When calling epub.EmbedImages() the added images are automatically named from the library. The problem is that the images are not cleaned when they come from URLs.

For example, a image of name: image.jpg?width=300 will try to create the same filename in the system, and some do not support symbols like that.

An example of this can be seen at: https://github.com/go-shiori/shiori/actions/runs/7819243851/job/21331283455?pr=837#step:5:369

The offending line: https://github.com/go-shiori/go-epub/blob/main/epub.go#L629-L630

Another possibility is to always create random filenames when calling EmbedImages(), since it's automatic nobody should care.

A subsection should belong to its section in toc.ncx, for proper show in reader.

Currently subsection is the same as section in toc.ncx

currently AddCSS just add file and that to manifest but not add that on document too, maybe it is not bad if we have option for that.
this make files not pass epubcheck

https://github.com/blang/vfs

Currently the library can only use system files, but I would like to be able to use vfs too

What is the problem?

if user try to add different or one file with same filename go-epub not add the second one.

	e.AddImage("path/to/file1.jpg", "filename.jpg")
	e.AddImage(t"path/to/file2.jpg", "filename.jpg")

What do you expected?

add second file with a new name and show a warning to the user

as an example this code use custom name SubSectionName it generate SubSectionName file without extension(xhtml) that can create error in some of ebook reader.
to avoid that go-epub should add extension automatically or return error.

I see how in the future we could allow creating a new Epub instance providing a custom logger from log/slog, but we can plan that later on.

Originally posted by @fmartingr in #5 (review)

Processing of chapters into subchapters does not work correctly in the AddSubSection function, I get an error like ParentDoesNotExistError, I cannot correctly process more complex nesting.

It would be nice if it was possible for user of the API to provide translations for English strings that are currently hardcoded. There is at least Table of contents. Maybe more.