|
Sorry, I don't understand. That's what I think I did.
There is TOC in the text of the document - { TOC \o "1-2" \h \z \u }.
It's displayed correctly in the text of DOCX and in the text of calibre-converted EPUB.
But generated NCX file - ie. the TOC displayed on the side of the Calibre Viewer (or in ebook reader) is flattened. That means that Calibre has correctly found all TOC items, but works with them as with 'level 1' items (incorrectly, as they use level 1 and level 2).
|