提高协程细粒度

docutranslate/app.py

@@ -141,14 +141,14 @@ async def _perform_translation(task_id: str, params: Dict[str, Any], file_conten
             refine=params['refine_markdown'], save=False
         )
         md_content = ft.export_to_markdown()
-        md_zip_content = ft.export_to_unembed_markdown()
+        md_zip_content = await ft.export_to_unembed_markdown_async()
         try:
             await httpx_client.head("https://s4.zstatic.net/ajax/libs/KaTeX/0.16.9/contrib/auto-render.min.js",
                                     timeout=3)
-            html_content = ft.export_to_html(title=task_state["original_filename_stem"], cdn=True)
+            html_content = await ft.export_to_html_async(title=task_state["original_filename_stem"], cdn=True)
         except (httpx.TimeoutException, httpx.RequestError):
             task_logger.info("CDN连接失败，使用本地JS进行渲染。")
-            html_content = ft.export_to_html(title=task_state["original_filename_stem"], cdn=False)
+            html_content = await ft.export_to_html_async(title=task_state["original_filename_stem"], cdn=False)
 
         end_time = time.time()
         duration = end_time - task_state["task_start_time"]

docutranslate/converter/converter_mineru.py

@@ -120,7 +120,7 @@ class ConverterMineru(Converter):
         time1 = time.time()
         batch_id = await self.upload_async(document)
         file_url = await self.get_file_url_async(batch_id)
-        result = await get_md_from_zip_url_with_inline_images_async(file_url)
+        result = await get_md_from_zip_url_with_inline_images_async(zip_url=file_url)
         self.logger.info(f"已转换为markdown，耗时{time.time() - time1}秒")
         return result
 
@@ -194,7 +194,7 @@ async def get_md_from_zip_url_with_inline_images_async(
         response = await client_async.get(zip_url)  # 增加超时
         response.raise_for_status()
         print("ZIP文件下载完成。")
-        return await asyncio.to_thread(embed_inline_image_from_zip(response.content, filename_in_zip=filename_in_zip, encoding=encoding))
+        return await asyncio.to_thread(embed_inline_image_from_zip,response.content, filename_in_zip=filename_in_zip, encoding=encoding)
 
 
     except httpx.HTTPStatusError as e:

docutranslate/translater.py

@@ -1,3 +1,4 @@
+import asyncio
 import html
 import io
 import logging
@@ -330,6 +331,9 @@ class FileTranslater:
         self._markdown_format()
         return unembed_base64_images_to_zip(self.markdown, markdown_name=filename.name)
 
+    async def export_to_unembed_markdown_async(self, filename: str | Path | None = None) -> bytes:
+        return await asyncio.to_thread(self.export_to_unembed_markdown,filename)
+
     def save_as_html(self, filename: str | Path | None = None, output_dir: str | Path = "./output"):
         if isinstance(filename, str):
             filename = Path(filename)
@@ -403,7 +407,8 @@ class FileTranslater:
             mermaid=mermaid,
         )
         return render
-
+    async def export_to_html_async(self, title="title", cdn=True):
+        return await asyncio.to_thread(self.export_to_html,title,cdn)
     def translate_file(self, file_path: Path | str | None = None, to_lang="中文", output_dir="./output",
                        formula=True,
                        code=True, output_format: Literal["markdown", "html"] = "markdown", refine=False,