Polar Bookshelf is a PDF manager created using Electron framework and PDF.js with added support for incremental reading, pagemarks, and progress tracking.

Pagemarks are a new proof of concept reading style inspired from incremental reading. Essentially they allow suspend and resume of reading for weeks and months in the future until you're ready to resume, without losing your place. This works even if you jump around in a book (which is often in technical or research work).

The long term goal is to implement the following functionality:

• Automatic metadata extraction of thumbnails, marked up text, and notes and support migration into 3rd party spaced repetition systems like Anki.

  • We have a proof of concept of this using chrome headless which needs to be ported here.

  • Integration of https://github.com/burtonator/pdf-annotation-exporter to enable this functionality.

• Additional annotation types like area highlight, text highlight, plus a complex feature set like notes and tags for these objects.

• Flashcards will be notes designed and marked specifically to be converted to flashcards for usage in spaced repitition systems. They will have extended metadata including front and back of the card as well as support for cloze deletions, and the ability to include extended metadata like the source text to which the flashcard was attached.

• Fully distributed. You control your notes. You can export them to Evernote, Google Drive, etc but Polar keeps track of your notes for you.

• Distributed collaboration with other Polar users.

• Tagging system and the ability to perform advanced functions on the tags.

• Native cloud sync.

• Management UI for all the notes you've worked on (editing, changing them, adding metadata, etc).

• Ability to pull down ISBN metadata

• Abilty to pull down metadata by academic paper ID using various platform APIs.

We believe the following design principles are core to seeing this as a successful project.

• All the data should support long term file formats. The on disk format we use is JSON.

• Portability to all platforms is critical. We're initially targeting Linux (Ubuntu), MacOS, and Windows. You shouldn't have to pick a tool, which you might be using for the next 5-10 years, and then get stuck to a platform which may or may not exist in the future.

We currently only support pagemarks but this is the main functionality I wanted implemented.

The pagemarks are persisted on disk in your ~/.polar directory and when you re-open a PDF your pagemarks are restored.

Additionally there is a progress bar that tracks the progress of the document based on the number of pagemarks you've created.

Right now usage is only via keyboard bindings (for pagemarks).

• Control Alt N - create a new pagemark on the current page
• Control Alt click - create a pagemark on the page up until the current mouse click
• Control Alt E - erase the current pagemark

• Meta-Command N - create a new pagemark on the current page
• Meta-Command click - create a pagemark on the page up until the current mouse click
• Meta-Command E - erase the current pagemark

Packages for MacOS and Linux are available in Releases

Windows builds are still alpha quality as I can't test them. The builds work and generate .exe setup files but I need feedback from the community on whether they work and how well they work.

Install NodeJS and npm for your platform.

$ git clone https://github.com/burtonator/polar-bookshelf
$ cd Electron-PDF-Viewer
$ npm install && npm start

./node_modules/.bin/electron --enable-remote-debugging --interactive --enable-console-logging .

PDF.js is available under Apache License. Electron is released under MIT License. Rest of the code is MIT licensed.