Leon/docutranslate
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

config去掉system_prompt

Browse Source
This commit is contained in:
xunbu
2025-08-28 16:43:23 +08:00
parent 56f5f4023e
commit 7192b961c7
16 changed files with 236 additions and 150 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
docutranslate/agents/agent.py
View File

@@ -26,7 +26,6 @@ class AgentConfig:
baseurl: str baseurl: str
key: str key: str
model_id: str model_id: str
system_prompt: str | None
temperature: float = 0.7 temperature: float = 0.7
max_concurrent: int = 30 max_concurrent: int = 30
timeout: int = 2000 timeout: int = 2000
@@ -68,8 +67,8 @@ class PromptsCounter:
PreSendHandlerType = Callable[[str, str], tuple[str, str]] PreSendHandlerType = Callable[[str, str], tuple[str, str]]
ResultHandlerType = Callable[[str, str, logging.Logger], str] ResultHandlerType = Callable[[str, str, logging.Logger], Any]
ErrorResultHandlerType = Callable[[str, logging.Logger], str] ErrorResultHandlerType = Callable[[str, logging.Logger], Any]
class Agent: class Agent:
@@ -90,7 +89,7 @@ class Agent:
self.domain = urlparse(self.baseurl).netloc self.domain = urlparse(self.baseurl).netloc
self.key = config.key.strip() or "xx" self.key = config.key.strip() or "xx"
self.model_id = config.model_id.strip() self.model_id = config.model_id.strip()
self.system_prompt = config.system_prompt or "" self.system_prompt = ""
self.temperature = config.temperature self.temperature = config.temperature
self.max_concurrent = config.max_concurrent self.max_concurrent = config.max_concurrent
self.timeout = config.timeout self.timeout = config.timeout
@@ -158,6 +157,8 @@ class Agent:
self.logger.warning(f"AI请求连接错误 (async): {repr(e)}") self.logger.warning(f"AI请求连接错误 (async): {repr(e)}")
except (KeyError, IndexError) as e: except (KeyError, IndexError) as e:
raise Exception(f"AI响应格式错误 (async): {repr(e)}") raise Exception(f"AI响应格式错误 (async): {repr(e)}")
except ValueError as e:
self.logger.warning(f"{e.__repr__()}")
# 如果没有正常获取结果则重试 # 如果没有正常获取结果则重试
if retry and retry_count < MAX_RETRY_COUNT: if retry and retry_count < MAX_RETRY_COUNT:
if self.total_error_counter.add(): if self.total_error_counter.add():
@@ -181,7 +182,7 @@ class Agent:
) -> list[Any]: ) -> list[Any]:
max_concurrent = self.max_concurrent if max_concurrent is None else max_concurrent max_concurrent = self.max_concurrent if max_concurrent is None else max_concurrent
total = len(prompts) total = len(prompts)
self.logger.info(f"base-url:{self.baseurl},model-id:{self.model_id}") self.logger.info(f"base-url:{self.baseurl},model-id:{self.model_id},concurrent:{self.max_concurrent},temperature:{self.temperature}")
self.logger.info(f"预计发送{total}个请求，并发请求数:{max_concurrent}") self.logger.info(f"预计发送{total}个请求，并发请求数:{max_concurrent}")
self.total_error_counter.max_errors_count = len(prompts) // MAX_REQUESTS_PER_ERROR # 允许多少个异常 self.total_error_counter.max_errors_count = len(prompts) // MAX_REQUESTS_PER_ERROR # 允许多少个异常
count = 0 count = 0
@@ -258,7 +259,8 @@ class Agent:
pre_send_handler, pre_send_handler,
result_handler, result_handler,
error_result_handler) -> Any: error_result_handler) -> Any:
result = self.send(client, prompt, system_prompt, pre_send_handler=pre_send_handler,result_handler=result_handler, result = self.send(client, prompt, system_prompt, pre_send_handler=pre_send_handler,
result_handler=result_handler,
error_result_handler=error_result_handler) error_result_handler=error_result_handler)
count.add() count.add()
return result return result
@@ -267,11 +269,11 @@ class Agent:
self, self,
prompts: list[str], prompts: list[str],
system_prompt: str | None = None, system_prompt: str | None = None,
pre_send_handler:PreSendHandlerType=None, pre_send_handler: PreSendHandlerType = None,
result_handler: ResultHandlerType = None, result_handler: ResultHandlerType = None,
error_result_handler: ErrorResultHandlerType = None error_result_handler: ErrorResultHandlerType = None
) -> list[Any]: ) -> list[Any]:
self.logger.info(f"base-url:{self.baseurl},model-id:{self.model_id}") self.logger.info(f"base-url:{self.baseurl},model-id:{self.model_id},concurrent:{self.max_concurrent},temperature:{self.temperature}")
self.logger.info(f"预计发送{len(prompts)}个请求，并发请求数:{self.max_concurrent}") self.logger.info(f"预计发送{len(prompts)}个请求，并发请求数:{self.max_concurrent}")
self.total_error_counter.max_errors_count = len(prompts) // MAX_REQUESTS_PER_ERROR # 允许多少个异常 self.total_error_counter.max_errors_count = len(prompts) // MAX_REQUESTS_PER_ERROR # 允许多少个异常
# 创建单个计数器实例 # 创建单个计数器实例
@@ -280,7 +282,7 @@ class Agent:
# 使用 itertools.repeat 将同一个实例传递给每个 map 调用 # 使用 itertools.repeat 将同一个实例传递给每个 map 调用
system_prompts = itertools.repeat(system_prompt, len(prompts)) system_prompts = itertools.repeat(system_prompt, len(prompts))
counters = itertools.repeat(counter, len(prompts)) counters = itertools.repeat(counter, len(prompts))
pre_send_handlers=itertools.repeat(pre_send_handler,len(prompts)) pre_send_handlers = itertools.repeat(pre_send_handler, len(prompts))
result_handlers = itertools.repeat(result_handler, len(prompts)) result_handlers = itertools.repeat(result_handler, len(prompts))
error_result_handlers = itertools.repeat(error_result_handler, len(prompts)) error_result_handlers = itertools.repeat(error_result_handler, len(prompts))
output_list = [] output_list = []

