Converting digital DRM protected ebook to PDF and epub for offline reading

I recently bought a digital, DRM protected, ebook. Since it is protected, you are not allowed to download it, so I always have to use a browser. Since I prefer an ereader, I looked for legal ways to convert my book to a PDF or epub.

A simple approach and, according to ChatGPT legal approach, is the following:

1. Create screenshots of the book pages in the online reader
2. Bundle the screenshots to epub or pdf

Taking screenshots

Taking screenshots of 300+ pages is no fun, even on a Mac. I needed to press for each page:

• CMD + SHIFT + 5: It opens the screenshot tool. The first time, select the page area and specify the output folder <OUTPUT_FOLDER>. On subsequent times, the same area is selected.
• Press "ENTER" to save the page
• Press "Page Down" to move to the next page.

AppleScript to the rescue:

Press CMD+SPACE and type "Script Editor".

1. Create a new script
2. Paste below code
3. Press RUN
4. Quickly click the browser (so it receives keyboard focus for Page Down)

delay 3
tell application "System Events"
	repeat 322 times
		
		-- Open screenshot tool
		key code 23 using {command down, shift down} -- Cmd+Shift+5
		delay 1
		
		-- Press Enter (take screenshot)
		key code 36
		delay 1
		
		-- Go to next page (Page Down)
		key code 121
		delay 1
		
	end repeat
end tell

TIP: If you have the option to display the page in landscape format, e.g. rotate page 90 degrees counter-clockwise, the image resolution is better.

Tools you need (on Mac)

brew install imagemagick ocrmypdf ghostscript
brew install --cask calibre

Post-processing

cd <OUTPUT_FOLDER>

# Rename to page_000.png
i=0; for f in *.png(.n); do mv "$f" "$(printf "page_%03d.png" "$i")"; ((i++)); done

# Prepare outputs
mkdir -p rotated/cropped

# Rotate
mogrify -path rotated/ \
  -rotate 90 \
  -density 300 \
  -format png \
  *.png

cd rotated

# Create regular PDF
magick *.png \
  -units PixelsPerInch \
  -density 300 \
  -resize 1158x1851! \
  input.pdf

# Add searchable text using ocrmypdf
ocrmypdf --optimize 3 --rotate-pages input.pdf ocred.pdf

# Compress
gs -sDEVICE=pdfwrite \
   -dCompatibilityLevel=1.4 \
   -dPDFSETTINGS=/ebook \
   -dNOPAUSE -dQUIET -dBATCH \
   -sOutputFile=output.pdf \
   ocred.pdf
 
# Do the same for the cropped version
mogrify -path cropped -crop 1158x1851+95+115 +repage *.png

cd cropped

# Create regular PDF
magick *.png \
  -units PixelsPerInch \
  -density 300 \
  -resize 1158x1851! \
  input.pdf

# Add searchable text using ocrmypdf
ocrmypdf --optimize 3 --rotate-pages input.pdf ocred.pdf

# Compress
gs -sDEVICE=pdfwrite \
   -dCompatibilityLevel=1.4 \
   -dPDFSETTINGS=/ebook \
   -dNOPAUSE -dQUIET -dBATCH \
   -sOutputFile=output.pdf \
   ocred.pdf

# To epub using calibre
ebook-convert output.pdf book.epub \
  --enable-heuristics \
  --no-default-epub-cover  

Alternatively, use docling to convert your pdf to markdown first:

PYTORCH_MPS_HIGH_WATERMARK_RATIO=0.8 \
PYTORCH_MPS_LOW_WATERMARK_RATIO=0.5 \
docling --from pdf --to md --image-export-mode embedded ocred.pdf