From e2bb898c54cad6505d6bd351ef99b0af9611b625 Mon Sep 17 00:00:00 2001 From: xunbu Date: Tue, 4 Nov 2025 18:03:58 +0800 Subject: [PATCH] =?UTF-8?q?=E6=94=AF=E6=8C=81mineru=E9=83=A8=E7=BD=B2?= =?UTF-8?q?=E6=9C=8D=E5=8A=A1?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- docutranslate/app.py | 99 ++++++++++++---- docutranslate/static/i18nData.json | 182 ++++++++++++++++------------- docutranslate/static/index.html | 2 +- 3 files changed, 177 insertions(+), 106 deletions(-) diff --git a/docutranslate/app.py b/docutranslate/app.py index c0a2f53..d89c0eb 100644 --- a/docutranslate/app.py +++ b/docutranslate/app.py @@ -37,7 +37,6 @@ from fastapi.staticfiles import StaticFiles from pydantic import ( BaseModel, Field, - field_validator, model_validator, AliasChoices, ConfigDict, @@ -46,7 +45,6 @@ from pydantic import ( from docutranslate import __version__ from docutranslate.agents.agent import ThinkingMode from docutranslate.agents.glossary_agent import GlossaryAgentConfig -from docutranslate.exporter.md.types import ConvertEngineType # --- 核心代码 Imports --- from docutranslate.global_values.conditional_import import DOCLING_EXIST @@ -78,6 +76,9 @@ from docutranslate.workflow.xlsx_workflow import XlsxWorkflow, XlsxWorkflowConfi if DOCLING_EXIST or TYPE_CHECKING: from docutranslate.converter.x2md.converter_docling import ConverterDoclingConfig from docutranslate.converter.x2md.converter_mineru import ConverterMineruConfig +# --- 新增的 Import --- +from docutranslate.converter.x2md.converter_mineru_deploy import ConverterMineruDeployConfig +# ---------------------- from docutranslate.exporter.md.md2html_exporter import MD2HTMLExporterConfig from docutranslate.exporter.txt.txt2html_exporter import TXT2HTMLExporterConfig from docutranslate.translator.ai_translator.md_translator import MDTranslatorConfig @@ -399,31 +400,61 @@ class MarkdownWorkflowParams(BaseWorkflowParams): workflow_type: Literal["markdown_based"] = Field( ..., description="指定使用基于Markdown的翻译工作流。" ) - convert_engine: ConvertEngineType = Field( + convert_engine: Literal["identity", "mineru", "docling", "mineru_deploy"] = Field( "identity", - description="选择将文件解析为markdown的引擎。如果输入文件是.md，此项可为`null`或不传。", - examples=["identity", "mineru", "docling"], - ) - mineru_token: Optional[str] = Field( - None, description="当 `convert_engine` 为 'mineru' 时必填的API令牌。" - ) - formula_ocr: bool = Field( - True, description="是否对公式进行OCR识别。对 `mineru` 和 `docling` 均有效。" - ) - code_ocr: bool = Field( - True, description="是否对代码块进行OCR识别。仅 `docling` 引擎有效。" - ) - model_version: Literal["pipeline", "vlm"] = Field( - "vlm", description="Mineru模型的版本，'vlm'是更新的版本。仅 `mineru` 引擎有效。" + description="选择将文件解析为markdown的引擎。'mineru_deploy' 适用于本地部署的 MinerU 服务。如果输入文件是.md，此项可为`identity`或不传。", + examples=["identity", "mineru", "docling", "mineru_deploy"], ) - @field_validator("mineru_token") - def check_mineru_token(cls, v, values): - if values.data.get("convert_engine") == "mineru" and not v: + # --- Engine-Specific Parameters --- + + # -- For "mineru" (Cloud API) -- + mineru_token: Optional[str] = Field( + None, description="[仅当 convert_engine='mineru'] 必填的API令牌。" + ) + model_version: Literal["pipeline", "vlm"] = Field( + "vlm", description="[仅当 convert_engine='mineru'] Mineru Cloud模型的版本。" + ) + formula_ocr: bool = Field( + True, description="[仅当 convert_engine='mineru' 或 'docling'] 是否对公式进行OCR识别。" + ) + + # -- For "docling" -- + code_ocr: bool = Field( + True, description="[仅当 convert_engine='docling'] 是否对代码块进行OCR识别。" + ) + + # -- For "mineru_deploy" (Local Deployment) -- + mineru_deploy_base_url: Optional[str] = Field( + "http://127.0.0.1:8000", + description="[仅当 convert_engine='mineru_deploy'] 本地部署的 MinerU 服务地址。", + ) + mineru_deploy_backend: Literal["pipeline", "vlm"] = Field( + "pipeline", + description="[仅当 convert_engine='mineru_deploy'] 本地部署的 MinerU 服务使用的后端。", + ) + mineru_deploy_formula_enable: bool = Field( + True, + description="[仅当 convert_engine='mineru_deploy'] 本地部署的服务是否启用公式解析。", + ) + mineru_deploy_start_page_id: int = Field( + 0, description="[仅当 convert_engine='mineru_deploy'] 起始解析页面。" + ) + mineru_deploy_end_page_id: int = Field( + 99999, description="[仅当 convert_engine='mineru_deploy'] 结束解析页面。" + ) + + @model_validator(mode="after") + def check_engine_params(self): + if self.convert_engine == "mineru" and not self.mineru_token: raise ValueError( "当 `convert_engine` 为 'mineru' 