Help on Epub correct numbering and index

[AlexHoang]

Hello guys,

I have a bunch of epub reading files. I read on my desktop. I have tried a bunch of epub reading apps. But eventually I am settling down with SumatraPDF because its lightness.

There are a lot of good books I like so much that I would like to print these out as physical books so I can have the feeling of flipping pages.

I know and already installed the software Calibre which helps convert Epub to PDF for printing as it looks like the only solution.

But the thing is that the Index part in my original epub book is already doesn't fit with the correct page the term I want to look up.

I think the index part is legitimate as correctly as the physical book but on epub it doesn't fit at all. For example, "nutrition 45," I can click on 45 and it takes me to the page it has the first word "nutrition" but that page is not 45th one.

Do you think whether I figure out to have the size of pages fit well regarding to index?

In the attachments, the first one is one of my pdf books, which is perfectly for printing out as it looks as the legit physical book.
The 2nd is one of my epub books (I think all my epub books have this kind of problem: indexing is correctly as the physical books but doesn't fit with page numbering) and the last one is a pdf book I converted from epub by Calibre.

How can I adjust or converting my epub book into pdf which I can read as its legit physical book like my first attachment? Is it possible? Or I have to live with having a print-out with wrong page numbering regarding to the index like my 3rd attachment? it is not appealing enough for me to read after being printed out anymore.

[DiapDealer]

Moderator Notice
Please do not upload and attach copyrighted material to Mobileread. It is a violation of both copyright law and Mobileread rules. This will be the only warning you receive concerning copyright violation.

[Hitch]

Originally Posted by AlexHoang

Hello guys,

I have a bunch of epub reading files. I read on my desktop. I have tried a bunch of epub reading apps. But eventually I am settling down with SumatraPDF because its lightness.

There are a lot of good books I like so much that I would like to print these out as physical books so I can have the feeling of flipping pages.

I know and already installed the software Calibre which helps convert Epub to PDF for printing as it looks like the only solution.

But the thing is that the Index part in my original epub book is already doesn't fit with the correct page the term I want to look up.

I think the index part is legitimate as correctly as the physical book but on epub it doesn't fit at all. For example, "nutrition 45," I can click on 45 and it takes me to the page it has the first word "nutrition" but that page is not 45th one.

Do you think whether I figure out to have the size of pages fit well regarding to index?

In the attachments, the first one is one of my pdf books, which is perfectly for printing out as it looks as the legit physical book.
The 2nd is one of my epub books (I think all my epub books have this kind of problem: indexing is correctly as the physical books but doesn't fit with page numbering) and the last one is a pdf book I converted from epub by Calibre.

How can I adjust or converting my epub book into pdf which I can read as its legit physical book like my first attachment? Is it possible? Or I have to live with having a print-out with wrong page numbering regarding to the index like my 3rd attachment? it is not appealing enough for me to read after being printed out anymore.

Realistically, you can't. How can you possibly know if it was a 6" x 9" book, or 4.5" x 7", or, or or? What font and font size? How many words per-page? How would you test it?

Even if you could manage to locate, say, 20 different index entries that all go on the same "page," How many iterations of a print would you have to do, to try to get all those indexed items to print, on this new page 20? Can you even be 100% sure that the indexer means the incidence of a given word, that YOU think she means? You can't.

If you want an index to work, you have to buy the print book. Using a reproduced print index, that's been copied into an ebook, when you're THEN printing it out to PDF--that's just bloody hopeless. It's too many iterations of changes with which you're not familiar.

Hitch

[Tex2002ans]

Originally Posted by AlexHoang

I know and already installed the software Calibre which helps convert Epub to PDF for printing as it looks like the only solution.

PrincePDF is also a solution for generating EPUB->PDF. It's one of the most advanced tools for generating high-quality PDFs from HTML+CSS+Javascript.

Jellby released a Calibre Plugin called "Prince PDF". (More information can be found in the topic.)

Originally Posted by AlexHoang

But the thing is that the Index part in my original epub book is already doesn't fit with the correct page the term I want to look up.

I think the index part is legitimate as correctly as the physical book but on epub it doesn't fit at all. For example, "nutrition 45," I can click on 45 and it takes me to the page it has the first word "nutrition" but that page is not 45th one.

There isn't a way to dynamically renumber EPUB "pages"/links in an Index to newly-generated PDF pages.

