[AlanHK]

I'm looking at a book that has a dozen images, all embedded inline as base64:

<img src="data:image/jpeg;base64,/9j/.....

It slows down editing the files, and takes more space than normal images, so I'd like to convert these to binary image files.

I can copy and paste each one to a file and run a base64 decode, but is there a faster way?

PS -- found a better way, still not ideal: merged the book into a single HTML file in Sigil, extracted that and open in Firefox. I can save each image as a normal image file. Have to manually import and replace the b64 images though.