时，`mineru_token` 字段是必须的。" ) - return v + if self.convert_engine == "mineru_deploy" and not self.mineru_deploy_base_url: + raise ValueError( + "当 `convert_engine` 为 'mineru_deploy' 时，`mineru_deploy_base_url` 字段是必须的。" + ) + return self class TextWorkflowParams(BaseWorkflowParams): @@ -612,6 +643,21 @@ class TranslateServiceRequest(BaseModel): "model_version": "vlm", }, }, + { + "file_name": "local_test.pdf", + "file_content": "JVBERi0xLjcKJeLjz9MKMSAwIG9iago8PC9...", + "payload": { + "workflow_type": "markdown_based", + "skip_translate": True, + "to_lang": "中文", + "convert_engine": "mineru_deploy", + "mineru_deploy_base_url": "http://127.0.0.1:8000", + "mineru_deploy_backend": "pipeline", + "mineru_deploy_formula_enable": True, + "mineru_deploy_start_page_id": 0, + "mineru_deploy_end_page_id": 5, + }, + }, { "file_name": "product_info.json", "file_content": "ewogICAgImlkIjogIjEyMzQ1IiwK...", @@ -874,6 +920,14 @@ async def _perform_translation( formula_ocr=payload.formula_ocr, model_version=payload.model_version, ) + elif payload.convert_engine == "mineru_deploy": + converter_config = ConverterMineruDeployConfig( + base_url=payload.mineru_deploy_base_url, + backend=payload.mineru_deploy_backend, + formula_enable=payload.mineru_deploy_formula_enable, + start_page_id=payload.mineru_deploy_start_page_id, + end_page_id=payload.mineru_deploy_end_page_id, + ) elif payload.convert_engine == "docling" and DOCLING_EXIST: converter_config = ConverterDoclingConfig( logger=task_logger, @@ -1458,7 +1512,6 @@ async def _start_translation_task( initial_log_msg = f"收到新的翻译请求: {original_filename}" print(f"[{task_id}] {initial_log_msg}") - log_history.append(initial_log_msg) await log_queue.put(initial_log_msg) try: @@ -2093,7 +2146,7 @@ async def service_content( "/engin-list", tags=["Application"], description="返回正在进行的可用的转换引擎" ) async def service_get_engin_list(): - engin_list = ["mineru"] + engin_list = ["mineru", "mineru_deploy"] if DOCLING_EXIST: engin_list.append("docling") return JSONResponse(content=engin_list) diff --git a/docutranslate/static/i18nData.json b/docutranslate/static/i18nData.json index 7776e21..0a79abc 100644 --- a/docutranslate/static/i18nData.json +++ b/docutranslate/static/i18nData.json @@ -46,12 +46,9 @@ "jsonSettingsTitleText": "JSON路径配置", "jsonPathLabel": "需要翻译的JSON路径", "jsonPathPlaceholder": "每行一个路径, 例如:\n$.name\n$.*", - "jsonPathHelp": "采用jsonpath-ng的路径选择语法，每一行表示一个json路径。 将翻译路径匹配对象内的所有字符串", + "jsonPathHelp": "采用jsonpath-ng的路径选择语法，每一行表示一个json路径。\n 将翻译路径匹配对象内的所有字符串", "parsingSettingsTitleText": "解析配置", "parsingEngineLabel": "解析引擎", - "engineOptionIdentity": "已经是markdown格式", - "engineOptionMineru": "Mineru (推荐)", - "engineOptionDocling": "Docling (本地解析)", "parsingEngineHelp": "如果上传的文件本身是.md格式，此项可不选。", "getMineruTokenTitle": "获取Mineru Token", "mineruTokenPlaceholder": "使用Mineru引擎时需要", @@ -59,13 +56,20 @@ "modelVersionVlm": "VLM", "modelVersionPipline": "Pipeline", "modelVersionHelp": "mineru VLM是更新的内测模型。", + "mineruDeployBaseUrlLabel": "服务地址 (Base URL)", + "mineruDeployBaseUrlPlaceholder": "例如: http://127.0.0.1:8000", + "mineruDeployBackendLabel": "后端类型", + "mineruDeployBackendPipeline": "Pipeline", + "mineruDeployBackendVlm": "VLM", + "mineruDeployStartPageLabel": "起始页面", + "mineruDeployEndPageLabel": "结束页面", + "mineruDeployFormulaEnableLabel": "启用公式解析", "formulaOcrLabel": "公式识别", "codeOcrLabel": "代码识别", "aiSettingsTitleText": "翻译模型", "skipTranslationLabel": "跳过翻译", "platformLabel": "选择平台", "platformCustom": "自定义接口", - "apiHrefInfo302ai": "👈 通过此链接注册可享1美元免费额度", "baseUrlLabel": "API 地址 (Base URL)", "baseUrlPlaceholder": "OpenAi兼容地址", "getApiKeyTitle": "获取API Key", @@ -101,7 +105,7 @@ "glossaryGenConfigCustom": "自定义", "importConfigBtn": "导入配置", "exportConfigBtn": "导出配置", - "githubInfo": "GitHub主页(欢迎star❤):
https://github.com/xunbu/docutranslate", + "githubInfo": "GitHub主页(欢迎star❤):
\n https://github.com/xunbu/docutranslate", "qqGroupInfo": "交流QQ群: 1047781902", "taskListTitle": "任务列表", "newTaskBtn": "新建任务", @@ -110,7 +114,7 @@ "taskCardIdPlaceholder": "等待提交...", "taskCardFileDrop": "点击或拖拽文件到此处", "taskCardFileSelected": "文件已选择", - "taskCardFilenameLabel": "文件名:", + "taskCardFilenameLabel": "文件名: ", "taskCardLogLabel": "日志", "copyLogsTooltip": "复制日志", "taskCardStatusWaiting": "等待上传文件...", @@ -130,7 +134,7 @@ "closeBtn": "关闭", "downloadBtn": "下载", "tutorialModalTitle": "使用教程", - "tutorialModalBody": "

