Ok, i'm trying to convert http://www.ussc.gov/guidelines/2010_guidelines/Manual_HTML/Chapter_2_A-C.htm from HTML into a TOC'd Epub using Calibre. I am able to get it to identify <h1> or <h2> tags for entries. However I'm trying to get it to recognize all of the individual sections that are marked like this "<p><b><span id="ident">§2A1.1.</span> <u>First Degree Murder</u></b></p>" as TOC entries. Ok great it did. It created a list like the following:
§2A.1...............7
§2A.2...............7
§2A.3...............7

Everything is linked to the same page. That's not great. I don't know why but it created all the correct entries but linked them to the same place. I've searched and can't seem to find an answer.

caliber should not be creating ToC with ..... <page number>. Is it possible that there is already a ToC in the file and that caliber only convert it?

If you want sub-section, you probably want to edit your source html to useh3, h4 and so forth. calibre can only convert to three levels of toc so you might have to use Sigil to edit the ToC in the resulting ePub instead of having calibre convert everything.

That's my experience with it anyways, your millage may vary.

Best