counts))
-                    yield get_return_value()
-                    break
-                history[-1] = construct_assistant(partial_words)
-                chatbot[-1] = (parse_text(inputs), parse_text(partial_words))
-                all_token_counts[-1] += 1
-                yield get_return_value()
-
-
-def predict_all(openai_api_key, system_prompt, history, inputs, chatbot, all_token_counts, top_p, temperature, selected_model):
-    logging.info("一次性回答模式")
-    history.append(construct_user(inputs))
-    history.append(construct_assistant(""))
-    chatbot.append((parse_text(inputs), ""))
-    all_token_counts.append(count_token(construct_user(inputs)))
-    try:
-        response = get_response(openai_api_key, system_prompt, history, temperature, top_p, False, selected_model)
-    except requests.exceptions.ConnectTimeout:
-        status_text = standard_error_msg + connection_timeout_prompt + error_retrieve_prompt
-        return chatbot, history, status_text, all_token_counts
-    except requests.exceptions.ProxyError:
-        status_text = standard_error_msg + proxy_error_prompt + error_retrieve_prompt
-        return chatbot, history, status_text, all_token_counts
-    except requests.exceptions.SSLError:
-        status_text = standard_error_msg + ssl_error_prompt + error_retrieve_prompt
-        return chatbot, history, status_text, all_token_counts
-    response = json.loads(response.text)
-    content = response["choices"][0]["message"]["content"]
-    history[-1] = construct_assistant(content)
-    chatbot[-1] = (parse_text(inputs), parse_text(content))
-    total_token_count = response["usage"]["total_tokens"]
-    all_token_counts[-1] = total_token_count - sum(all_token_counts)
-    status_text = construct_token_message(total_token_count)
-    return chatbot, history, status_text, all_token_counts
-
-
-def predict(openai_api_key, system_prompt, history, inputs, chatbot, all_token_counts, top_p, temperature, stream=False, selected_model = MODELS[0], use_websearch_checkbox = False, should_check_token_count = True):  # repetition_penalty, top_k
-    logging.info("输入为：" +colorama.Fore.BLUE + f"{inputs}" + colorama.Style.RESET_ALL)
-    if use_websearch_checkbox:
-        results = ddg(inputs, max_results=3)
-        web_results = []
-        for idx, result in enumerate(results):
-            logging.info(f"搜索结果{idx + 1}：{result}")
-            web_results.append(f'[{idx+1}]"{result["body"]}"\nURL: {result["href"]}')
-        web_results = "\n\n".join(web_results)
-        today = datetime.datetime.today().strftime("%Y-%m-%d")
-        inputs = websearch_prompt.replace("{current_date}", today).replace("{query}", inputs).replace("{web_results}", web_results)
-    if len(openai_api_key) != 51:
-        status_text = standard_error_msg + no_apikey_msg
-        logging.info(status_text)
-        chatbot.append((parse_text(inputs), ""))
-        if len(history) == 0:
-            history.append(construct_user(inputs))
-            history.append("")
-            all_token_counts.append(0)
-        else:
-            history[-2] = construct_user(inputs)
-        yield chatbot, history, status_text, all_token_counts
-        return
-    if stream:
-        yield chatbot, history, "开始生成回答……", all_token_counts
-    if stream:
-        logging.info("使用流式传输")
-        iter = stream_predict(openai_api_key, system_prompt, history, inputs, chatbot, all_token_counts, top_p, temperature, selected_model)
-        for chatbot, history, status_text, all_token_counts in iter:
-            yield chatbot, history, status_text, all_token_counts
-    else:
-        logging.info("不使用流式传输")
-        chatbot, history, status_text, all_token_counts = predict_all(openai_api_key, system_prompt, history, inputs, chatbot, all_token_counts, top_p, temperature, selected_model)
-        yield chatbot, history, status_text, all_token_counts
-    logging.info(f"传输完毕。当前token计数为{all_token_counts}")
-    if len(history) > 1 and history[-1]['content'] != inputs:
-        logging.info("回答为：" +colorama.Fore.BLUE + f"{history[-1]['content']}" + colorama.Style.RESET_ALL)
-    if stream:
-        max_token = max_token_streaming
-    else:
-        max_token = max_token_all
-    if sum(all_token_counts) > max_token and should_check_token_count:
-        status_text = f"精简token中{all_token_counts}/{max_token}"
-        logging.info(status_text)
-        yield chatbot, history, status_text, all_token_counts
-        iter = reduce_token_size(openai_api_key, system_prompt, history, chatbot, all_token_counts, top_p, temperature, stream=False, selected_model=selected_model, hidden=True)
-        for chatbot, history, status_text, all_token_counts in iter:
-            status_text = f"Token 达到上限，已自动降低Token计数至 {status_text}"
-            yield chatbot, history, status_text, all_token_counts
-
-
-def retry(openai_api_key, system_prompt, history, chatbot, token_count, top_p, temperature, stream=False, selected_model = MODELS[0]):
-    logging.info("重试中……")
-    if len(history) == 0:
-        yield chatbot, history, f"{standard_error_msg}上下文是空的", token_count
-        return
-    history.pop()
-    inputs = history.pop()["content"]
-    token_count.pop()
-    iter = predict(openai_api_key, system_prompt, history, inputs, chatbot, token_count, top_p, temperature, stream=stream, selected_model=selected_model)
-    logging.info("重试完毕")
-    for x in iter:
-        yield x
-
-
-def reduce_token_size(openai_api_key, system_prompt, history, chatbot, token_count, top_p, temperature, stream=False, selected_model = MODELS[0], hidden=False):
-    logging.info("开始减少token数量……")
-    iter = predict(openai_api_key, system_prompt, history, summarize_prompt, chatbot, token_count, top_p, temperature, stream=stream, selected_model = selected_model, should_check_token_count=False)
-    logging.info(f"chatbot: {chatbot}")
-    for chatbot, history, status_text, previous_token_count in iter:
-        history = history[-2:]
-        token_count = previous_token_count[-1:]
-        if hidden:
-            chatbot.pop()
-        yield chatbot, history, construct_token_message(sum(token_count), stream=stream), token_count
-    logging.info("减少token数量完毕")
-
 
