[shotsky]

I use Calibre to convert from various ebook formats to 'htmlz' at this point, which is the only way I know of to get all the html files into a single file. Unfortunately, Calibre replaces a lot of classes with its own made up classes, which 'breaks' the meaning of various existing classes. All I really need is one single html file, unmodified, and the photos. The htmlz export joins the files, but it converts the classes as well, even though there is no need to do so.
I've just installed the KindleUnpack plugin, and it does exactly what I need - merge the html files into one file, without using Calibre's conversion routines, which renames classes from whatever they were to 'calibexx', which causes a loss of the original class names.
The only problem I still have is that I need the same feature for Epub files. Would it be impossible to add that as an import format?
What would be even better would be to add a 'unpack' feature to Calibre itself, which would take any input format, merge the html files, and provide a zipped output.
So, my three most important needs are:
1. Merge the html files into a single file
2. Not change or rename classes.
3. Run from the command line.
Thanks!