Internal Filenames in MOBI/AZW to EPUB Conversion

[Jaws]

This is a really minor thing, but it crops up every time.

After converting a MOBI-family (including AZW) book to EPUB, and then starting to "Edit book," the top file is named "titlepage."

It's almost never the title page. It's the cover (usually).

I'd be much happier if Calibre didn't actually misname things using assumptions. It's better to not "name" anything (and begin with *_000.htm) than to misname things that, say, might be referenced by external indexing programs looking for the actual title page.

Failing that, is there a way to adjust/tweak any parameter so that "titlepage.htm" is instead imported as "cover.htm"? I've tried to find such info, but failed.

[kovidgoyal]

What external indexing tool relies on filenames instead of metadata to do indexing. Ifthere exists such a tool, it should be fixed. Internal filenames in EPUB have no semantic meaning. The guide in the opf in epub2 or the landmarks in epub 3 need tobe used for that.

[Jaws]

There are a lot of public-domain and other older books out there with utterly useless and inaccurate metadata. There are several useful scripts that look for the title page and copyright page and use those as sources to correct bad metadata.

Put another way: Metadata is a GIGO issue that can and should be cross-checked against the actual content of the book (and publishers are awful about it, especially when there's a "new foreword by cringeworthy celebrity" that the publisher has put into the metadata as a coauthor). The cover doesn't include that data in a useful form, so it shouldn't be named in a way indicating that it does have that data in a useful form.

[kovidgoyal]

It does include it in useful form, useful to *humans*. Just not to your scripts.