这是翻译后的HTML内容...
" + } + }, + "markdown_zip_base64": { + "summary": "ZIP 内容 (Base64)", + "value": { + "file_type": "markdown_zip", + "filename": "my_doc_translated.zip", + "content": "UEsDBBQAAAAIA... (base64-encoded string)" + } + } + } + } + } + }, + 404: { + "description": "资源未找到。可能的原因包括：任务ID不存在、任务结果尚未就绪、或请求了无效的文件类型。", + "content": {"application/json": {"example": {"detail": "内容尚未准备好。"}}} + }, + } +) +async def service_download_content( + task_id: str = FastApiPath(..., description="已完成任务的ID", example="task-b2865b93"), + file_type: FileType = FastApiPath(..., description="要获取内容的文件类型。", example="html") +): + """根据任务ID和文件类型，以JSON格式返回内容。zip文件会进行Base64编码。""" + task_state = tasks_state.get(task_id) + if not task_state: + raise HTTPException(status_code=404, detail=f"找不到任务ID '{task_id}'。") + + if not task_state["download_ready"]: + raise HTTPException(status_code=404, detail="内容尚未准备好。") + + content_map = { + "markdown": (task_state.get("markdown_content"), f"{task_state['original_filename_stem']}_translated.md"), + "markdown_zip": (task_state.get("markdown_zip_content"), + f"{task_state['original_filename_stem']}_translated.zip"), + "html": (task_state.get("html_content"), f"{task_state['original_filename_stem']}_translated.html"), + } + + raw_content, filename = content_map.get(file_type, (None, None)) + + if raw_content is None: + raise HTTPException(status_code=404, detail=f"'{file_type}' 类型的内容不可用或生成失败。") + + # 如果内容是字节串 (zip)，则进行Base64编码；否则直接使用字符串。 + final_content = base64.b64encode(raw_content).decode('utf-8') if isinstance(raw_content, bytes) else raw_content + + return JSONResponse(content={ + "file_type": file_type, + "filename": filename, + "content": final_content + }) + + @service_router.get( "/engin-list", summary="获取可用解析引擎", @@ -876,10 +964,13 @@ async def temp_translate( api_key: str = Body(..., description="LLM API的密钥。", example="sk-xxxxxxxxxx"), model_id: str = Body(..., description="使用的模型ID。", example="gpt-4-turbo"), mineru_token: str = Body(..., description="Mineru引擎的Token。"), - file_name: str = Body(..., description="文件名，用以判断文件类型。当后缀为txt时该接口返回普通文本，为其他后缀时返回翻译后的markdown文本", examples=["test.txt","test.md","test.pdf"]), + file_name: str = Body(..., + description="文件名，用以判断文件类型。当后缀为txt时该接口返回普通文本，为其他后缀时返回翻译后的markdown文本", + examples=["test.txt", "test.md", "test.pdf"]), file_content: str = Body(..., description="文件内容，可以是纯文本或Base64编码的字符串。"), - to_lang: str = Body("中文", description="目标语言。",examples=["中文","英文","English"]), - concurrent: int = Body(30, description="ai翻译请求并发数") + to_lang: str = Body("中文", description="目标语言。", examples=["中文", "英文", "English"]), + concurrent: int = Body(30, description="ai翻译请求并发数"), + custom_prompt_translate: str | None = Body(None, description="翻译自定义提示词") ): """一个用于快速测试的同步翻译接口。""" @@ -894,12 +985,13 @@ async def temp_translate( key=api_key, model_id=model_id, mineru_token=mineru_token, - concurrent=concurrent + concurrent=concurrent, ) try: decoded_content = base64.b64decode(file_content) if is_base64(file_content) else file_content.encode('utf-8') - await ft.translate_bytes_async(name=file_name, file=decoded_content, to_lang=to_lang, save=False) + await ft.translate_bytes_async(name=file_name, file=decoded_content, to_lang=to_lang, save=False, + custom_prompt_translate=custom_prompt_translate) return {"success": True, "content": ft.export_to_markdown()} except Exception as e: print(f"翻译出现错误：{e.__repr__()}") @@ -931,4 +1023,4 @@ def run_app(port: int | None = None): if __name__ == "__main__": - run_app() \ No newline at end of file + run_app() diff --git a/docutranslate/cacher/document_cacher.py b/docutranslate/cacher/document_cacher.py index 6e78983..5e934c8 100644 --- a/docutranslate/cacher/document_cacher.py +++ b/docutranslate/cacher/document_cacher.py @@ -1,12 +1,13 @@ import os +from collections import OrderedDict from docutranslate.converter import Document -# DOCUTRANSLATE_CHACHE_NUM=os.getenv("DOCUTRANSLATE_CHACHE_NUM") +CACHE_NUM=os.getenv("DOCUTRANSLATE_CACHE_NUM",default=10) class DocumentCacher: def __init__(self): - self.cache_dict:dict[str:str] = {} + self.cache_dict = OrderedDict() @staticmethod def _get_hashcode(document: Document, formula: bool, code: bool, convert_engin: str) -> str: obj = (document.suffix, document.filebytes, formula, code, convert_engin) @@ -17,6 +18,8 @@ class DocumentCacher: def cache_result(self, result: str, document: Document, formula: bool, code: bool, convert_engin: str): hash_code = self._get_hashcode(document, formula, code, convert_engin) + if len(self.cache_dict)>=CACHE_NUM: + self.cache_dict.popitem(last=False) self.cache_dict[hash_code] = result return result