Depending on your EPUB's code, a compromise may be to visibly display the original page numbers, and float them into the margin of the PDF, similar to this:



I'm not familiar at all with Calibre's PDF output, but you should be able to finagle something using HTML+CSS and negative margins:

HTML:

Spoiler:

<p>It does not make any difference in this regard whether the new <span class="pagenumber">508</span> invention is or is not protected by a patent. A firm that has acquired a license has already expended money for the new invention. [...]</p>

[...]

<p>If an invention is so superior to the old processes that it makes the old equipment obsolete and peremptorily demands its immediate replacement by new machines, the transformation will be effected no matter whether the privilege conferred by the patent is in the hands of the owners of the old equipment or of an independent firm. The assertions to the contrary are based on the assumption that not only the inventor and his attorneys but also all people already active in <span class="pagenumber">509</span> [...]</p>

CSS:

Spoiler:

span.pagenumber {
	float: right;
	font-weight: bold;
	padding: .5em;
	margin-left: -1em;
	margin-right: -2em;
}

Here is how the EPUB appears in Sigil:



Of course, your book's exact code + PDF settings will be different, so you'll have to experiment with what negative margins work well for your book.

Originally Posted by Hitch

If you want an index to work, you have to buy the print book. Using a reproduced print index, that's been copied into an ebook, when you're THEN printing it out to PDF--that's just bloody hopeless. It's too many iterations of changes with which you're not familiar.

I agree.

Too many variables change (page size, font size, margins), and the littlest tweak to even ONE variable (footnotes being indented 1mm to the left) can throw all subsequent PDF pages off.

There's a reason why layouts of physical books are completely locked down when books are passed to Indexers. Trying to reverse engineer all that work is... extremely tedious.

[AlexHoang]

Thanks guys, it seems like I ought to live with it.

@Tex: I have tried Prince as you suggested. It is bloody difficult. The pdf file coma out to me as error. It seems to take some work to get used to it.

@Hitch, Yeah, it sounds hopeless to me now!

[Hitch]

Originally Posted by Tex2002ans

<snippage>



I agree.

Too many variables change (page size, font size, margins), and the littlest tweak to even ONE variable (footnotes being indented 1mm to the left) can throw all subsequent PDF pages off.

There's a reason why layouts of physical books are completely locked down when books are passed to Indexers. Trying to reverse engineer all that work is... extremely tedious.

First, Tex, I wanted to say that you have some incredibly clever solutions in there, truly.

Secondly, the lack of understanding, of most people, of the reality of indices--it's boggling. (@AlexHoang, this is absolutely NOT directed at you; we've had some run-ins about indices, at my shop lately, and I'm taking this moment to spew a bit.)

We've had clients that want us to magically take the index from a previously-printed book, make a new printed book (with new paging and page numbers, right?) and somehow, "convert" the old index to work in the new. Uh...nope? I mean, can you imagine the time and effort in that? You'd have to read the entire bloody book, and then work through the index, item by item, and try to figure out what it was, on original page 36, that the indexer thought was worth noting, and then find it in the new, re-tag it, lather-rinse-repeat for a thousand index entries, and then generate a new index. Uh, no. Not for what WE make, that's for damn sure.

We've had to do things like, extract text (scan/ocr/proof) to a new Word file, send it to the client, TEACH THEM HOW TO USE THE TAGGING TOOL, to mark things for an index, and then generate a new index, THEN import the cursed thing into INDD, tag the entries...I mean, in the Cheap Seats, so to speak, making new indices is a freaking nightmare. (On a bright note, I've had two clients lately buy indexing services from an associate, so at least those are done well.)

Oy. If you're a non-fiction author, before you publish, learn how indices are made, not only through some "Then A Miracle Occurs" mindset, but the actual mechanics. Unless you get picked up by Random House, pal, you're going to need to handle it yourself, so, learn now. Save your designer some hassle.

Hitch

[Tex2002ans]

Originally Posted by Hitch

Secondly, the lack of understanding, of most people, of the reality of indices--it's boggling. (@AlexHoang, this is absolutely NOT directed at you; we've had some run-ins about indices, at my shop lately, and I'm taking this moment to spew a bit.)

lol. Yep AlexHoang, you just landed in an age-old discussion. Me + Hitch have discussed Indexes a ton on MobileRead. :P

(It also gets brought up every so often when people also bring up "Page Numbers" in ebooks.)

