wxReader

wxReader is a lightweight, high-performance document reader built with wxWidgets (wxPython), MuPDF (PyMuPDF), OpenGL (PyOpenGL), libvips (pyvips), and Python.

wxReaderVoiceCtrl is a hands-free external controller app for wxReader, recognizing voice command offline by vosk and pyaudio. GUI is built with wxPython.

wxReaderEyeTrackCtrl is a hands-free external controller app for wxReader, recognizing eye movement and blinking gesture offline by EyeTrax(https://doi.org/10.5281/zenodo.17188537) through a webcam. GUI is also built with wxPython.

Version 1.3.4 Add customizable keyboard shortcuts. Add GUI control on uStrength of shaders. Add a folder tab in the sidebar. Gallery and cache performance optimized.

~Version 1.3.7 External controlled page tuning, and add example apps of voice and eye gesture command. UI/UX enhancements. Bugs fixed. 7z format support. Cache optimizations.

Version 1.3.8 Add internationalization and localization for ja_JP, zh_TW and zh_SG.

---

Features

Feature	Description
File Format Support	PDF, EPUB, ePub, ZIP, CBZ (manga or comics archive files).
Paging Strategy	Single Page and Two-Page spreads (with optional blank start page). Supports Right-to-Left (RTL) reading.
Content Extraction	Extract text and images directly from specific pages to the clipboard or disk.
Image Processing	Real-time built-in filters and enhancements. Support custom OpenGL frag shaders.
Zoom and View	Fit Width, Fit Page, Manual Zoom, Fullscreen Mode, and customizable background color.
Navigation	Sidebar with Outline (TOC) and File Browser tabs. Text search dialog, and a standalone TOC dialog.
Gallery Mode	View all thumbnails or frontpages of all books inside a same folder like a gallery.
File History	Automatically saves recent files and reading progress on close.
Interaction	Drag-and-drop file loading and full keyboard operation support.
External control	(Default off.) Send command+token to a local UDP port to do page turning.

Accessibility features: The two independent external control programs, wxReaderVoiceCtrl and wxReaderEyeTrackCtrl, respectively provides offline voice-activated page turning commands and offline eye gesture based page-turning via computer camera recognition, triggered by “blinking twice while looking at a designated area of the screen.” (The are fully offline and requires a not-potato CPU.)

---

Installation

1. From sauce (latest dev):

   1. Install Python 3.12 and uv, uv --project . sync.
   2. Download and put the libvips shared library *.dll files inside .\src.
   3. Sync dependencies with uv.
   4. Build with .\build.bat on Windows x64. (Run the build script inside /src folder.)
   5. (Notes: upgrade pymupdf will fail the compilation, and this is a Nuitka issue.)
   6. Optional: uv --project .\extctrl\voice sync, uv --project .\extctrl\eye_track sync, then build with .\build_*.bat.

2. Pre-compiled binary (stable): portable .exe files are provided on the Releases page. Here are links to download (for win10+ x86_64):

   1. wxReader
   2. wxReaderVoiceCtrl
   3. wxReaderEyeTrackCtrl

3. Upgrade from older versions: simply copy the wxReader.cfg and pswd.txt files containing all user settings to the folder of a new version. You (probably) can also just unzip and overwrite existing files in the old folder.

TODO

• keys for navigation can conflict with input fields