While confirming that the toc.ncx gets recreated, I also noticed the same lang= issue for both it (toc.ncx) and nav.xhtml. Sort of. Toc.ncx doesn't even have an <html...> line (I guess since it's not an html/xhtml file). Instead, it has:
Code:
<ncx xmlns="http://www.daisy.org/z3986/2005/ncx/" version="2005-1" xml:lang="en">
Nav.xhtml has:
Code:
<html xmlns="http://www.w3.org/1999/xhtml" xmlns:epub="http://www.idpf.org/2007/ops" lang="en" xml:lang="en">
The reason I'm bringing these things up is that since I've been specifying the lang= line in the regular book files, I've had issues with the spellchecker. With all the text files using "lang="en-us" xml:lang="en-us"," the spellchecker comes up with the Language tagged as English (US). Except it's also picking up some of the same words from toc.xhtml, toc.ncx and jacket.xhtmo with their language tagged as either nothing or plain-ol' English. If I add one of those words to the dictionary, I can't do anything with the duplicate (quasi different language) word. To clear it after adding a "duplicate," I have to close the spellchecker, shut down the Editor and re-open everything. Only then will that word disappear (Refreshing the list doesn't help). This gets in the way of spellchecking the book.
EDIT: Also, manually editing the <html...> line in nav.xhtml does seem to remove its "duplicate" entries in the spellchecker (I assume they're getting folded into the ones for the rest of the document). But, changing the language in the <nxc...> line of toc.ncx doesn't make any difference. Its misspellings continue to show up as plain English.