Hacking an e-Reader with React

React for websites is cool and all, but how about we use it for eBooks? I'm here to show you how to hack an eReader to make it show whatever you want, such as a tea menu. My name is Tiger Oaks. I'm a software engineer working at Microsoft. I previously worked on three different web browsers. You can find me at Mountainwoods on all the socials, as in, not a certain golfer.

Besides for my cool name, if there's one thing to know about me, it's that I like tea quite a lot. I've bought tea from different shops around the world, and I've gotten lots of tea as gifts from friends, and I just went to a tea festival last month. Basically, I now have about 30 different tea flavors in my apartment. But when I have guests over, I just have too many teas to list. By the time I get about halfway through, their eyes have already glazed over. The solution? Putting my collection onto a menu. Now, I can just hand guest a list that they can peruse as they like. Rather than reprinting a piece of paper whenever my collection changes, I can just update an e-reader. Now, I'll show you how you can do this yourself, creating an image and e-book with React, and then loading them onto the e-reader.

Let's start by breaking down how to make an image to later display. First, we start with a background to make the menu look a little nicer. I took inspiration from a Dungeons and Dragons-themed tea shop and created a custom background on the paper texture.

Then, I moved on to creating the foreground text using SVG. We frequently use SVG to create vector art on our websites, and it resembles HTML and lets us use React and CSS. For a single tea flavor in the menu, I created text elements for the tea name and description, as well as a line to separate them. I then wrapped it in an SVG group, like a div.

I wrote a small program in TypeScript to read the database with my tea, and then write SVG files using React. This SVG snippet has been turned into a React component, so you can translate a data object into a rendered image using text elements and SVG positioning. Then I can move each of these groups using transform-translate, like a CSS transformation or using position-absolute. By setting different x and y values, I can reuse the same text and line elements, but move them around the image.

With the top section nicely laid out, I can then wrap all these elements in a group and repeat the same process, running the component with a different translation for the bottom section. SVG is very flexible, so you can make your own customizations. You could switch out the font to CSS or display different number of items and show a menu for other things in your house. It's also easy to swap the background image around to anything you want. For an e-reader, I convert the SVG file into a png image, which you can do using free tools like Inkscape and Squoosh. Some e-readers let you just upload a custom lockscreen image, I can connect it to my computer, then drag this image onto its hard drive in a screensaver folder. But other e-readers do not have this feature. However, most of them will show the cover of whatever book you're reading when locked. We can abuse this by writing our own e-book, using the menu image as the cover of the book.

Now, when I say e-book, you probably think of a digital version of a big book like this with lots of characters. But our e-book is going to be more like a pamphlet. If you have at least one page, you can write an e-book. E-books are written in a format called EPUB. This is a standardized format used by all e-readers except Kindles, but Amazon lets you easily convert from EPUB into Kindle's custom format. EPUBs are actually just zip files that have their extension renamed. Just like websites, e-books need to be responsive since you can have different e-reader screen sizes. As a result, writing an e-book is very similar to writing a website. You use familiar HTML, CSS, and images to create each chapter of the book. In a website, each HTML file is a page on your site.

But in an e-book, each HTML file is a chapter of your book. EPUB uses a stricter version of HTML called XHTML. You need to make sure that your self-closing tags have a slash and that your tag names are all lowercase. This might sound familiar, because these are the exact same restrictions that JSX has. But other than that, everything is the same.

You can include CSS and use the same tags like h1, and p, and image. EPUBs also need a file called content.opf, which acts like a sitemap. This file includes data like the title of the book, every file in a zip folder, the book chapters in order, and bookmarks to some important pages. Here's an example. The content file is written in XML. Let's hide the unimportant parts. You can see it broken down into a few tags, similar to how an HTML file is broken down into head and body tags. Metadata includes the title of the book and the cover image. Manifest includes all the files in the book. Some files have been assigned an ID that can be referenced in upper sections. Spying indicates the order of the chapters. And guide points to important pages like the table of contents.

You probably don't want to deal with maintaining a file like this, but don't worry. Just like HTML has static site generators to simplify writing a website, there are EPUB generators to simplify writing ebooks. I used Pandoc, a program that can convert a markdown file into an EPUB file. The front matter, at the top of the markdown file, is used to populate the EPUB metadata. Each top level heading in a markdown also gets converted into an EPUB chapter. So this makes it easy to include images and descriptions of each t-flare as EPUB chapters. Then I ran it all with Dino, which is a really useful tool to run JavaScript on your desktop, like node, but with many more built-in features. It has built-in support for TypeScript and JSX, so I didn't need a build step. It also included the built-in lenture formatter so I didn't need ESLint or Prettier. And it has a huge standard library, so you don't even need that many npm modules. I switched over to using Dino when writing monast scripts, because it's just so much easier to quickly build something with it.

With that, we have all the building blocks you need to create a menu like this yourself. You can also find all the code I wrote on GitHub. You might ask, why go through all this trouble when you could say, use a pen and paper? Well, it's more fun. This is also a great learning process for me to better understand how e-readers work, and whenever we change our tees out, it's much faster to just re-run the script. Thank you all for listening, and feel free to get in touch with me if you have any questions.