视频教程可以在B站搜索 docutranslate 获取。

欢迎使用 DocuTranslate！请按照以下步骤完成文档翻译：

  1. 第一步：选择工作流

    在左侧配置面板的顶部，首先选择最适合您文件类型的处理流程。

    提示: 默认已开启“自动选择工作流”。您只需上传文件，系统会自动为您匹配合适的工作流，简化操作。

    • 转Markdown再翻译: 适用于翻译PDF、markdown、图片等文件。这是最通用和强大的模式。
    • 纯文本翻译: 用于翻译 .txt 纯文本文件。
    • EPUB翻译: 用于翻译 .epub 电子书文件。
    • DOCX翻译: 用于翻译 .docx Word文档。
    • XLSX翻译: 用于翻译 .xlsx 或 .csv 电子表格文件。
    • SRT字幕翻译: 用于翻译 .srt 字幕文件。
    • ASS字幕翻译: 用于翻译 .ass 特效字幕文件。
    • JSON翻译: 用于翻译 .json 文件中的特定字段。
    • HTML翻译: 用于翻译 .html 网页文件。
  2. 第二步：配置参数

    选择工作流后，下方会显示相关的配置选项。请依次完成设置（所有配置都会自动保存在您的浏览器中）：

    A. 工作流特定选项 (根据您第一步的选择出现):

    • 如果选择“转Markdown再翻译”，请配置 解析配置：
      • 解析引擎: 选择一个引擎将您的文件（如PDF）转换为适合翻译的Markdown格式。如果您的文件已经是Markdown格式，则无需选择。
      • Mineru Token: 如果您选择 minerU 引擎，需要在此处填入您的Token。
    • 如果选择“纯文本/DOCX/XLSX/SRT/ASS/EPUB/HTML”，请配置其 翻译选项：
      • 插入模式: 定义翻译结果如何放入文档。您可以选择直接“替换”原文，或是在原文之后“附加”，或是在原文之前“前置”。
      • 分隔符: 当选择“附加”或“前置”模式时，此项用于在原文和译文之间插入分隔符（例如ASS格式常用 \\N，EPUB格式常用 <br /> 作为换行分隔符）。
    • 如果选择“JSON翻译”，请配置 JSON路径：
      • 需要翻译的JSON路径: 每行输入一个 JSONPath 表达式，将翻译所有与路径匹配的对象中的字符串。例如:$.*(翻译全部字符串)，$..description(翻译所有键为description的值)。

    B. 通用选项 (适用于所有工作流):

    • 翻译模型:
      • 选择平台/API 地址/API Key/模型ID: 配置您希望使用的AI翻译服务。模型能力指令遵循越强，出错、漏翻的概率越低。
      • 跳过翻译: 勾选此项后，将只执行文档解析和格式转换，不调用AI进行翻译。
    • 翻译配置:
      • 目标语言: 指定翻译的目标语言。
      • 自定义Prompt: 可选，添加额外指令，如“人名保持原文不翻译”。
      • 思考模式: 针对部分支持混合推理的模型进行设置，建议选择“禁用(推荐)”。
      • 分块大小/并发数等: 高级参数，用于调整性能和API请求行为，通常保持默认即可。
    • 术语表:
      • 上传术语表 (可选): 上传CSV文件（需包含'src'和'dst'列）来保证特定术语翻译的统一性和准确性。
      • 自动生成术语表: 启用后，程序会先从原文中提取术语并生成一个术语表，然后再进行翻译。
  3. 第三步：上传文件

    在右侧的任务列表中，点击或拖拽您的文档到文件上传区域。

  4. 第四步：开始翻译

    文件选择成功后，点击任务卡片右下角的 开始翻译 按钮。系统将开始处理任务，您可以在日志区域查看实时进度。

  5. 第五步：查看与下载

    翻译完成后，任务卡片下方会出现操作按钮：

    • 预览: 在右侧滑出的面板中进行原文和译文的对照预览。
    • 下载: 下载包括 PDF, DOCX, Markdown 等多种格式的译文。
    • 附件: 如果翻译过程中生成了附加文件（如自动生成的术语表），可在此处下载。
重要提示: 所有配置都会自动保存在您的浏览器本地，方便下次使用。您也可以使用新增的“导出配置”和“导入配置”按钮来备份和恢复您的设置。
", + "tutorialModalBody": "

视频教程可以在B站搜索 docutranslate 获取。

\n

欢迎使用 DocuTranslate！请按照以下步骤完成文档翻译：

\n
    \n
  1. \n 第一步：选择工作流\n

    \n 在左侧配置面板的顶部，首先选择最适合您文件类型的处理流程。\n

    \n 提示:\n 默认已开启“自动选择工作流”。您只需上传文件，系统会自动为您匹配合适的工作流，简化操作。\n
    \n

    \n
      \n
    • 转Markdown再翻译: 适用于翻译PDF、markdown、图片等文件。这是最通用和强大的模式。
      • \n
    • 纯文本翻译: 用于翻译 .txt 纯文本文件。
      • \n
    • EPUB翻译: 用于翻译 .epub 电子书文件。
      • \n
    • DOCX翻译: 用于翻译 .docx Word文档。
      • \n
    • XLSX翻译: 用于翻译 .xlsx 或 .csv 电子表格文件。
      • \n
    • SRT字幕翻译: 用于翻译 .srt 字幕文件。
      • \n
    • ASS字幕翻译: 用于翻译 .ass 特效字幕文件。
      • \n
    • JSON翻译: 用于翻译 .json 文件中的特定字段。
      • \n
    • HTML翻译: 用于翻译 .html 网页文件。
      • \n
    \n
    2. \n
  2. \n 第二步：配置参数\n