9
docutranslate/agents/glossary_agent.py
View File

@@ -46,6 +46,8 @@ You are a professional machine translation engine.
""" """
def _result_handler(self, result: str, origin_prompt: str, logger: Logger): def _result_handler(self, result: str, origin_prompt: str, logger: Logger):
if result == "":
return []
try: try:
result = json_repair.loads(result) result = json_repair.loads(result)
if not isinstance(result, list): if not isinstance(result, list):
@@ -56,6 +58,8 @@ You are a professional machine translation engine.
return result return result
def _error_result_handler(self, origin_prompt: str, logger: Logger): def _error_result_handler(self, origin_prompt: str, logger: Logger):
if origin_prompt == "":
return []
try: try:
return json_repair.loads(origin_prompt) return json_repair.loads(origin_prompt)
except: except:
@@ -79,11 +83,11 @@ You are a professional machine translation engine.
self.logger.info(f"json解析错误，解析文本:{chunk}，错误:{e.__repr__()}") self.logger.info(f"json解析错误，解析文本:{chunk}，错误:{e.__repr__()}")
except Exception as e: except Exception as e:
self.logger.info(f"send_segments发生错误:{e.__repr__()}") self.logger.info(f"send_segments发生错误:{e.__repr__()}")
self.logger.info("术语表提取完成")
return result return result
async def send_segments_async(self, segments: list[str], chunk_size: int): async def send_segments_async(self, segments: list[str], chunk_size: int):
self.logger.info("开始提取术语表") self.logger.info("开始术语表提取")
result = {} result = {}
indexed_originals, chunks, merged_indices_list = await asyncio.to_thread(segments2json_chunks, segments, indexed_originals, chunks, merged_indices_list = await asyncio.to_thread(segments2json_chunks, segments,
chunk_size) chunk_size)
@@ -101,4 +105,5 @@ You are a professional machine translation engine.
except Exception as e: except Exception as e:
self.logger.info(f"send_segments发生错误:{e.__repr__()}") self.logger.info(f"send_segments发生错误:{e.__repr__()}")
print(f"术语表:\n{result}") print(f"术语表:\n{result}")
self.logger.info("术语表提取完成")
return result return result

