MobileRead Forums - View Single Post - Sigil to Kindlegen -- No Table of Contents

Elonex500eb-User · 04-11-2012, 03:58 PM

I have had a lot of 'battles' getting the TOC for Kindle to work.

The first thing I do is convert the source to epub if it's not already in epub format.

I create my own css for the chapter heading - thus I can be 100% sure its unique.

I set all the headings to 'h1' and apply my css

I use the Sigil TOC function to generate the content list

I check the TOC links in Sigil sometimes they dont jump to the chapter text - in my experience this has always been because the 'id' value is missing in the toc.ncx so edit as needed.

If everything is ok in Sigil I then check to make sure that Calibre reads this edited epub correctly.

Now comes the Calibre conversion from epub to mobi.

Earlier I said I create my own css for the chapter headings I call this class 'chapter' - Calibre's Structure Detection function is already set up to look for this class in H tags !!!

So there's no need for xpath expressions in Calibres TOC function.

On Calibres Mobi output function I leave the 'Do not add TOC' option UNTICKED

So I run the conversion I usually open the 'Job' function so I can see how Calibre is working I want to see.... 'Chapters detected'.

When its finished I check my mobi conversion and it always works !

My process may not be 'good' Sigil or 'good' Calibre, but it works for me.

I'm not publishing or distributing books, and I'd rather read books than spend time processing them !

I whole heartedly agree with Whackatagin's comment about "simple is best".

04-11-2012, 03:58 PM	#63
Elonex500eb-User Now a Kindle3 User ..lol Posts: 5 Karma: 10 Join Date: Apr 2011 Device: Kindle3+wiFi no 3g	I have had a lot of 'battles' getting the TOC for Kindle to work. The first thing I do is convert the source to epub if it's not already in epub format. I create my own css for the chapter heading - thus I can be 100% sure its unique. I set all the headings to 'h1' and apply my css I use the Sigil TOC function to generate the content list I check the TOC links in Sigil sometimes they dont jump to the chapter text - in my experience this has always been because the 'id' value is missing in the toc.ncx so edit as needed. If everything is ok in Sigil I then check to make sure that Calibre reads this edited epub correctly. Now comes the Calibre conversion from epub to mobi. Earlier I said I create my own css for the chapter headings I call this class 'chapter' - Calibre's Structure Detection function is already set up to look for this class in H tags !!! So there's no need for xpath expressions in Calibres TOC function. On Calibres Mobi output function I leave the 'Do not add TOC' option UNTICKED So I run the conversion I usually open the 'Job' function so I can see how Calibre is working I want to see.... 'Chapters detected'. When its finished I check my mobi conversion and it always works ! My process may not be 'good' Sigil or 'good' Calibre, but it works for me. I'm not publishing or distributing books, and I'd rather read books than spend time processing them ! I whole heartedly agree with Whackatagin's comment about "simple is best".