    \n 选择工作流后，下方会显示相关的配置选项。请依次完成设置（所有配置都会自动保存在您的浏览器中）：

    \n

    A. 工作流特定选项 (根据您第一步的选择出现):

    \n
      \n
    • 如果选择“转Markdown再翻译”，请配置 解析配置：\n
        \n
      • 解析引擎:\n 选择一个引擎将您的文件（如PDF）转换为适合翻译的Markdown格式。如果您的文件已经是Markdown格式，则无需选择。\n
        • \n
      • Mineru Token: 如果您选择 minerU 引擎，需要在此处填入您的Token。\n
        • \n
      \n
      • \n
    • 如果选择“纯文本/DOCX/XLSX/SRT/ASS/EPUB/HTML”，请配置其 翻译选项：\n
        \n
      • 插入模式: 定义翻译结果如何放入文档。您可以选择直接“替换”原文，或是在原文之后“附加”，或是在原文之前“前置”。\n
        • \n
      • 分隔符: 当选择“附加”或“前置”模式时，此项用于在原文和译文之间插入分隔符（例如ASS格式常用 \\N，EPUB格式常用 <br /> 作为换行分隔符）。\n
        • \n
      \n
      • \n
    • 如果选择“JSON翻译”，请配置 JSON路径：\n
        \n
      • 需要翻译的JSON路径: 每行输入一个 JSONPath\n 表达式，将翻译所有与路径匹配的对象中的字符串。例如:$.*(翻译全部字符串)，$..description(翻译所有键为description的值)。\n
        • \n
      \n
      • \n
    \n

    B. 通用选项 (适用于所有工作流):

    \n
      \n
    • 翻译模型:\n
        \n
      • 选择平台/API 地址/API Key/模型ID: 配置您希望使用的AI翻译服务。模型能力指令遵循越强，出错、漏翻的概率越低。\n
        • \n
      • 跳过翻译: 勾选此项后，将只执行文档解析和格式转换，不调用AI进行翻译。\n
        • \n
      \n
      • \n
    • 翻译配置:\n
        \n
      • 目标语言: 指定翻译的目标语言。
        • \n
      • 自定义Prompt: 可选，添加额外指令，如“人名保持原文不翻译”。
        • \n
      • 思考模式: 针对部分支持混合推理的模型进行设置，建议选择“禁用(推荐)”。\n
        • \n
      • 分块大小/并发数等: 高级参数，用于调整性能和API请求行为，通常保持默认即可。\n
        • \n
      \n
      • \n
    • 术语表:\n
        \n
      • 上传术语表 (可选): 上传CSV文件（需包含'src'和'dst'列）来保证特定术语翻译的统一性和准确性。\n
        • \n
      • 自动生成术语表: 启用后，程序会先从原文中提取术语并生成一个术语表，然后再进行翻译。\n
        • \n
      \n
      • \n
    \n
    3. \n
  3. \n 第三步：上传文件\n

    在右侧的任务列表中，点击或拖拽您的文档到文件上传区域。

    \n
    4. \n
  4. \n 第四步：开始翻译\n

    文件选择成功后，点击任务卡片右下角的 开始翻译 按钮。系统将开始处理任务，您可以在日志区域查看实时进度。\n

    \n
    5. \n
  5. \n 第五步：查看与下载\n

    翻译完成后，任务卡片下方会出现操作按钮：

    \n
      \n
    • 预览: \n 在右侧滑出的面板中进行原文和译文的对照预览。\n
      • \n
    • 下载: \n 下载包括 PDF, DOCX, Markdown 等多种格式的译文。\n
      • \n
    • 附件: \n 如果翻译过程中生成了附加文件（如自动生成的术语表），可在此处下载。\n
      • \n
    \n
    6. \n
\n
\n 重要提示: \n 所有配置都会自动保存在您的浏览器本地，方便下次使用。您也可以使用新增的“导出配置”和“导入配置”按钮来备份和恢复您的设置。\n
", "tutorialUnderstandBtn": "我明白了", "contributorsModalTitle": "感谢贡献", "contributorsPara1": "DocuTranslate是一个开源项目！大家的需求与使用是项目进步的动力。", @@ -143,36 +147,41 @@ "glossaryModalTitle": "当前术语表", "glossaryTableSource": "原文 (src)", "glossaryTableDestination": "译文 (dst)", - "glossaryEmpty": "术语表为空。", "init_i18n_failed_alert": "加载界面翻译资源失败，请检查网络连接或联系管理员。", "init_failed_alert": "初始化失败，无法连接到后端服务。请检查服务是否运行或刷新页面。", - "status_selectFileFirst": "请先选择文件", - "status_fillRequired": "请填写所有必填项", + "apiHrefInfo302ai": "👈 通过此链接注册可享1美金免费额度", + "glossaryEmpty": "术语表为空。", + "engineOptionIdentity": "已经是Markdown格式", + "engineOptionMineru": "Mineru（推荐）", + "engineOptionDocling": "Docling", + "engineOptionMineruDeploy": "Mineru部署服务", + "preview_loading": "正在加载预览...", + "preview_cantReadOriginal": "无法读取原始文件内容。", + "preview_cantPreviewType": "无法预览此文件类型", + "preview_noOriginalCache": "无原始文件缓存可供预览。", + "preview_loadFailed": "预览加载失败。", + "pdf_preparing": "正在准备PDF，请稍候...", + "pdf_print_failed": "调用打印功能失败。请检查浏览器设置或手动保存为PDF。", + "pdf_fetch_failed": "获取译文内容失败，无法生成PDF。", + "preview_bilingual": "双语对照预览", + "preview_translatedOnly": "仅译文预览", + "admin_tasklist_failed": "管理员模式：获取任务列表失败。", + "status_selectFileFirst": "请先选择文件！", + "status_fillRequired": "请填写所有必填项！", "btn_initializing": "初始化中...", - "status_encodingAndSubmitting": "正在编码并提交请求...", - "status_requestOk": "请求成功，任务已开始", + "status_encodingAndSubmitting": "正在编码并提交任务...", + "status_requestOk": "请求成功，任务已开始。", "btn_cancelTranslation": "取消翻译", "status_requestFail": "请求失败", - "status_initFail": "初始化失败", - "status_cancelling": "正在取消...", - "status_cancelSent": "已发送取消请求", + "status_initFail": "任务初始化失败", + "status_cancelling": "取消中...", + "status_cancelSent": "已发送取消请求。", "status_cancelFail": "取消失败", "status_gettingStatus": "正在获取状态...", "btn_reTranslate": "重新翻译", "status_updateError": "状态更新出错", - "preview_loading": "

正在加载预览...