11
docutranslate/agents/segments_agent.py
View File

@@ -57,16 +57,19 @@ Warning: Never wrap the entire JSON object in quotes to make it a single string.
return system_prompt, prompt return system_prompt, prompt
def _result_handler(self, result: str, origin_prompt: str, logger: Logger): def _result_handler(self, result: str, origin_prompt: str, logger: Logger):
if result == "":
return {}
try: try:
result = json_repair.loads(result) result = json_repair.loads(result)
if not isinstance(result, dict): if not isinstance(result, dict):
raise ValueError("agent返回结果不是dict的json形式") raise ValueError(f"agent返回结果不是dict的json形式,result:{result}")
except: except RuntimeError as e:
logger.error("结果不能正确解析") raise ValueError(f"结果不能正确解析:{e.__repr__()}")
return self._error_result_handler(origin_prompt, logger)
return result return result
def _error_result_handler(self, origin_prompt: str, logger: Logger): def _error_result_handler(self, origin_prompt: str, logger: Logger):
if origin_prompt == "":
return {}
try: try:
return json_repair.loads(origin_prompt) return json_repair.loads(origin_prompt)
except: except:

176
docutranslate/app.py
View File

@@ -22,6 +22,7 @@ from pydantic import BaseModel, Field, field_validator
from docutranslate import __version__ from docutranslate import __version__
from docutranslate.agents.agent import ThinkingMode from docutranslate.agents.agent import ThinkingMode
from docutranslate.agents.glossary_agent import GlossaryAgentConfig
from docutranslate.exporter.md.types import ConvertEngineType from docutranslate.exporter.md.types import ConvertEngineType
# --- 核心代码 Imports --- # --- 核心代码 Imports ---
from docutranslate.global_values.conditional_import import DOCLING_EXIST from docutranslate.global_values.conditional_import import DOCLING_EXIST
@@ -220,6 +221,16 @@ app.mount("/static", StaticFiles(directory=STATIC_DIR), name="static")
# --- Pydantic Models for Service API --- # --- Pydantic Models for Service API ---
# =================================================================== # ===================================================================
class GlossaryAgentConfigPayload(BaseModel):
baseurl: str = Field(..., description="用于术语表生成的Agent的LLM API基础URL。", examples=["https://api.openai.com/v1"])
key: str = Field(..., description="用于术语表生成的Agent的LLM API密钥。", examples=["sk-agent-api-key"])
model_id: str = Field(..., description="用于术语表生成的Agent的模型ID。", examples=["gpt-4-turbo"])
temperature: float = Field(default=0.7, description="用于术语表生成的Agent的温度参数。")
max_concurrent: int = Field(default=30, description="Agent的最大并发请求数。")
timeout: int = Field(default=2000, description="Agent的API调用超时时间。")
thinking: ThinkingMode = Field(default="default", description="Agent的思考模式。")
# 1. 定义所有工作流共享的基础参数 # 1. 定义所有工作流共享的基础参数
class BaseWorkflowParams(BaseModel): class BaseWorkflowParams(BaseModel):
base_url: str = Field(..., description="LLM API的基础URL。", examples=["https://api.openai.com/v1"]) base_url: str = Field(..., description="LLM API的基础URL。", examples=["https://api.openai.com/v1"])
@@ -233,6 +244,15 @@ class BaseWorkflowParams(BaseModel):
examples=["default", "enable", "disable"]) examples=["default", "enable", "disable"])
custom_prompt: Optional[str] = Field(None, description="用户自定义的翻译Prompt。", alias="custom_prompt") custom_prompt: Optional[str] = Field(None, description="用户自定义的翻译Prompt。", alias="custom_prompt")
glossary_dict: Optional[Dict[str, str]] = Field(None, description="术语表字典，key为原文，value为译文。") glossary_dict: Optional[Dict[str, str]] = Field(None, description="术语表字典，key为原文，value为译文。")
glossary_generate_enable: bool = Field(default=False, description="是否开启术语表自动生成。")
glossary_agent_config: Optional[GlossaryAgentConfigPayload] = Field(None,
description="用于术语表生成的Agent的配置。如果 `glossary_generate_enable` 为 `True`，此项必填。")
@field_validator('glossary_agent_config')
def check_glossary_config(cls, v, values):
if values.data.get('glossary_generate_enable') and not v:
raise ValueError("当 `glossary_generate_enable` 为 `True` 时, `glossary_agent_config` 字段是必须的。")
return v
