重构代码，新增了MarkdownBasedManager和TXTManager实现

docutranslate/translater/base.py

@@ -0,0 +1,16 @@
+from dataclasses import dataclass
+from logging import Logger
+
+
+@dataclass
+class AiTranslateConfig:
+    base_url: str
+    api_key: str
+    model_id: str
+    to_lang: str
+    custom_prompt: str | None = None
+    temperature: float = 0.7
+    timeout: int = 2000
+    chunk_size: int = 3000
+    concurrent: int = 30
+    logger: Logger | None = None

docutranslate/translater/interfaces.py

@@ -0,0 +1,21 @@
+from typing import runtime_checkable, Protocol, TypeVar
+
+from docutranslate.agents import Agent
+from docutranslate.ir.document import Document
+
+T=TypeVar('T',bound=Document)
+V=TypeVar('V',bound=Agent)
+
+@runtime_checkable
+class Translator(Protocol[T,V]):
+    """
+    翻译中间文本（原地替换），Translator不做格式转换
+    """
+    def translate(self, document:T) -> Document:
+        ...
+
+    async def translate_async(self, document: T) -> Document:
+        ...
+
+    def log(self,info:str):
+        ...

docutranslate/translater/md_translator.py

@@ -0,0 +1,70 @@
+import asyncio
+from dataclasses import dataclass
+from logging import Logger
+from typing import Self
+
+from docutranslate.agents import MDTranslateAgent
+from docutranslate.document_context.md_mask_context import MDMaskUrisContext
+from docutranslate.ir.markdown_document import MarkdownDocument
+from docutranslate.logger import global_logger
+from docutranslate.translater.base import AiTranslateConfig
+from docutranslate.translater.interfaces import Translator
+from docutranslate.utils.markdown_splitter import split_markdown_text, join_markdown_texts
+from docutranslate.utils.markdown_utils import clean_markdown_math_block
+
+
+@dataclass
+class MDTranslateConfig(AiTranslateConfig):
+    ...
+
+
+
+class MDTranslator(Translator):
+    def __init__(self, config: MDTranslateConfig):
+        self.logger = config.logger or global_logger
+        self.chunk_size = config.chunk_size
+        self.translate_agent = MDTranslateAgent(custom_prompt=config.custom_prompt,
+                                                to_lang=config.to_lang,
+                                                baseurl=config.base_url,
+                                                key=config.api_key,
+                                                model_id=config.model_id,
+                                                system_prompt=None,
+                                                temperature=config.temperature,
+                                                max_concurrent=config.concurrent,
+                                                timeout=config.timeout,
+                                                logger=self.logger)
+
+    def translate(self, document: MarkdownDocument) -> Self:
+        self.logger.info("正在翻译markdown")
+        with MDMaskUrisContext(document):
+            chunks: list[str] = split_markdown_text(document.content.decode(), self.chunk_size)
+            self.logger.info(f"markdown分为{len(chunks)}块")
+            result: list[str] = self.translate_agent.send_prompts(chunks)
+            content = join_markdown_texts(result)
+            # 做一些加强鲁棒性的操作
+            content = content.replace(r'\（', r'\(')
+            content = content.replace(r'\）', r'\)')
+            content = clean_markdown_math_block(content)
+
+            document.content = content.encode()
+        self.logger.info("翻译完成")
+        return self
+
+    async def translate_async(self, document: MarkdownDocument) -> Self:
+        self.logger.info("正在翻译markdown")
+        with MDMaskUrisContext(document):
+            chunks: list[str] = split_markdown_text(document.content.decode(), self.chunk_size)
+            self.logger.info(f"markdown分为{len(chunks)}块")
+            result: list[str] = await self.translate_agent.send_prompts_async(chunks)
+
+            def run():
+                content = join_markdown_texts(result)
+                # 做一些加强鲁棒性的操作
+                content = content.replace(r'\（', r'\(')
+                content = content.replace(r'\）', r'\)')
+                content = clean_markdown_math_block(content)
+                document.content = content.encode()
+
+            await asyncio.to_thread(run)
+        self.logger.info("翻译完成")
+        return self

docutranslate/translater/txt_translator.py

@@ -0,0 +1,50 @@
+from dataclasses import dataclass
+from typing import Self
+
+from docutranslate.agents.txt_agent import TXTTranslateAgent
+from docutranslate.ir.document import Document
+from docutranslate.logger import global_logger
+from docutranslate.translater.base import AiTranslateConfig
+from docutranslate.translater.interfaces import Translator
+from docutranslate.utils.markdown_splitter import split_markdown_text
+
+
+@dataclass
+class TXTTranslateConfig(AiTranslateConfig):
+    ...
+
+
+class TXTTranslator(Translator):
+    def __init__(self, config: TXTTranslateConfig):
+        self.logger = config.logger or global_logger
+        self.chunk_size = config.chunk_size
+        self.translate_agent = TXTTranslateAgent(custom_prompt=config.custom_prompt,
+                                                 to_lang=config.to_lang,
+                                                 baseurl=config.base_url,
+                                                 key=config.api_key,
+                                                 model_id=config.model_id,
+                                                 system_prompt=None,
+                                                 temperature=config.temperature,
+                                                 max_concurrent=config.concurrent,
+                                                 timeout=config.timeout,
+                                                 logger=self.logger)
+
+    def translate(self, document: Document) -> Self:
+        self.logger.info("正在翻译txt")
+        chunks: list[str] = split_markdown_text(document.content.decode(), max_block_size=self.chunk_size)
+        self.logger.info(f"txt分为{len(chunks)}块")
+        result: list[str] = self.translate_agent.send_prompts(chunks)
+        content = "\n".join(result)
+        document.content = content.encode()
+        self.logger.info("翻译完成")
+        return self
+
+    async def translate_async(self, document: Document) -> Self:
+        self.logger.info("正在翻译txt")
+        chunks: list[str] = split_markdown_text(document.content.decode(), max_block_size=self.chunk_size)
+        self.logger.info(f"txt分为{len(chunks)}块")
+        result: list[str] = await self.translate_agent.send_prompts_async(chunks)
+        content = "\n".join(result)
+        document.content = content.encode()
+        self.logger.info("翻译完成")
+        return self