", - "preview_cantReadOriginal": "无法读取原始文件内容。", - "preview_cantPreviewType": "无法预览此文件类型", - "preview_noOriginalCache": "没有缓存的原始文件用于预览。", - "preview_loadFailed": "

加载预览失败

", - "pdf_preparing": "正在准备PDF，请稍候...", - "pdf_print_failed": "启动打印失败。请尝试手动右键点击预览内容并选择打印。", - "pdf_fetch_failed": "获取翻译内容失败，无法生成PDF。", - "preview_bilingual": "双语预览", - "preview_translatedOnly": "仅译文预览", - "admin_tasklist_failed": "管理员模式：加载任务列表失败。", - "configImportSuccess": "配置导入成功！界面已刷新。", - "configImportError": "导入失败，请检查配置文件是否为有效的JSON格式。" + "configImportSuccess": "配置导入成功！设置已应用。", + "configImportError": "导入失败！文件格式错误或内容不合法。" }, "en": { "pageTitle": "DocuTranslate - Interactive Document Translation", @@ -191,60 +200,64 @@ "autoWorkflowLabel": "Auto-select workflow", "txtSettingsTitleText": "TXT Translation Options", "insertModeLabel": "Insert Mode", - "insertModeReplace": "Replace", - "insertModeAppend": "Append", - "insertModePrepend": "Prepend", + "insertModeReplace": "Replace Original", + "insertModeAppend": "Append to Original", + "insertModePrepend": "Prepend to Original", "insertModeHelpTxt": "Choose how to insert the translated text.", "separatorLabel": "Separator", "separatorPlaceholderSimple": "e.g., \\n---\\n", - "separatorHelp": "Characters to separate original and translated text in append/prepend mode. \\n for newline.", + "separatorHelp": "Characters to separate original and translated text in append/prepend mode. \\n for new line.", "docxSettingsTitleText": "DOCX Translation Options", "insertModeHelpDocx": "Choose how to insert the translated text.", "separatorPlaceholderStructured": "e.g., ---", - "separatorHelpDocx": "In append mode, translation starts in a new paragraph. This adds extra content between paragraphs. \\n for internal newlines.", + "separatorHelpDocx": "In append mode, translation starts in a new paragraph. This adds extra content between paragraphs. \\n for new lines inside.", "xlsxSettingsTitleText": "XLSX Translation Options", "insertModeHelpXlsx": "Choose how to insert the translated text into cells.", - "xlsxTranslateRegionsLabel": "Translate Regions (Optional)", - "xlsxTranslateRegionsPlaceholder": "One region per line, e.g., Sheet1!A1:B10 (applies to all sheets if name is omitted)", + "xlsxTranslateRegionsLabel": "Translation Regions (Optional)", + "xlsxTranslateRegionsPlaceholder": "One region per line, e.g., Sheet1!A1:B10 (applies to all sheets if sheet name is omitted)", "srtSettingsTitleText": "SRT Translation Options", "insertModeHelpSrt": "Choose how to insert the translated text.", "epubSettingsTitleText": "EPUB Translation Options", "insertModeHelpEpub": "Choose how to insert the translated text.", - "separatorHelpEpub": "In append mode, translation starts in a new block. This adds extra content between blocks. \\n for internal newlines.", + "separatorHelpEpub": "In append mode, translation starts in a new block. This adds extra content between blocks. \\n for new lines inside.", "htmlSettingsTitleText": "HTML Translation Options", "insertModeHelpHtml": "Choose how to insert the translated text.", - "separatorHelpHtml": "In append mode, translation starts in a new block. This adds extra content between blocks. \\n for internal newlines.", + "separatorHelpHtml": "In append mode, translation starts in a new block. This adds extra content between blocks. \\n for new lines inside.", "assSettingsTitleText": "ASS Translation Options", "insertModeHelpAss": "Choose how to insert the translated text.", "separatorPlaceholderAss": "e.g., \\N (newline)", "separatorHelpAss": "Characters to separate original and translated text in append/prepend mode. \\N is the newline for ASS format.", "jsonSettingsTitleText": "JSON Path Configuration", "jsonPathLabel": "JSON Paths to Translate", - "jsonPathPlaceholder": "One path per line, e.g.,\n$.name\n$.*", - "jsonPathHelp": "Uses jsonpath-ng syntax. One JSON path per line. Translates all strings within matching objects.", + "jsonPathPlaceholder": "One path per line, e.g.:\n$.name\n$.*", + "jsonPathHelp": "Uses jsonpath-ng syntax. Each line is a JSON path. All strings within matching objects will be translated.", "parsingSettingsTitleText": "Parsing Configuration", "parsingEngineLabel": "Parsing Engine", - "engineOptionIdentity": "Already in Markdown format", - "engineOptionMineru": "Mineru (Recommended)", - "engineOptionDocling": "Docling (Local Parsing)", - "parsingEngineHelp": "Not required if the uploaded file is already in .md format.", + "parsingEngineHelp": "Not required if your uploaded file is already in .md format.", "getMineruTokenTitle": "Get Mineru Token", "mineruTokenPlaceholder": "Required when using Mineru engine", "modelVersionLabel": "Mineru Model Version", "modelVersionVlm": "VLM", "modelVersionPipline": "Pipeline", "modelVersionHelp": "Mineru VLM is a newer, internal test model.", + "mineruDeployBaseUrlLabel": "Service Address (Base URL)", + "mineruDeployBaseUrlPlaceholder": "e.g., http://127.0.0.1:8000", + "mineruDeployBackendLabel": "Backend Type", + "mineruDeployBackendPipeline": "Pipeline", + "mineruDeployBackendVlm": "VLM", + "mineruDeployStartPageLabel": "Start Page", + "mineruDeployEndPageLabel": "End Page", + "mineruDeployFormulaEnableLabel": "Enable Formula Parsing", "formulaOcrLabel": "Formula Recognition", "codeOcrLabel": "Code Recognition", "aiSettingsTitleText": "Translation Model", "skipTranslationLabel": "Skip Translation", "platformLabel": "Select Platform", "platformCustom": "Custom Endpoint", - "apiHrefInfo302ai": "👈 Register through this link to enjoy a $1 free credit", - "baseUrlLabel": "API Base URL", - "baseUrlPlaceholder": "OpenAI-compatible URL", + "baseUrlLabel": "API Address (Base URL)", + "baseUrlPlaceholder": "OpenAI-compatible address", "getApiKeyTitle": "Get API Key", - "apiKeyPlaceholder": "Enter your API Key", + "apiKeyPlaceholder": "Please enter your API Key", "modelIdLabel": "Model ID", "modelIdPlaceholder": "e.g., gpt-4o, glm-4", "systemProxyLabel": "Enable System Proxy", @@ -253,12 +266,12 @@ "targetLanguageCustom": "Other (Custom)", "customLangPlaceholder": "Enter target language, e.g., Italian", "thinkingModeLabel": "Thinking Mode", - "thinkingModeTooltip": "Configure whether the hybrid reasoning model should 'think'. Supported by Zhipu's glm4.5 series, Volcengine's seed1.6 series, SiliconFlow, Google's Gemini series, and some 302AI models. Disabling is recommended.", + "thinkingModeTooltip": "Sets whether the hybrid inference model should 'think'. Currently supports Zhipu's glm4.5 series, Volcengine's seed1.6 series, SiliconFlow platform, Google's Gemini series, and 302AI (partial). Disabling is recommended.", "thinkingModeEnable": "Enable", "thinkingModeDisable": "Disable (Recommended)", "thinkingModeDefault": "Default", "customPromptLabel": "Custom Prompt", - "customPromptPlaceholder": "Optional, e.g., 'Keep proper names in the original language'", + "customPromptPlaceholder": "Optional, e.g., 'Keep names in their original language.'", "chunkSizeLabel": "Chunk Size", "resetBtn": "Reset", "concurrentLabel": "Concurrency", @@ -268,7 +281,7 @@ "glossaryHelp": "Select one or more CSV files. Files must have 'src' and 'dst' columns for source and destination terms.", "viewGlossaryBtn": "View Glossary", "clearGlossaryBtn": "Clear", - "glossaryGenEnableLabel": "Auto-generate Glossary", + "glossaryGenEnableLabel": "Auto-generate glossary", "glossaryCustomPromptLabel": "Custom Prompt", "glossaryCustomPromptPlaceholder": "Prompt for glossary generation", "glossaryGenConfigLabel": "Glossary Generation Config", @@ -276,7 +289,7 @@ "glossaryGenConfigCustom": "Custom", "importConfigBtn": "Import Config", "exportConfigBtn": "Export Config", - "githubInfo": "GitHub Homepage (Star us ❤):
https://github.com/xunbu/docutranslate", + "githubInfo": "GitHub Homepage (stars❤ are welcome):
\n https://github.com/xunbu/docutranslate", "qqGroupInfo": "QQ Group: 1047781902", "taskListTitle": "Task List", "newTaskBtn": "New Task", @@ -285,7 +298,7 @@ "taskCardIdPlaceholder": "Awaiting submission...", "taskCardFileDrop": "Click or drag file here", "taskCardFileSelected": "File selected", - "taskCardFilenameLabel": "Filename:", + "taskCardFilenameLabel": "Filename: ", "taskCardLogLabel": "Logs", "copyLogsTooltip": "Copy logs", "taskCardStatusWaiting": "Waiting for file upload...", @@ -298,55 +311,60 @@ "downloadAss": "ASS", "previewTitle": "Preview", "previewBilingualBtn": "Bilingual", - "previewTranslatedOnlyBtn": "Translated Only", + "previewTranslatedOnlyBtn": "Translation Only", "syncScrollTooltip": "Sync Scrolling", "previewOriginal": "Original", - "previewTranslated": "Translated", + "previewTranslated": "Translation", "closeBtn": "Close", "downloadBtn": "Download", "tutorialModalTitle": "User Guide", - "tutorialModalBody": "