# 2. 为每个工作流创建独立的参数模型 # 2. 为每个工作流创建独立的参数模型
@@ -410,6 +430,27 @@ class TranslateServiceRequest(BaseModel):
} }
} }
}, },
{
"summary": "XLSX 带术语表生成",
"value": {
"file_name": "complex_terms.xlsx",
"file_content": "UEsDBBQAAAAIA... (base64-encoded xlsx)",
"payload": {
"workflow_type": "xlsx",
"base_url": "https://api.openai.com/v1",
"api_key": "sk-your-main-translator-key",
"model_id": "gpt-4o",
"to_lang": "简体中文",
"glossary_generate_enable": True,
"glossary_agent_config": {
"baseurl": "https://api.openai.com/v1",
"key": "sk-your-agent-key-for-glossary",
"model_id": "gpt-4-turbo",
"temperature": 0.5
}
}
}
},
{ {
"summary": "DOCX 工作流示例", "summary": "DOCX 工作流示例",
"value": { "value": {
@@ -508,15 +549,28 @@ async def _perform_translation(
workflow: Workflow workflow: Workflow
# 辅助函数：构建术语表生成配置
def build_glossary_agent_config():
if payload.glossary_generate_enable and payload.glossary_agent_config:
agent_payload = payload.glossary_agent_config
return GlossaryAgentConfig(
logger=task_logger,
to_lang=payload.to_lang,
**agent_payload.model_dump()
)
return None
# 2. 根据 payload 的具体类型构建配置并实例化 workflow # 2. 根据 payload 的具体类型构建配置并实例化 workflow
if isinstance(payload, MarkdownWorkflowParams): if isinstance(payload, MarkdownWorkflowParams):
task_logger.info("构建 MarkdownBasedWorkflow 配置。") task_logger.info("构建 MarkdownBasedWorkflow 配置。")
translator_config = MDTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent', 'glossary_dict'
'temperature', 'thinking', 'chunk_size', 'concurrent', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = MDTranslatorConfig(**translator_args)
converter_config = None converter_config = None
if payload.convert_engine == 'mineru': if payload.convert_engine == 'mineru':
converter_config = ConverterMineruConfig(logger=task_logger, mineru_token=payload.mineru_token, converter_config = ConverterMineruConfig(logger=task_logger, mineru_token=payload.mineru_token,
@@ -535,12 +589,14 @@ async def _perform_translation(
elif isinstance(payload, TextWorkflowParams): elif isinstance(payload, TextWorkflowParams):
task_logger.info("构建 TXTWorkflow 配置。") task_logger.info("构建 TXTWorkflow 配置。")
translator_config = TXTTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent', 'glossary_dict'
'temperature', 'thinking', 'chunk_size', 'concurrent', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = TXTTranslatorConfig(**translator_args)
html_exporter_config = TXT2HTMLExporterConfig(cdn=True) html_exporter_config = TXT2HTMLExporterConfig(cdn=True)
workflow_config = TXTWorkflowConfig( workflow_config = TXTWorkflowConfig(
translator_config=translator_config, html_exporter_config=html_exporter_config, translator_config=translator_config, html_exporter_config=html_exporter_config,
@@ -550,13 +606,15 @@ async def _perform_translation(
elif isinstance(payload, JsonWorkflowParams): elif isinstance(payload, JsonWorkflowParams):
task_logger.info("构建 JsonWorkflow 配置。") task_logger.info("构建 JsonWorkflow 配置。")
translator_config = JsonTranslatorConfig( translator_args = payload.model_dump(include={
json_paths=payload.json_paths, 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
**payload.model_dump(include={ 'temperature', 'thinking', 'chunk_size', 'concurrent', 'glossary_dict',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'json_paths'
'temperature', 'thinking', 'chunk_size', 'concurrent', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = JsonTranslatorConfig(**translator_args)
html_exporter_config = Json2HTMLExporterConfig(cdn=True) html_exporter_config = Json2HTMLExporterConfig(cdn=True)
workflow_config = JsonWorkflowConfig( workflow_config = JsonWorkflowConfig(
translator_config=translator_config, html_exporter_config=html_exporter_config, translator_config=translator_config, html_exporter_config=html_exporter_config,
@@ -566,13 +624,15 @@ async def _perform_translation(
elif isinstance(payload, XlsxWorkflowParams): elif isinstance(payload, XlsxWorkflowParams):
task_logger.info("构建 XlsxWorkflow 配置。") task_logger.info("构建 XlsxWorkflow 配置。")
translator_config = XlsxTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent',
