Unofficial fork of SumatraPDF · 非官方社区增强版
Not affiliated with sumatrapdfreader.org · 与官方站点无关
A Windows document reader (PDF, EPUB, MOBI, and more) with improvements for Chinese ebooks, offline lookup, themes, smart PDF dark mode, and local OCR.
Windows 下的 PDF / 电子书阅读器，针对中文 EPUB/MOBI、离线查词、主题、PDF 智能暗黑模式与本地 OCR 等做了增强。
- Repository / 本仓库： source code for GPLv3 compliance · 满足 GPLv3 源码随同分发要求
- User guide (Chinese) / 中文详细说明： readme.txt
- Upstream / 上游项目： https://github.com/sumatrapdfreader/sumatrapdf
- License / 许可： COPYING (GPLv3) + AUTHORS
Highlights / 重点
- OCR — offline RapidOCR / PP-OCR for scanned pages. Auto OCR while reading, recognize all pages, save a searchable PDF, or drag a region; then search, select, look up, read aloud, and extract bookmarks. Keep
{exe}\ocr\.
OCR — 扫描页本地识别（RapidOCR / PP-OCR）。翻页自动识别、全文识别、保存可搜索 PDF、框选识别；之后可搜索、选中、查词、朗读、提取书签。模型在{exe}\ocr\。 - Smart PDF dark mode — Original / Match theme for PDF and ebooks; toolbar Light / Dark (Light-Warm, Light-White).
智能暗黑 — 原稿 / 匹配主题（PDF、EPUB、MOBI 等）；工具栏亮/暗主题（暖色护眼、中性浅色）。 - Read Aloud (TTS) — word-by-word highlight; start from top, cursor, or selection; pause/continue; voice and speed 0.25×–2.0×.
朗读 (TTS) — 逐词高亮；从页首/光标/选中开始；暂停/继续；可选语音与语速。 - Offline dictionary — double-click a word;
.idx/.datin{exe}\dict\.
离线查词 — 双击词语；词典放在{exe}\dict\。 - Chinese EPUB/MOBI — mixed text/image layout; TOC navigation; faster open for large EPUB/MOBI/AZW.
中文 EPUB/MOBI — 图文混排与目录跳转；大型文件打开更快。 - Selection toolbar — highlight, underline, strike out, Ask AI on PDF selection.
划词工具栏 — 选中后高亮、下划线、删除线、Ask AI。
Also / 其他： per-monitor DPI (sidebar, toolbar, menu, tabs, TOC search; tab-detach font fix) · session restore no longer crashes on double-click open · Win11-style caption and theme-aware chrome.
另有： 双屏 DPI（书签/工具栏/菜单/标签/目录搜索字号；拖标签分窗字体）· 恢复会话后双击打开不再崩溃 · Win11 标题栏与主题配色。
Supported formats / 常见格式： PDF（含扫描件 OCR）、EPUB、MOBI、AZW/AZW3、FB2、CHM、CBZ/CBR、DjVu、XPS 等。Scanned PDF, DjVu, images, and comics can be recognized locally.
扫描 PDF、DjVu、图片和漫画可本地 OCR。
Product name: Sumatra PDF Plus. Executable: SumatraPDF-Plus.exe.
显示名称：Sumatra PDF Plus，可执行文件为 SumatraPDF-Plus.exe。
- Extract the full folder to any path (avoid special characters in the path). 解压整个文件夹到任意目录（路径尽量不要含特殊符号）。
- Run
SumatraPDF-Plus.exe. Keep thefontsfolder beside the exe. 双击SumatraPDF-Plus.exe运行；不要只复制 exe，请保留同目录的fonts文件夹。 For Traditional Chinese PDF lookup, also keep theopenccfolder.
繁体 PDF 查词还需保留opencc文件夹。 For scanned-page OCR, keep theocrfolder (onnxruntime.dlland.onnxmodels).
扫描页 OCR 还需保留ocr文件夹（含onnxruntime.dll与模型）。 - Use the toolbar to toggle light/dark theme and word lookup. 可通过工具栏切换亮/暗主题和查词开关。
- Click Auto OCR to recognize scanned pages as you view them, or use the dropdown for recognize-all / save / region. 点击自动 OCR 可在翻页时识别扫描页；下拉菜单可全文识别、保存可搜索 PDF、或框选识别。
- Click the speaker icon (after OCR) or open Read Aloud (TTS) in the menu bar to read aloud. 点击工具栏喇叭图标（在 OCR 之后），或菜单栏 Read Aloud (TTS) 开始朗读。
Recognize scanned pages (little or no text layer) so you can select, search, look up words, read aloud, and extract bookmarks. Runs offline with RapidOCR / PP-OCR Chinese mobile models in {exe}\ocr\.
识别几乎没有文字层的扫描页，便于选择、搜索、查词、朗读和提取书签。使用 {exe}\ocr\ 里的 RapidOCR / PP-OCR 中文模型，完全离线。
How to use / 使用方法
- Toolbar / 工具栏 — Auto OCR (before the speaker). Click to toggle recognition as you view or search scanned pages (default off). Dropdown: Recognize all pages, Recognize all pages and save, OCR region.
自动 OCR（在喇叭按钮前）。单击后翻页即识别扫描页（默认关）。下拉：全文识别、全文识别并保存、框选识别。 - Recognize all pages / 全文识别 — re-OCRs every page even if a text layer already exists (garbled dual-layer PDFs). Results stay in memory for this session. If there is no outline, bookmarks are extracted after OCR.