 def delete_last_conversation(chatbot, history, previous_token_count):
     if len(chatbot) > 0 and standard_error_msg in chatbot[-1][1]:
@@ -320,7 +89,12 @@ def delete_last_conversation(chatbot, history, previous_token_count):
     if len(previous_token_count) > 0:
         logging.info("删除了一组对话的token计数记录")
         previous_token_count.pop()
-    return chatbot, history, previous_token_count, construct_token_message(sum(previous_token_count))
+    return (
+        chatbot,
+        history,
+        previous_token_count,
+        construct_token_message(sum(previous_token_count)),
+    )
 
 
 def save_file(filename, system, history, chatbot):
@@ -340,6 +114,7 @@ def save_file(filename, system, history, chatbot):
     logging.info("保存对话历史完毕")
     return os.path.join(HISTORY_DIR, filename)
 
+
 def save_chat_history(filename, system, history, chatbot):
     if filename == "":
         return
@@ -347,6 +122,7 @@ def save_chat_history(filename, system, history, chatbot):
         filename += ".json"
     return save_file(filename, system, history, chatbot)
 
+
 def export_markdown(filename, system, history, chatbot):
     if filename == "":
         return
@@ -382,9 +158,11 @@ def load_chat_history(filename, system, history, chatbot):
         logging.info("没有找到对话历史文件，不执行任何操作")
         return filename, system, history, chatbot
 
+
 def sorted_by_pinyin(list):
     return sorted(list, key=lambda char: lazy_pinyin(char)[0][0])
 
