Update 2022.07.02.
I’ve updated the article by no longer recommending split and spread as the last step, mainly because I couldn’t always keep up with the schedule thus the order of the splits changed which resulted in me seeing subsequent chapters without first having a chance to process the preceding ones. I’m sure these scheduling problems are fixable but I’m not sure how to do it quickly and easily. So my recommendation is NOT to split and spread books just read them incrementally as a whole.
Intro
I’m a rather quick learner. However, I’ve been struggling with actually remembering the material for a longer period. Recently I’ve found Anki and SuperMemo. Both applications are wonderful, but ultimately I’ve chosen SuperMemo, mainly because of its Incremental Reading feature. I’m still very early in my journey to learn using it proficiently. (SuperMemo is a Windows-only software that I’m using with Parallels Desktop for Mac)
I’m planning to read a lot of books with incremental reading but one of the first problems I’ve encountered is that SuperMemo doesn’t support .epub natively. This short guide presents my findings on importing a book into SuperMemo.
Main steps
- Have a DRM free
epub - Import it to Calibre and convert it from
epubtoepub - Convert to
html - Clean up
htmland optimize images - Import to SuperMemo
- Fix images, set up references
- Adjust priority
Have a DRM free .epub
In my experience, most of the epubs out there are suitable candidates for conversion. There are also ways to remove DRM from az epub, but due to my understanding, it’s a grey area.
I’ve wanted to import the official Swift book, The Swift Programming Language. This book is written by Apple and is available for free in the official Apple Books application. Exporting from Books is as simple as dragging the book to Finder. However, I had a small hiccup importing this book to Calibre because Apple uses a non-standard epub format. Erica Sadun has found a simple solution to resolve this issue.
Import to Calibre and convert
Calibre is a wonderful piece of software full of well-written features. I’ve found that even if I already have an .epub it might be worth converting it from epub to epub as it seems to be producing a somewhat cleaner result.
Convert to html
Having an .epub then from Calibre, I’m using Pandoc, the universal document converter to convert it to html. Install Pandoc on macOS via Homebrew:
1
brew install pandoc
SuperMemo’s support for styles and images is somewhat limited. Even if the .epub has some fancy fonts and colors it would probably look bad in SuperMemo. This is why we have to take a minimalistic approach during the conversion.
1
pandoc -M document-css=false -s --extract-media=Apple-TheSwift55ProgrammingLanguage Apple-TheSwift55ProgrammingLanguage.epub -o Apple-TheSwift55ProgrammingLanguage.html
This command upon success will generate an html file and a folder full of images.
Clean up html and optimize images
Opening the html file in our favorite editor we can see a lot of noise/clutter that seamingly doesn’t provide anything valuable. I got rid of them by searching for the following regular expressions and replacing them with an empty string:
( class="[^"]+")and(class="[^"]+")( id="[^"]+")and(id="[^"]+")
As a final step before the import, I ran all images through ImageOptim saving a little space without compromising on the quality.
Import to SuperMemo
- Copy the
.htmlfile and the folder full of images to the root ofC:\. - Open the
.htmlfile with Internet Explorer (😢) - Select all (
Ctrl + A) and copy (Ctrl + C) - Open SuperMemo, create a new article/topic (
Ctrl + N) and paste (Ctrl + V) - Remove anything that you deem unnecessary. (Table of Contents, Index, Acknowledgements, etc.)
- Using
Alt + Qset the Reference fields accordingly. For me, the bare minimum is to set the- author
- title
- date
Fix images, set up references
The book is imported, we could jump to the next, final step, however, the image references are currently pointing to the root where we’ve copied the folder. (e.g.: C:/Apple-TheSwift55ProgrammingLanguage/) If the images are removed from there they won’t show up in SuperMemo again. I haven’t found a standard solution for this, but I wanted to know these images in safety so I did the following.
- Depending on where SuperMemo is installed (for me it’s
C:/SuperMemo) insidesystems, next to my collection I’ve created a newImagesfolder. - I’ve copied all the images here:
C:/SuperMemo/systems/Images/Books/Apple-TheSwift55ProgrammingLanguage/ - The last step is to open the book’s
.HTMfile from SuperMemo in an external editor (Ctrl + F9) and replace the image paths.
1
2
file:///C:/Apple-TheSwift55ProgrammingLanguage/
file:///C:/SuperMemo/systems/Images/Books/Apple-TheSwift55ProgrammingLanguage/
Later on it’s recommended to gradually import the images into image components (that will be saved to the image registry) instead of leaving it in the html with reference.
Parallels Desktop for Mac is amazing, somehow it integrates the Mac apps into Windows’s interface. I could set Visual Studio Code (Mac) as a default app for
.HTMfile extension on Windows!
Adjust priority
After fixing up everything one last step remains. Adjusting the book’s priority. It’s done by pressing Alt + P then specifying either a percentage or a position. Please note that position 1 equals to 0% that is 0% means this topic is top priority.
Split & Spread
The only case where I had to do splitting, either way, was when the imported book was so massively big that it caused performance issues with SuperMemo and the UI became unresponsive for a long time every time the book was being opened.
I’ve mostly followed Pleasurable Learning SuperMemo’s great step-by-step video to do this.
One of the most useful tricks I’ve learned is to NOT to use SuperMemo’s default splitting as it messes up the topics a little. Instead specify a custom tag, that’s usually <h1>, <h2>, <h3>, <h4> with epubs converted to html.
Splitting up this way is a little cumbersome, but doable in a relatively short time. After the split, we can spread the material according to our schedule.