var fichubToWget=url=>fetch(`https://fichub.net/api/v0/epub?q=${encodeURIComponent(url)}`).then(r => r.json()).then(data=>{
  var fullEpubUrl=`https://fichub.net${data.epub_url}`;
  var baseFileName=new URL(fullEpubUrl).pathname.split('/').pop().split('-').slice(0,-1).join('-')+'.epub';
  console.log(`wget "${fullEpubUrl}" -O "${baseFileName}"`);
});
var url=''; // 填入阅读链接
fichubToWget(url);

{"user": "{text}", "system": "你是作家余光中, 你正在给一本文学书的一段话(有时是章节名)翻译为中文。作为忠实的译者, 你绝对不会尝试续写, 也绝对不会省略。无法理解的词语和人名地名将保持原样。如果有专业术语和专有名词，第一次出现时需用括号附上原文。请直接返回翻译结果。"}

diff --git a/book_maker/cli.py b/book_maker/cli.py
index 57ac7b7..4b23185 100644
--- a/book_maker/cli.py
+++ b/book_maker/cli.py
@@ -35,7 +35,7 @@ def parse_prompt_arg(prompt_arg):
     else:
         raise FileNotFoundError(f"{prompt_arg} not found")
 
-    if prompt is None or any(c not in prompt["user"] for c in ["{text}", "{language}"]):
+    if prompt is None: #or any(c not in prompt["user"] for c in ["{text}", "{language}"]):
         raise ValueError("prompt must contain `{text}` and `{language}`")
 
     if "user" not in prompt:
diff --git a/book_maker/translator/chatgptapi_translator.py b/book_maker/translator/chatgptapi_translator.py
index 22f2796..7e8b12a 100644
--- a/book_maker/translator/chatgptapi_translator.py
+++ b/book_maker/translator/chatgptapi_translator.py
@@ -78,7 +78,7 @@ class ChatGPTAPI(Base):
     def create_chat_completion(self, text):
         content = self.prompt_template.format(
             text=text, language=self.language, crlf="\n"
-        )
+        ) if "{language}" in self.prompt_template else self.prompt_template.format(text=text, crlf="\n")
         sys_content = self.system_content or self.prompt_sys_msg.format(crlf="\n")
         messages = [
             {"role": "system", "content": sys_content},
@@ -111,10 +111,10 @@ class ChatGPTAPI(Base):
         start_time = time.time()
         # todo: Determine whether to print according to the cli option
         if needprint:
-            print(re.sub("\n{3,}", "\n\n", text))
+            print(re.sub("\n{3,}", "\n\n", text).replace('[/','').replace(r'[\\',''))
 
         attempt_count = 0
-        max_attempts = 3
+        max_attempts = 12
         t_text = ""
 
         while attempt_count < max_attempts:
@@ -126,7 +126,7 @@ class ChatGPTAPI(Base):
                 # 1. openai server error or own network interruption, sleep for a fixed time
                 # 2. an apikey has no money or reach limit, don`t sleep, just replace it with another apikey
                 # 3. all apikey reach limit, then use current sleep
-                sleep_time = int(60 / self.key_len)
+                sleep_time = int(10 / self.key_len)
                 print(e, f"will sleep {sleep_time} seconds")
                 time.sleep(sleep_time)
                 attempt_count += 1
@@ -134,12 +134,17 @@ class ChatGPTAPI(Base):
                     print(f"Get {attempt_count} consecutive exceptions")
                     raise
             except Exception as e:
-                print(str(e))
-                return
+                sleep_time = 10+20*attempt_count
+                print(str(e), f"will sleep {sleep_time} seconds")
+                time.sleep(sleep_time)
+                attempt_count += 1
+                if attempt_count == max_attempts:
+                    print(f"Get {attempt_count} consecutive exceptions")
+                    return
 
         # todo: Determine whether to print according to the cli option
         if needprint:
-            print("[bold green]" + re.sub("\n{3,}", "\n\n", t_text) + "[/bold green]")
+            print("[bold green]" + re.sub("\n{3,}", "\n\n", t_text).replace('[/','').replace(r'[\\','') + "[/bold green]")
 
         time.time() - start_time
         # print(f"translation time: {elapsed_time:.1f}s")
diff --git a/book_maker/translator/gemini_translator.py b/book_maker/translator/gemini_translator.py
index 0dc43a6..315e9f7 100644
--- a/book_maker/translator/gemini_translator.py
+++ b/book_maker/translator/gemini_translator.py
@@ -6,6 +6,8 @@ from google.generativeai.types.generation_types import (
     StopCandidateException,
     BlockedPromptException,
 )
+from google.generativeai.types import HarmCategory, HarmBlockThreshold
+
 from rich import print
 
 from .base_translator import Base
@@ -14,22 +16,29 @@ generation_config = {
     "temperature": 0.7,
     "top_p": 1,
     "top_k": 1,
-    "max_output_tokens": 2048,
+    "max_output_tokens": 8192,
 }
 
 safety_settings = [
-    {"category": "HARM_CATEGORY_HARASSMENT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
-    {"category": "HARM_CATEGORY_HATE_SPEECH", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
+    {"category": "HARM_CATEGORY_HARASSMENT", "threshold": "BLOCK_NONE"},
+    {"category": "HARM_CATEGORY_HATE_SPEECH", "threshold": "BLOCK_NONE"},
     {
         "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT",
-        "threshold": "BLOCK_MEDIUM_AND_ABOVE",
+        "threshold": "BLOCK_NONE",
     },
     {
         "category": "HARM_CATEGORY_DANGEROUS_CONTENT",
-        "threshold": "BLOCK_MEDIUM_AND_ABOVE",
+        "threshold": "BLOCK_NONE",
     },
 ]
 
