there used to be an internet explorer option to save a html page - some sort of save all- that would put all the images into a subfolder, alongside of a file with all the text in, but it did not call it rtf.
maybe your "books" are actually grabs of html versions , badly saved ?
there's no way that calibre, or any other program, can put humpty dumpty back together again without some saved info defining where the images need to go
|