即使已有文字层也会再扫一遍（乱码双层 PDF）。结果先留在本次打开的文档里。没有目录时识别后提取书签。 - Recognize all pages and save / 全文识别并保存 — scan with progress, then overwrite the current PDF (no Save As). Prompts before replacing an existing text layer or outline. PDF only.
显示进度后直接覆盖当前 PDF，不弹另存。已有文字层或目录时先询问。仅 PDF。 - OCR region / 框选识别 — drag a rectangle; text is copied.
拖选矩形，识别结果复制到剪贴板。 - Cancel / 取消 — stop queued jobs; pages already done are kept.
取消排队中的页；已识别的页保留。
Keep the ocr folder next to the exe. Details: docs/md/OCR.md.
请保留 exe 同目录的 ocr 文件夹。详见 docs/md/OCR.md。
Advanced setting: AutoOcrScanPages (default false).
高级设置：AutoOcrScanPages（默认 false）。
Read documents aloud with word-by-word highlighting synced to speech. Works on PDF, EPUB, MOBI, and other formats with extractable text.
逐词高亮跟随朗读进度，支持 PDF、EPUB、MOBI 等可提取文本的格式。
How to use / 使用方法
- Toolbar / 工具栏 — speaker icon after Auto OCR; click to start/pause/continue; dropdown for voice, speed, and start options
OCR 按钮后的喇叭；单击开始/暂停/继续；下拉菜单可选语音、语速与起始方式 - Menu bar / 菜单栏 — Read Aloud (TTS) → start from top, cursor, or selection; Voice and Speed submenus
Read Aloud (TTS) → 从页首/光标/选中开始；Voice 选语音、Speed 选调速 - Right-click / 右键 — Start Reading From Cursor Position; Pause Reading / Continue Reading while active
从光标处开始朗读；朗读中可暂停朗读 / 继续朗读 - Speed presets / 语速 — 0.25×, 0.5×, 0.75×, 1.0×, 1.25×, 1.5×, 2.0×
Natural voices / 自然语音
Default Windows SAPI voices (e.g. Huihui, Kangkang, Yaoyao) sound robotic. For natural speech, install NaturalVoiceSAPIAdapter and configure local Narrator voices or Edge online voices — see docs/md/Read-Aloud-TTS.md.
系统默认 SAPI 语音（如慧慧、康康、瑶瑶）较机械。自然语音需安装 NaturalVoiceSAPIAdapter，详见 docs/md/Read-Aloud-TTS.md。
Advanced settings: ReadAloudVoiceId, ReadAloudSpeakingRate.
高级设置：ReadAloudVoiceId（语音）、ReadAloudSpeakingRate（语速）。
Dictionary code is in this repo;
查词功能代码在本仓库；
- Create
{exe}\dict\next to the executable. 在 exe 同目录建立dict文件夹。 - Add dictionary files, for example: 放入词典文件，例如：
SumatraDict.idx/SumatraDict.dat /SumatraDictAudio.dat— English–Chinese · 英汉SumatraDictZh.idx/SumatraDictZh.dat— Chinese dictionary · 汉语
- Enable the dictionary button on the toolbar, then double-click a word in the document. 打开工具栏「查词」按钮，在正文中双击词语即可。
Override the dictionary folder with the OfflineDictionaryPath advanced setting.
可用高级设置 OfflineDictionaryPath 指定其他词典目录。
Traditional Chinese PDF lookup / 繁体 PDF 查词 — keep the opencc folder next to the exe (OpenCC t2s data). Upgrading only the exe without opencc disables Traditional lookup.
繁体 PDF 查词需保留 exe 同目录的 opencc 文件夹；只升级 exe 将无法繁体查词。
Settings are stored in SumatraPDF-settings.txt:
- Portable / 便携版: next to
SumatraPDF-Plus.exe - Installed / 安装版:
%LOCALAPPDATA%\SumatraPDF\SumatraPDF-settings.txt
Annotated reference / 带中文注释的参考模板: SumatraPDF-settings-annotated.txt
Copy only the lines you need into your live settings file. Comments must be on their own # lines (inline comments after values are not supported).
只复制需要的配置行到真实配置文件；注释必须单独成行，不能写在值后面。
User guide (Chinese) / 中文说明: readme.txt
Requirements: Visual Studio 2022 build tools, Bun.
bun ./cmd/build.ts # debug → out/dbg64/SumatraPDF-Plus.exe
bun ./cmd/build-all.ts # release → out/rel64/See AGENTS.md for contributor notes.
贡献者说明见 AGENTS.md。
Bug reports and pull requests are welcome on this repository only — not on the official SumatraPDF issue tracker.
欢迎在本仓库提交 Issue 和 PR，请勿发到官方 SumatraPDF 项目。
When reporting problems, please include the file type, book title (if ebook), and steps to reproduce.
反馈问题时请说明文件类型、书名（电子书）和复现步骤。
Sumatra PDF Plus is a community modification for testing and personal use.
Sumatra PDF Plus 为基于 GPLv3 的社区修改版，供测试与个人使用。
SumatraPDF is a multi-format reader for Windows under (A)GPLv3, with some code under BSD license (see AUTHORS).