'temperature', 'thinking', 'chunk_size', 'concurrent', 'insert_mode', 'separator', 'translate_regions', 'glossary_dict'
'insert_mode', 'separator', 'translate_regions', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = XlsxTranslatorConfig(**translator_args)
html_exporter_config = Xlsx2HTMLExporterConfig(cdn=True) html_exporter_config = Xlsx2HTMLExporterConfig(cdn=True)
workflow_config = XlsxWorkflowConfig( workflow_config = XlsxWorkflowConfig(
translator_config=translator_config, translator_config=translator_config,
@@ -583,13 +643,15 @@ async def _perform_translation(
elif isinstance(payload, DocxWorkflowParams): elif isinstance(payload, DocxWorkflowParams):
task_logger.info("构建 DocxWorkflow 配置。") task_logger.info("构建 DocxWorkflow 配置。")
translator_config = DocxTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent',
'temperature', 'thinking', 'chunk_size', 'concurrent', 'insert_mode', 'separator', 'glossary_dict'
'insert_mode', 'separator', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = DocxTranslatorConfig(**translator_args)
html_exporter_config = Docx2HTMLExporterConfig(cdn=True) html_exporter_config = Docx2HTMLExporterConfig(cdn=True)
workflow_config = DocxWorkflowConfig( workflow_config = DocxWorkflowConfig(
translator_config=translator_config, translator_config=translator_config,
@@ -600,13 +662,15 @@ async def _perform_translation(
elif isinstance(payload, SrtWorkflowParams): elif isinstance(payload, SrtWorkflowParams):
task_logger.info("构建 SrtWorkflow 配置。") task_logger.info("构建 SrtWorkflow 配置。")
translator_config = SrtTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent',
'temperature', 'thinking', 'chunk_size', 'concurrent', 'insert_mode', 'separator', 'glossary_dict'
'insert_mode', 'separator', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = SrtTranslatorConfig(**translator_args)
html_exporter_config = Srt2HTMLExporterConfig(cdn=True) html_exporter_config = Srt2HTMLExporterConfig(cdn=True)
workflow_config = SrtWorkflowConfig( workflow_config = SrtWorkflowConfig(
translator_config=translator_config, translator_config=translator_config,
@@ -617,13 +681,15 @@ async def _perform_translation(
elif isinstance(payload, EpubWorkflowParams): elif isinstance(payload, EpubWorkflowParams):
task_logger.info("构建 EpubWorkflow 配置。") task_logger.info("构建 EpubWorkflow 配置。")
translator_config = EpubTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent',
'temperature', 'thinking', 'chunk_size', 'concurrent', 'insert_mode', 'separator', 'glossary_dict'
'insert_mode', 'separator', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = EpubTranslatorConfig(**translator_args)
html_exporter_config = Epub2HTMLExporterConfig(cdn=True) html_exporter_config = Epub2HTMLExporterConfig(cdn=True)
workflow_config = EpubWorkflowConfig( workflow_config = EpubWorkflowConfig(
translator_config=translator_config, translator_config=translator_config,
@@ -635,13 +701,15 @@ async def _perform_translation(
# --- HTML WORKFLOW LOGIC START --- # --- HTML WORKFLOW LOGIC START ---
elif isinstance(payload, HtmlWorkflowParams): elif isinstance(payload, HtmlWorkflowParams):
task_logger.info("构建 HtmlWorkflow 配置。") task_logger.info("构建 HtmlWorkflow 配置。")
translator_config = HtmlTranslatorConfig( translator_args = payload.model_dump(include={
**payload.model_dump(include={ 'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt',
'base_url', 'api_key', 'model_id', 'to_lang', 'custom_prompt', 'temperature', 'thinking', 'chunk_size', 'concurrent',
'temperature', 'thinking', 'chunk_size', 'concurrent', 'insert_mode', 'separator', 'glossary_dict'
'insert_mode', 'separator', 'glossary_dict' }, exclude_none=True)
}, exclude_none=True) translator_args['glossary_generate_enable'] = payload.glossary_generate_enable
) translator_args['glossary_agent_config'] = build_glossary_agent_config()
translator_config = HtmlTranslatorConfig(**translator_args)
workflow_config = HtmlWorkflowConfig( workflow_config = HtmlWorkflowConfig(
translator_config=translator_config, translator_config=translator_config,
logger=task_logger logger=task_logger