+
 def get_file_names(dir, plain=False, filetypes=[".json"]):
     logging.info(f"获取文件名列表，目录为{dir}，文件类型为{filetypes}，是否为纯文本列表{plain}")
     files = []
@@ -401,10 +179,12 @@ def get_file_names(dir, plain=False, filetypes=[".json"]):
     else:
         return gr.Dropdown.update(choices=files)
 
+
 def get_history_names(plain=False):
     logging.info("获取历史记录文件名列表")
     return get_file_names(HISTORY_DIR, plain)
 
+
 def load_template(filename, mode=0):
     logging.info(f"加载模板文件{filename}，模式为{mode}（0为返回字典和下拉菜单，1为返回下拉菜单，2为返回字典）")
     lines = []
@@ -414,22 +194,28 @@ def load_template(filename, mode=0):
             lines = json.load(f)
         lines = [[i["act"], i["prompt"]] for i in lines]
     else:
-        with open(os.path.join(TEMPLATES_DIR, filename), "r", encoding="utf8") as csvfile:
+        with open(
+            os.path.join(TEMPLATES_DIR, filename), "r", encoding="utf8"
+        ) as csvfile:
             reader = csv.reader(csvfile)
             lines = list(reader)
         lines = lines[1:]
     if mode == 1:
         return sorted_by_pinyin([row[0] for row in lines])
     elif mode == 2:
-        return {row[0]:row[1] for row in lines}
+        return {row[0]: row[1] for row in lines}
     else:
         choices = sorted_by_pinyin([row[0] for row in lines])
-        return {row[0]:row[1] for row in lines}, gr.Dropdown.update(choices=choices, value=choices[0])
+        return {row[0]: row[1] for row in lines}, gr.Dropdown.update(
+            choices=choices, value=choices[0]
+        )
+
 
 def get_template_names(plain=False):
     logging.info("获取模板文件名列表")
     return get_file_names(TEMPLATES_DIR, plain, filetypes=[".csv", "json"])
 
+
 def get_template_content(templates, selection, original_system_prompt):
     logging.info(f"应用模板中，选择为{selection}，原始系统提示为{original_system_prompt}")
     try:
@@ -437,9 +223,62 @@ def get_template_content(templates, selection, original_system_prompt):
     except:
         return original_system_prompt
 
+
 def reset_state():
     logging.info("重置状态")
     return [], [], [], construct_token_message(0)
 
+
 def reset_textbox():
-    return gr.update(value='')
+    return gr.update(value="")
+
+
+def reset_default():
+    global API_URL
+    API_URL = "https://api.openai.com/v1/chat/completions"
+    os.environ.pop("HTTPS_PROXY", None)
+    os.environ.pop("https_proxy", None)
+    return gr.update(value=API_URL), gr.update(value=""), "API URL 和代理已重置"
+
+
+def change_api_url(url):
+    global API_URL
+    API_URL = url
+    msg = f"API地址更改为了{url}"
+    logging.info(msg)
+    return msg
+
+
+def change_proxy(proxy):
+    os.environ["HTTPS_PROXY"] = proxy
+    msg = f"代理更改为了{proxy}"
+    logging.info(msg)
+    return msg
+
+
+def hide_middle_chars(s):
+    if len(s) <= 8:
+        return s
+    else:
+        head = s[:4]
+        tail = s[-4:]
+        hidden = "*" * (len(s) - 8)
+        return head + hidden + tail
+
+
+def submit_key(key):
+    key = key.strip()
+    msg = f"API密钥更改为了{hide_middle_chars(key)}"
+    logging.info(msg)
+    return key, msg
+
+
+def sha1sum(filename):
+    sha1 = hashlib.sha1()
+    sha1.update(filename.encode("utf-8"))
+    return sha1.hexdigest()
+
+
+def replace_today(prompt):
+    today = datetime.datetime.today().strftime("%Y-%m-%d")
+    return prompt.replace("{current_date}", today)