It's been 10 years since the definitive thread on this topic was published https://www.mobileread.com/forums/sh...ad.php?t=53311 Recently. trying to convert a mobi to an epub produced the same error, but with a different root cause than large pre.../pre blocks.
Turns out this book had been generated using blockquote tags to chunk the input. Just why blockquote tags were be used, especially without the cite modifier, is a mystery. Perhaps some book generator does that.
My fix was to run calibre in debug mode, and edit the index.html in the input subdirectory as described in the article referenced above. However, instead of looking for long pre blocks to chop up, use an editor to replace all instances of /blockquote with /div, and blockquote with div,in that order. Now use Sigil as discussed in the thread to build an epub from the modified index.html. When you have this epub, you can use Calibre to fill in the metadata by cutting and pasting fields from the opf.html file. (There's probably a way to tell Sigil to use the original opf data, but I didn't bother digging that far.)
Now you have the book as an epub.
Turns out this book had been generated using blockquote tags to chunk the input. Just why blockquote tags were be used, especially without the cite modifier, is a mystery. Perhaps some book generator does that.
My fix was to run calibre in debug mode, and edit the index.html in the input subdirectory as described in the article referenced above. However, instead of looking for long pre blocks to chop up, use an editor to replace all instances of /blockquote with /div, and blockquote with div,in that order. Now use Sigil as discussed in the thread to build an epub from the modified index.html. When you have this epub, you can use Calibre to fill in the metadata by cutting and pasting fields from the opf.html file. (There's probably a way to tell Sigil to use the original opf data, but I didn't bother digging that far.)
Now you have the book as an epub.