Video tutorials are available on Bilibili by searching for docutranslate.

Welcome to DocuTranslate! Follow these steps to translate your documents:

  1. Step 1: Select Workflow

    At the top of the left settings panel, choose the processing flow that best suits your file type.

    Tip: 'Auto-select workflow' is enabled by default. Simply upload your file, and the system will automatically match the appropriate workflow for you.

    • To Markdown then Translate: Ideal for PDF, markdown, and image files. This is the most versatile and powerful mode.
    • Plain Text Translation: For translating .txt files.
    • EPUB Translation: For translating .epub e-books.
    • DOCX Translation: For translating .docx Word documents.
    • XLSX Translation: For translating .xlsx or .csv spreadsheets.
    • SRT Subtitle Translation: For translating .srt subtitle files.
    • ASS Subtitle Translation: For translating .ass advanced subtitle files.
    • JSON Translation: For translating specific fields within .json files.
    • HTML Translation: For translating .html web pages.
  2. Step 2: Configure Parameters

    After selecting a workflow, relevant options will appear below. Configure them sequentially (all settings are saved automatically in your browser):

    A. Workflow-Specific Options (Appear based on your Step 1 choice):

    • If 'To Markdown then Translate' is selected, configure Parsing Configuration:
      • Parsing Engine: Choose an engine to convert your file (like a PDF) into a translation-friendly Markdown format. Not needed if your file is already Markdown.
      • Mineru Token: Required if you select the minerU engine.
    • If 'Plain Text/DOCX/XLSX/SRT/ASS/EPUB/HTML' is selected, configure its Translation Options:
      • Insert Mode: Define how the translation is placed. You can 'Replace' the original, 'Append' after it, or 'Prepend' before it.
      • Separator: Used in 'Append' or 'Prepend' mode to insert a separator between the original and translated text (e.g., \\N for ASS, <br /> for EPUB).
    • If 'JSON Translation' is selected, configure JSON Paths:
      • JSON Paths to Translate: Enter one JSONPath expression per line to translate all string values at matching paths. E.g., $.* (all strings), $..description (all values for the key description).

