how to build toc when chapter headers are just images

[cybmole]

I have a book in which images have been used as chapter headers - each chapter has a unique image.

the chapter header code looks like this

<h1 class="calibre8" id="calibre_pb_13"><span class="calibre9"><a class="calibre10"></a><img alt="" class="calibre11" src="../Images/record-522.jpg" /></span></h1>

the image name changes for each chapter, and the actual chapter number has been embedded into each image.
the id ="calibre_pb-nn" changes its value but nn increments 2 at a time & starts at 09 so would be tricky to extract & adjust that

I cannot think of any way to auto-generate a workable TOC from this - is there one ?
I could manually add text to each chapter header but that would take a long time.

[Rob Lister]

What do you want the end result to look like?

[DiapDealer]

If something in the id or the image name could be matched up to to the chapter number, you could probably get a regexp to at least add that info to the title attribute. But even then, nothing's going to be able to extract the chapter "name" from the different images. As it is (with the image name starting at 9 and incrementing by twos), I don't think you're going to be able to do either... automatically. Your best bet is to add the ToC entry you prefer to the title attribute of each chapter-header manually.

[Perkin]

Isn't there a title attribute you can insert in the h# tags that get recognised?

[Jabby]

Originally Posted by cybmole

I have a book in which images have been used as chapter headers - each chapter has a unique image.

the chapter header code looks like this

<h1 class="calibre8" id="calibre_pb_13"><span class="calibre9"><a class="calibre10"></a><img alt="" class="calibre11" src="../Images/record-522.jpg" /></span></h1>

the image name changes for each chapter, and the actual chapter number has been embedded into each image.
the id ="calibre_pb-nn" changes its value but nn increments 2 at a time & starts at 09 so would be tricky to extract & adjust that

I cannot think of any way to auto-generate a workable TOC from this - is there one ?
I could manually add text to each chapter header but that would take a long time.

There are probable more elegant ways to do this, but I would try this:
Find <h1
Replace with <h1 title="Chapter 1"
For me, it beats copying the above to the clipboard and then pasting it one at a time.

You still have to fix the chapter count one-by-one..

Regards - John

[cybmole]

Originally Posted by Rob Lister

What do you want the end result to look like?

I just want to be able to select a chapter from the toc menu when I put the book onto my Sony reader - but judging from previous comments I either have to do a long winded manual fix or accept that is not going to happen!

if I ask sigil to create a toc there are no visible entries- I guess it finds the h1 tags OK but because there is no text it then gives up

John's solution is the best so far.

if I weren't so OCD about tweaking books to meet my idealised requirement then 'twould not be an issue as I expect I'll only read it once; but I use challenges like this to improve my regex / book layout skills & I get frustrated when I cannot find an elegant solution

[DiapDealer]

Yeah, you're just going to have to bite the bullet and manually add a title="Chapter XX - The Beginning" attribute to each of your chapter <h> tags. Then... you'll be able to generate a ToC by clicking the button.

[theducks]

Seeing the Images have unrecognisable generic names
Make them meaningful (eg. ..\Images\chapter_1), hopefully using the great Bulk renamer Meme built.
(Sigil adjusts the references )

Now that is done,
REGEX time

Code fragments

your search term: will include:
<h1 class="calibre8" id="calibre_pb_(\d+)">
Images\\chapter (\d+) <- note the SECOND forward reference

You are going to replace

<h1 class="calibre8" title="Chapter \2" > The Bold is what will be in the TOC

and repair the captured text

Images\chapter \2

I just threw away the id tag as Sigil's TOC generator will make a new one if missing

[DiapDealer]

Nice catch on the mass image renaming!

It won't help getting the chapter name/title from the image, but it will get things a lot closer a lot quicker.

[Jabby]

Originally Posted by theducks

Seeing the Images have unrecognisable generic names
Make them meaningful (eg. ..\Images\chapter_1), hopefully using the great Bulk renamer Meme built.
(Sigil adjusts the references )

Now that is done,
REGEX time

Code fragments

your search term: will include:
<h1 class="calibre8" id="calibre_pb_(\d+)">
Images\\chapter (\d+) <- note the SECOND forward reference

You are going to replace

<h1 class="calibre8" title="Chapter \2" > The Bold is what will be in the TOC

and repair the captured text

Images\chapter \2

I just threw away the id tag as Sigil's TOC generator will make a new one if missing

If you have books that come from Calibre and your book has other images scattered around, you could have a problem.

Calibre assigns a starting image number and increments as it goes. So,you could have a chapter 5 (00005.jpg) image followed by a photo image (00006.jpg) followed by a chapter 6 (00007.jpg) image. This sequence would put a photo to depict chapter 6 and all after would be off.

This works just fine if the chapters are in strict sequential order. You needn't ask how I know.

Regards - John

[theducks]

Originally Posted by Jabby

If you have books that come from Calibre and your book has other images scattered around, you could have a problem.

Calibre assigns a starting image number and increments as it goes. So,you could have a chapter 5 (00005.jpg) image followed by a photo image (00006.jpg) followed by a chapter 6 (00007.jpg) image. This sequence would put a photo to depict chapter 6 and all after would be off.

This works just fine if the chapters are in strict sequential order. You needn't ask how I know.

Regards - John

If you use Calibre. Open the EPUB with TweaK, Locate the images (easy if the have been touched by Sigil, they will be in the Images folder), Select All (images): Print (the photo wizard should start): Print a Proof sheet (thumbnails)
Now you know which to rename
Yo also get to see which images are really repeats (usually scene breakes and chapter logo's that are not numbered) and pick just one to use, making the book smaller, less bloat
Sigil 5 allow multiple placement of the SAME image

[theducks]

deleted duplicate answer

[cybmole]

thanks but I went instead with a variation of John's solution, I regexed in some place holder text into each h1 tag, so that sigil could build a toc. I then edited the toc page , adding chapter numbers 1, 2, 3 etc into title field & then regexed the placeholder text back out again.

it will be find now so long as a don't touch the recreate to button!

messing with the images would have taken much longer as ever image contained a chapter number & they look good as-is.

[Dopedangel]

Originally Posted by cybmole

thanks but I went instead with a variation of John's solution, I regexed in some place holder text into each h1 tag, so that sigil could build a toc. I then edited the toc page , adding chapter numbers 1, 2, 3 etc into title field & then regexed the placeholder text back out again.

it will be find now so long as a don't touch the recreate to button!

messing with the images would have taken much longer as ever image contained a chapter number & they look good as-is.

If the chapters are split properly using epub fixer you can just select all the chapters add and then rename them automatically I know I am a bit late in giving an answer but for future use.

[cybmole]

Originally Posted by Dopedangel

If the chapters are split properly using epub fixer you can just select all the chapters add and then rename them automatically I know I am a bit late in giving an answer but for future use.

please explain how to do that, given that chapters are split ( = 1 x xhtlml per chapter)
but that chapter headers are image only