+safety_settings2={
+        HarmCategory.HARM_CATEGORY_HATE_SPEECH: HarmBlockThreshold.BLOCK_NONE,
+        HarmCategory.HARM_CATEGORY_HARASSMENT: HarmBlockThreshold.BLOCK_NONE,
+        HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT: HarmBlockThreshold.BLOCK_NONE,
+        HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT: HarmBlockThreshold.BLOCK_NONE,
+    }
+
 
 class Gemini(Base):
     """
@@ -42,9 +51,9 @@ class Gemini(Base):
         genai.configure(api_key=key)
         super().__init__(key, language)
         model = genai.GenerativeModel(
-            model_name="gemini-pro",
+            model_name="gemini-1.5-flash-latest",
             generation_config=generation_config,
-            safety_settings=safety_settings,
+            safety_settings=safety_settings2,
         )
         self.convo = model.start_chat()
 
@@ -85,7 +94,7 @@ class Gemini(Base):
 
         print("[bold green]" + re.sub("\n{3,}", "\n\n", t_text) + "[/bold green]")
         # for limit
-        time.sleep(0.5)
+        time.sleep(10)
         if num:
             t_text = str(num) + "\n" + t_text
         return t_text

#!/bin/bash

# 检查是否提供了参数
if [ -z "$1" ]; then
  echo "Usage: $0 <search_pattern>"
  exit 1
fi

# 执行搜索命令
echo "Files and directories in the current directory matching '$1':"
ls *$1*

echo "\nFiles and directories in subdirectories matching '$1':"
ls */*$1*

echo "\nFiles and directories in all subdirectories (recursive) matching '$1':"
ls **/*$1*

echo "\nFile contents matching '$1' (using ripgrep):"
rg "$1"

#!/bin/bash
find . -name "*.epub" -print0 | while IFS= read -r -d $'\0' file; do

  # Check if the filename already contains "_bilingual" or "temp.epub"
  if [[ "$file" == *_bilingual.epub ]] || [[ "$file" == *temp.epub ]]; then
    continue  # Skip to the next iteration of the loop
  fi

  # Extract the directory and base filename without the extension
  directory=$(dirname "$file")
  base_filename=$(basename "$file" .epub)

  # Construct the expected bilingual filename with directory
  bilingual_filename="${directory}/${base_filename}_bilingual.epub"

  # Check if the bilingual file exists
  if [ ! -f "${bilingual_filename}" ]; then
    # Print the filename of the unprocessed epub file
    echo "$file"
  fi
done

#!/bin/bash

# 获取当前目录下所有的 xxx.epub 文件
for file in *_bilingual.epub; do
    # 去除 '_bilingual.epub' 后缀以获取基本文件名
    base_name=${file%_bilingual.epub}

    # 检查对应的 xxx.epub 文件是否存在
    if [ -e "${base_name}.epub" ]; then
        # 获取两个文件的大小
        size_bilingual=$(stat -c%s "${base_name}_bilingual.epub")
        size_single=$(stat -c%s "${base_name}.epub")

        # 比较文件大小
        if [ $size_bilingual -lt $size_single ]; then
            echo "${base_name}_bilingual.epub"
            ls -l "${base_name}"*
        fi
    fi
done

#!/bin/bash

rm .*bin

# Find all potential file groups
for original in *.epub; do
  bilingual="${original%.*}_bilingual.epub"
  temp="${original%.*}_bilingual_temp.epub"

  # Check if all three files exist
  if [[ -f "$original" && -f "$bilingual" && -f "$temp" ]]; then
    # Get file sizes
    original_size=$(stat -c%s "$original")
    bilingual_size=$(stat -c%s "$bilingual")
    temp_size=$(stat -c%s "$temp")

    # Check if temp file is smaller than both original and bilingual
    if (( temp_size < bilingual_size )); then
      # Print file details
      echo "Found a matching file group:"
      ls -lh "$original"
      ls -lh "$bilingual"
      ls -lh "$temp"

      # Ask for confirmation before deleting
      read -p "Delete $temp? (y/n) " confirm
      if [[ "$confirm" == "y" ]]; then
        rm "$temp"
        echo "$temp deleted."
      else
        echo "$temp not deleted."
      fi
      echo "" # Add an empty line for better readability
    fi
  fi
done

echo "Scan complete."