7
docutranslate/glossary/glossary.py
View File

@@ -8,9 +8,14 @@ class Glossary:
self.glossary_dict[src]=dst self.glossary_dict[src]=dst
def append_system_prompt(self,text:str): def append_system_prompt(self,text:str):
flag=False
prompt="\n以下为参考术语表:\n" prompt="\n以下为参考术语表:\n"
for src,dst in self.glossary_dict.items(): for src,dst in self.glossary_dict.items():
if src in text: if src in text:
prompt+=f"{src}=>{dst}\n" prompt+=f"{src}=>{dst}\n"
flag=True
prompt+="术语表结束\n" prompt+="术语表结束\n"
return prompt if flag:
return prompt
else:
return ""

150
docutranslate/static/i18nData.json
View File

File diff suppressed because one or more lines are too long

2
docutranslate/static/index.html
View File

File diff suppressed because one or more lines are too long

2
docutranslate/translator/ai_translator/base.py
View File

@@ -21,7 +21,7 @@ class AiTranslatorConfig(TranslatorConfig):
chunk_size: int = 3000 chunk_size: int = 3000
concurrent: int = 30 concurrent: int = 30
glossary_dict: dict[str:str] | None = None glossary_dict: dict[str:str] | None = None
glossary_generate_enable: bool = True glossary_generate_enable: bool = False
glossary_agent_config: GlossaryAgentConfig | None = None glossary_agent_config: GlossaryAgentConfig | None = None

1
docutranslate/translator/ai_translator/docx_translator.py
View File

@@ -44,7 +44,6 @@ class DocxTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,

2
docutranslate/translator/ai_translator/epub_translator.py
View File

@@ -32,7 +32,7 @@ class EpubTranslator(AiTranslator):
agent_config = SegmentsTranslateAgentConfig( agent_config = SegmentsTranslateAgentConfig(
custom_prompt=config.custom_prompt, to_lang=config.to_lang, custom_prompt=config.custom_prompt, to_lang=config.to_lang,
baseurl=config.base_url, key=config.api_key, model_id=config.model_id, baseurl=config.base_url, key=config.api_key, model_id=config.model_id,
system_prompt=None, temperature=config.temperature, thinking=config.thinking, temperature=config.temperature, thinking=config.thinking,
max_concurrent=config.concurrent, timeout=config.timeout, logger=self.logger, max_concurrent=config.concurrent, timeout=config.timeout, logger=self.logger,
glossary_dict=config.glossary_dict glossary_dict=config.glossary_dict
) )

1
docutranslate/translator/ai_translator/html_translator.py
View File

@@ -91,7 +91,6 @@ class HtmlTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,

1
docutranslate/translator/ai_translator/json_translator.py
View File

@@ -23,7 +23,6 @@ class JsonTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,

1
docutranslate/translator/ai_translator/md_translator.py
View File

@@ -24,7 +24,6 @@ class MDTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,

1
docutranslate/translator/ai_translator/srt_translator.py
View File

@@ -31,7 +31,6 @@ class SrtTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,

1
docutranslate/translator/ai_translator/txt_translator.py
View File

@@ -21,7 +21,6 @@ class TXTTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,

1
docutranslate/translator/ai_translator/xlsx_translator.py
View File

@@ -31,7 +31,6 @@ class XlsxTranslator(AiTranslator):
baseurl=config.base_url, baseurl=config.base_url,
key=config.api_key, key=config.api_key,
model_id=config.model_id, model_id=config.model_id,
system_prompt=None,
temperature=config.temperature, temperature=config.temperature,
thinking=config.thinking, thinking=config.thinking,
max_concurrent=config.concurrent, max_concurrent=config.concurrent,