    B. General Options (Apply to all workflows):

    • Translation Model:
      • Select Platform/API Base URL/API Key/Model ID: Configure the AI translation service you want to use. More powerful models lead to fewer errors and omissions.
      • Skip Translation: If checked, only document parsing and format conversion will be performed, without calling the AI for translation.
    • Translation Configuration:
      • Target Language: Specify the language to translate to.
      • Custom Prompt: Optional. Add extra instructions, e.g., 'Do not translate proper names'.
      • Thinking Mode: A setting for certain hybrid reasoning models. 'Disable (Recommended)' is suggested.
      • Chunk Size/Concurrency, etc.: Advanced parameters for performance tuning. Defaults are usually fine.
    • Glossary:
      • Upload Glossary (Optional): Upload a CSV file (with 'src' and 'dst' columns) to ensure consistency and accuracy for specific terms.
      • Auto-generate Glossary: If enabled, the program will first extract terms from the source text to create a glossary before proceeding with the translation.
  3. Step 3: Upload File

    In the task list on the right, click or drag your document into the file upload area.

  4. Step 4: Start Translation

    Once the file is selected, click the Start Translation button on the task card. The process will begin, and you can monitor real-time progress in the log area.

  5. Step 5: View and Download

    After translation is complete, action buttons will appear on the task card:

    • Preview: Opens a side panel for a side-by-side comparison of the original and translated text.
    • Download: Download the translated document in various formats, including PDF, DOCX, and Markdown.
    • Attachments: If any additional files were generated (like an auto-generated glossary), download them here.
Important: All settings are automatically saved in your browser for future use. You can also use the new 'Export Config' and 'Import Config' buttons to back up and restore your settings.
", - "tutorialUnderstandBtn": "I understand", - "contributorsModalTitle": "Thanks for Contributing", - "contributorsPara1": "DocuTranslate is an open-source project! Your needs and usage drive its progress.", - "contributorsPara2": "A heartfelt thank you to everyone who has sponsored the project, submitted code, provided valuable suggestions, or starred the repository!", + "tutorialModalBody": "

Video tutorials are available on Bilibili by searching for docutranslate.

Welcome to DocuTranslate! Follow these steps to translate your documents:

  1. Step 1: Select Workflow

    At the top of the left-side settings panel, choose the processing flow that best suits your file type.

    Tip: 'Auto-select workflow' is enabled by default. Just upload your file, and the system will automatically match the appropriate workflow for you, simplifying the process.
    • To Markdown then Translate: Suitable for translating PDFs, markdown files, images, etc. This is the most versatile and powerful mode.
    • Plain Text Translation: For translating .txt files.
    • EPUB Translation: For translating .epub e-books.
    • DOCX Translation: For translating .docx Word documents.
    • XLSX Translation: For translating .xlsx or .csv spreadsheets.
    • SRT Subtitle Translation: For translating .srt subtitle files.
    • ASS Subtitle Translation: For translating .ass advanced subtitle files.
    • JSON Translation: For translating specific fields within .json files.
    • HTML Translation: For translating .html web files.
  2. Step 2: Configure Parameters

    After selecting a workflow, relevant configuration options will appear below. Please complete the settings in order (all configurations are automatically saved in your browser):

    A. Workflow-Specific Options (appear based on your choice in Step 1):

    • If 'To Markdown then Translate' is selected, configure Parsing Configuration:
      • Parsing Engine: Choose an engine to convert your file (like a PDF) into a translation-friendly Markdown format. If your file is already Markdown, no selection is needed.
      • Mineru Token: If you select the minerU engine, you must enter your token here.
    • If 'Plain Text/DOCX/XLSX/SRT/ASS/EPUB/HTML' is selected, configure its Translation Options:
      • Insert Mode: Defines how the translation result is placed in the document. You can choose to 'Replace' the original, 'Append' after it, or 'Prepend' before it.
      • Separator: When 'Append' or 'Prepend' is selected, this is used to insert a separator between the original and translated text (e.g., \\N is common for ASS, <br /> for EPUB).
    • If 'JSON Translation' is selected, configure JSON Paths:
      • JSON Paths to Translate: Enter one JSONPath expression per line to translate all strings within matching objects. E.g., $.* (translates all strings), $..description (translates all values with the key description).

    B. General Options (apply to all workflows):

    • Translation Model:
      • Select Platform/API Address/API Key/Model ID: Configure the AI translation service you wish to use. The stronger the model's instruction-following capability, the lower the chance of errors or missed translations.
      • Skip Translation: If checked, only document parsing and format conversion will be performed, without calling an AI for translation.
    • Translation Configuration:
      • Target Language: Specify the language to translate to.
      • Custom Prompt: Optional, add extra instructions, like 'Keep personal names in their original language.'
      • Thinking Mode: For models that support hybrid inference; 'Disable (Recommended)' is suggested.
      • Chunk Size/Concurrency, etc.: Advanced parameters to tune performance and API request behavior. The defaults are usually fine.
    • Glossary:
      • Upload Glossary (Optional): Upload a CSV file (must have 'src' and 'dst' columns) to ensure consistency and accuracy for specific terms.
      • Auto-generate glossary: When enabled, the program first extracts terms from the original text to generate a glossary, then proceeds with the translation.
  3. Step 3: Upload File

    In the task list on the right, click or drag your document into the file upload area.

  4. Step 4: Start Translation

    Once the file is selected, click the Start Translation button on the task card. The system will begin processing, and you can watch the real-time progress in the log area.

