"metadata" (toc) in HTML documents

The documents I am trying to make into epub format have a nice TOC in them. Can I tell calibre (or ebook-convert in particular) to just use that TOC?

Thank you,
Perry

[Starson17]

Originally Posted by pedz

The documents I am trying to make into epub format have a nice TOC in them. Can I tell calibre (or ebook-convert in particular) to just use that TOC?

Thank you,
Perry

My wife makes documents for her organization in Microsoft Word. She marks the chapters with Heading1 format, tells Word to make a TOC, saves it out as HTML filtered and drags it into Calibre. Calibre makes a beautiful EPUB TOC that's identical. Your question would be easier to answer if you said something about your source document.

I'm trying mostly to get the w3.org and whatwg.org documents. Here is one: CSS 2.1

Anyone care to help here?

[chaley]

I just did it by:

1. downloading the zip version of the document and unzipping it to a temp folder
2. adding cover.html in that folder to calibre
3. starting a convert to epub of the ebook resulting from #2
4. setting table of contents level 1 detection to //h:h1
5. letting the convert run

I got a nice epub with a level 1 toc. If you also want level two, put //h:h2 into the table of contents level 2 detection box.

That wasn't my original question. My question is how do I specify a TOC to use -- not how calibre can find the toc entries. Someone asked me for an example (which didn't make sense to me at the time).

So... to repeat, how do I tell calibre what to use for the TOC? i.e. how do I give it a list of TOC entries?

[kovidgoyal]

Write an opf+ncx file use the automatically generated ones as a base

or stick special class attributes on all the elements you want in the toc in the source html files and use the toc autodetection options

[chaley]

Originally Posted by kovidgoyal

or stick special class attributes on all the elements you want in the toc in the source html files and use the toc autodetection options

The html file already uses H? tags. The TOC (which is not actually a TOC, but is a section in the cover page) appears to be automatically generated by finding H? tags and generating a reference to the <a > anchor within the H tag. That is why I thought that doing the same thing (finding the H? tags) would satisfy the need.

[kovidgoyal]

Originally Posted by chaley

The html file already uses H? tags. The TOC (which not actually a TOC, but is a section in the cover page) appears to be automatically generated by finding H? tags and generating a reference the the <a > anchor within the H tag. That is why I thought that doing the same thing (finding the H? tags) would satisfy the need.

It should. calibre can also create a TOC from <a href> elements, so with a little fone-tuning, the OP could even use the TOC section in the HTML