Update utils.py

utils.py

@@ -1,5 +1,6 @@
 # -*- coding:utf-8 -*-
 from __future__ import annotations
+from typing import TYPE_CHECKING, Any, Callable, Dict, List, Tuple, Type
 import logging
 import json
 import os
@@ -10,12 +11,10 @@ import gradio as gr
 from pypinyin import lazy_pinyin
 import tiktoken
 import mdtex2html
-from typing import TYPE_CHECKING, Any, Callable, Dict, List, Tuple, Type
 from markdown import markdown
 from pygments import highlight
 from pygments.lexers import get_lexer_by_name
 from pygments.formatters import HtmlFormatter
-
     
 use_websearch_checkbox=False
 use_streaming_checkbox=True
@@ -60,25 +59,6 @@ The default model role of the app is the original assistant of ChatGPT, but you
         
 MODELS = ["gpt-3.5-turbo", "gpt-3.5-turbo-0301",]  
 
-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
 
 if TYPE_CHECKING:
     from typing import TypedDict
@@ -94,6 +74,7 @@ def count_token(message):
     length = len(encoding.encode(input_str))
     return length
 
+
 def markdown_to_html_with_syntax_highlight(md_str):
     def replacer(match):
         lang = match.group(1) or "text"
@@ -115,6 +96,7 @@ def markdown_to_html_with_syntax_highlight(md_str):
     html_str = markdown(md_str)
     return html_str
 
+
 def normalize_markdown(md_text: str) -> str:
     lines = md_text.split("\n")
     normalized_lines = []
@@ -138,6 +120,24 @@ def normalize_markdown(md_text: str) -> str:
 
     return "\n".join(normalized_lines)
 
+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 or y == []:
+        return []
+    tag_regex = re.compile(r"^<\w+>[^<]+</\w+>")
+    if tag_regex.search(y[-1][1]):
+        y[-1] = (convert_user(y[-1][0]), y[-1][1])
+    else:
+        y[-1] = (convert_user(y[-1][0]), convert_mdtext(y[-1][1]))
+    return y
+
 def convert_mdtext(md_text):
     code_block_pattern = re.compile(r"```(.*?)(?:```|$)", re.DOTALL)
     inline_code_pattern = re.compile(r"`(.*?)`", re.DOTALL)
@@ -153,8 +153,6 @@ def convert_mdtext(md_text):
             else:
                 result.append(mdtex2html.convert(non_code, extensions=["tables"]))
         if code.strip():
-            # _, code = detect_language(code)  # 暂时去除代码高亮功能，因为在大段代码的情况下会出现问题
-            # code = code.replace("\n\n", "\n") # 暂时去除代码中的空行，因为在大段代码的情况下会出现问题
             code = f"```{code}\n\n```"
             code = markdown_to_html_with_syntax_highlight(code)
             result.append(code)
@@ -184,6 +182,7 @@ def construct_assistant(text):
 def construct_token_message(token, stream=False):
     return f"Token count: {token}"
 
+
 def save_file(filename, system, history, chatbot):
     logging.info("saving......")
     os.makedirs(HISTORY_DIR, exist_ok=True)
@@ -198,8 +197,10 @@ def save_file(filename, system, history, chatbot):
             md_s += f"\n{data['role']}: \n- {data['content']} \n"
         with open(os.path.join(HISTORY_DIR, filename), "w", encoding="utf8") as f:
             f.write(md_s)
+    # 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
@@ -207,6 +208,7 @@ 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
@@ -214,7 +216,9 @@ def export_markdown(filename, system, history, chatbot):
         filename += ".md"
     return save_file(filename, system, history, chatbot)
 
+
 def load_chat_history(filename, system, history, chatbot):
+    # logging.info("加载对话历史中……")
     if type(filename) != str:
         filename = filename.name
     try:
@@ -222,6 +226,7 @@ def load_chat_history(filename, system, history, chatbot):
             json_s = json.load(f)
         try:
             if type(json_s["history"][0]) == str:
+                # logging.info("历史记录格式为旧版，正在转换……")
                 new_history = []
                 for index, item in enumerate(json_s["history"]):
                     if index % 2 == 0:
@@ -231,14 +236,20 @@ def load_chat_history(filename, system, history, chatbot):
                 json_s["history"] = new_history
                 logging.info(new_history)
         except:
+            # 没有对话历史
             pass
+        # logging.info("加载对话历史完毕")
         return filename, json_s["system"], json_s["history"], json_s["chatbot"]
     except FileNotFoundError:
+        # 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 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 = []
     logging.info("Loading template...")
+    # filename='中文Prompts.json'
     if filename.endswith(".json"):
         with open(os.path.join(TEMPLATES_DIR, filename), "r", encoding="utf8") as f:
             lines = json.load(f)
@@ -261,6 +272,7 @@ def load_template(filename, mode=0):
 def sorted_by_pinyin(list):
     return sorted(list, key=lambda char: lazy_pinyin(char)[0][0])
 
+
 def get_template_content(templates, selection, original_system_prompt):
     logging.info(f"Prompt:  {selection}")
     try:
@@ -268,13 +280,16 @@ def get_template_content(templates, selection, original_system_prompt):
     except:
         return original_system_prompt
 
+
 def reset_state():
     logging.info("Reset")
     return [], [], [], construct_token_message(0)
 
+
 def reset_textbox():
     return gr.update(value="")
 
+
 def hide_middle_chars(s):
     if len(s) <= 8:
         return s
@@ -289,3 +304,5 @@ def submit_key(key):
     msg = f"API-Key: {hide_middle_chars(key)}"
     logging.info(msg)
     return key, msg
+
+