  5. Step 5: View & Download

    After the translation is complete, action buttons will appear on the task card:

    • Preview: Compare the original and translated text side-by-side in a slide-out panel.
    • Download: Download the translation in various formats, including PDF, DOCX, and Markdown.
    • Attachments: If any additional files were generated (like an auto-generated glossary), download them here.
Important Note: All settings are automatically saved locally in your browser for your convenience. You can also use the new 'Export Config' and 'Import Config' buttons to back up and restore your settings.
", + "tutorialUnderstandBtn": "I Understand", + "contributorsModalTitle": "Thank You for Contributing", + "contributorsPara1": "DocuTranslate is an open-source project! The community's needs and usage are the driving force behind its progress.", + "contributorsPara2": "A heartfelt thank you to everyone who has sponsored the project, submitted code, provided valuable suggestions, and starred the repository!", "contributorsWelcome": "You are welcome to contribute in the following ways:", - "contributorsGithub": "GitHub Home", + "contributorsGithub": "GitHub Page", "contributorsPR": "Submit a Pull Request", "contributorsIssue": "Report an Issue", "contributorsQQ": "Or contact the author via QQ Group: 1047781902", "glossaryModalTitle": "Current Glossary", "glossaryTableSource": "Source (src)", "glossaryTableDestination": "Destination (dst)", - "glossaryEmpty": "The glossary is empty.", "init_i18n_failed_alert": "Failed to load interface translations. Please check your network connection or contact an administrator.", "init_failed_alert": "Initialization failed, could not connect to the backend service. Please ensure the service is running and refresh the page.", - "status_selectFileFirst": "Please select a file first", - "status_fillRequired": "Please fill in all required fields", + "apiHrefInfo302ai": "👈 Register through this link to enjoy a $1 free credit", + "glossaryEmpty": "The glossary is empty.", + "engineOptionIdentity": "Already Markdown", + "engineOptionMineru": "Mineru (Recommended)", + "engineOptionDocling": "Docling", + "engineOptionMineruDeploy": "Mineru Deploy Service", + "preview_loading": "Loading preview...", + "preview_cantReadOriginal": "Could not read original file content.", + "preview_cantPreviewType": "Cannot preview this file type", + "preview_noOriginalCache": "No original file cached for preview.", + "preview_loadFailed": "Failed to load preview.", + "pdf_preparing": "Preparing PDF, please wait...", + "pdf_print_failed": "Failed to invoke print function. Please check browser settings or save as PDF manually.", + "pdf_fetch_failed": "Failed to fetch translated content, cannot generate PDF.", + "preview_bilingual": "Bilingual Preview", + "preview_translatedOnly": "Translation Only Preview", + "admin_tasklist_failed": "Admin Mode: Failed to fetch task list.", + "status_selectFileFirst": "Please select a file first!", + "status_fillRequired": "Please fill in all required fields!", "btn_initializing": "Initializing...", - "status_encodingAndSubmitting": "Encoding and submitting request...", - "status_requestOk": "Request successful, task started", - "btn_cancelTranslation": "Cancel Translation", + "status_encodingAndSubmitting": "Encoding and submitting task...", + "status_requestOk": "Request successful, task has started.", + "btn_cancelTranslation": "Cancel", "status_requestFail": "Request failed", - "status_initFail": "Initialization failed", + "status_initFail": "Task initialization failed", "status_cancelling": "Cancelling...", - "status_cancelSent": "Cancellation request sent", + "status_cancelSent": "Cancellation request sent.", "status_cancelFail": "Cancellation failed", "status_gettingStatus": "Getting status...", "btn_reTranslate": "Re-translate", "status_updateError": "Error updating status", - "preview_loading": "

Loading preview...

", - "preview_cantReadOriginal": "Could not read original file content.", - "preview_cantPreviewType": "Cannot preview this file type", - "preview_noOriginalCache": "No cached original file to preview.", - "preview_loadFailed": "

Failed to load preview

", - "pdf_preparing": "Preparing PDF, please wait...", - "pdf_print_failed": "Failed to initiate print. Please try right-clicking the preview and selecting 'Print'.", - "pdf_fetch_failed": "Failed to fetch translated content, cannot generate PDF.", - "preview_bilingual": "Bilingual Preview", - "preview_translatedOnly": "Translated-Only Preview", - "admin_tasklist_failed": "Admin mode: Failed to load task list.", - "configImportSuccess": "Configuration imported successfully! The interface has been updated.", - "configImportError": "Import failed. Please check if the configuration file is a valid JSON format." + "configImportSuccess": "Configuration imported successfully! Settings have been applied.", + "configImportError": "Import failed! Invalid file format or content." } } \ No newline at end of file diff --git a/docutranslate/static/index.html b/docutranslate/static/index.html index 4563daf..535549e 100644 --- a/docutranslate/static/index.html +++ b/docutranslate/static/index.html @@ -1 +1 @@ - DocuTranslate - 交互式文档翻译

DocuTranslate

如果上传的文件本身是.md格式，此项可不选。
mineru VLM是更新的内测模型。

Base URL:

选择一个或多个CSV文件。文件需包含'src'和'dst'两列标题，分别代表原文和译文。

GitHub主页(欢迎star❤):
https://github.com/xunbu/docutranslate

交流QQ群: 1047781902

version:

任务列表

LOGO

当前没有任务，点击“新建任务”开始吧！

预览
原文
译文
\ No newline at end of file + DocuTranslate - 交互式文档翻译

DocuTranslate

如果上传的文件本身是.md格式，此项可不选。

Base URL:

选择一个或多个CSV文件。文件需包含'src'和'dst'两列标题，分别代表原文和译文。

GitHub主页(欢迎star❤):
https://github.com/xunbu/docutranslate

交流QQ群: 1047781902

version:

任务列表

LOGO

当前没有任务，点击“新建任务”开始吧！

预览
原文
译文
\ No newline at end of file