Calibre actually does horrible things to the HTML source that hides in each epub. While Standard Ebooks do promote awesome HTML. If you run Calibre on it, it's ruined.
An epub is just a zip file. Open it up, look inside, see for yourself.
OK, I checked. Calibre did almost nothing to the epubs I downloaded and added to the Calibre library and then sent to my device from there (a KOBO Libra H2O). The only thing it did was to add a file `./META-INF/calibre_bookmarks.txt` to the zip archive. The HTML was completely untouched. Now obviously if you do some conversion it's going to be another mattter and it's possible that the Calibre checks for some things it considers incompatible with its reader and in those cases alters the HTML, but I didn't find anything like that in the books I checked.
Admittedly I don't think that Calibre should touch the contents of an ebook zip file without specific action from the user (such as requesting a conversion)... but what it does doesn't seem to be problematic.
* Read eBooks in just about any popular or semi-popular format
* Manage your library
* Convert books to and from any of the supported formats, specifying details of that conversion like tables of content, typography, chapter headings, etc.
* Edit book metadata
* With a plugin, strip DRMed ebooks of their protection so you can use them on other devices
It's just a one-stop GUI for doing just about any sort of ebook maintenance/library management/cleanup. Helpful if your source materials are free or pirated, or if your device is unsupported at the major ebook stores. Lets you take a Kindle-only ebook, for example, get rid of the DRM, rip it to a more standard format, and send it to some obscure ereader that you have.
I think you misunderstood my question.
The question was in regards to this sentence from the op:
> Calibre actually does horrible things to the HTML source that hides in each epub.
An epub is just a zip file. Open it up, look inside, see for yourself.