Originally Posted by Hitch

We've had clients that want us to magically take the index from a previously-printed book, make a new printed book (with new paging and page numbers, right?) and somehow, "convert" the old index to work in the new. Uh...nope? I mean, can you imagine the time and effort in that? You'd have to read the entire bloody book, and then work through the index, item by item, and try to figure out what it was, on original page 36, that the indexer thought was worth noting, and then find it in the new, re-tag it, lather-rinse-repeat for a thousand index entries, and then generate a new index.

Yep, I know that game. The screenshot I shared above was from a book I've been retypesetting over years. It took me months to recreate the Index, and I'm still not done with it. I wrote about that experience in previous threads. (Indexes are incredibly information-dense, and thousands of cross-references can be condensed into a very small amount of space.)

And like Hitch mentioned, you can't just "renumber the Index", you have to actually reread the entire page and figure out WHAT exactly the Indexer was referencing:

Take an example "Poodles, 15":

• Was it just a single word "poodle" in Page 15?
• Or were the top-to-bottom paragraphs all discussing "poodles"?
  
  • Imagine in your new PDF: Old Page 15 top paragraph = New Page 20 bottom of the page.
    
  • All of a sudden, a single page reference in the Index: 15 can become "20-22"
    

Now what if there was a "Dogs, 15"?

• Did the Indexer mean the same exact reference as the single "poodle" sentence?
• Or did the rest of the page have other discussions about dogs?

... multiply that work by 100, 200, 500, or however many individual PAGE NUMBERS appear in your Index... and you can see how the amount of time to recreate an Index properly explodes exponentially.

And things can only get more complicated from there (introduce footnotes, book titles, full author names, same things in different categories, [...]).

Originally Posted by Hitch

If you're a non-fiction author, before you publish, learn how indices are made, not only through some "Then A Miracle Occurs" mindset, but the actual mechanics. Unless you get picked up by Random House, pal, you're going to need to handle it yourself, so, learn now. Save your designer some hassle.

Indexing is also a highly specialized skill. And in order to create a fantastic Index, you also have to be extremely well-versed in the subject matter + "know HOW a typical reader will approach this book", because you have to categorize the text as well.

Think to something like a cookbook, a single recipe can belong under multiple categories:

Chicken Noodle Soup:

• Soups > Chicken Noodle
• Main Ingredients > Chicken > Chicken Noodle Soup
• Crockpot Recipes > Chicken Noodle Soup
• Method > Stovetop > Chicken Noodle Soup

Different readers may be approaching the recipes in different ways:

• Some may LOVE chicken and want every chicken recipe
• Others may have a crockpot and set-it-and-forget-it
• Others may LOVE soup

[Hitch]

Tex is my indexing hero. I want to be a bigtime indexing coder like him when I grow up. :-)

Seriously, he's one of the few people here who's tried to tackle this, manually, so he knows whereof he speaks.

Hitch

[Notjohn]

Originally Posted by Hitch

Unless you get picked up by Random House, pal, you're going to need to handle it yourself, so, learn now. Save your designer some hassle.

I've been published by Doubleday, HarperCollins, and Smithsonian Books, and in every case I had to do the index myself.

How much does your associate charge for indexing, do you know?

[Hitch]

Originally Posted by Notjohn

I've been published by Doubleday, HarperCollins, and Smithsonian Books, and in every case I had to do the index myself.

How much does your associate charge for indexing, do you know?

There are two levels. Actually, NJ, can you email me? You know the address; it's probably more appropriate to discuss it in private. I don't want the UberMods to think I'm pimping.

Hitch

[Notjohn]

Wilco. Thanks.

[Hitch]

Originally Posted by Notjohn

Wilco. Thanks.

Thanks.

Hitch

[democrite]

For reading EPUB 3s as they are and displaying page numbers, I’ve been meaning to file a request for the Calibre reader. As for rendering PDFs with correct numbering, unsure if it’s possible to CSS the page numbers so they cause a page break. If so, and perhaps it might require some post-cleanup, there are Acrobat plugins that can link numbers to page. All I’ve tried link to physical page so one may need to temporarily remove any front matter differently numbered.

[Hitch]

Originally Posted by democrite

For reading EPUB 3s as they are and displaying page numbers, I’ve been meaning to file a request for the Calibre reader. As for rendering PDFs with correct numbering, unsure if it’s possible to CSS the page numbers so they cause a page break. If so, and perhaps it might require some post-cleanup, there are Acrobat plugins that can link numbers to page. All I’ve tried link to physical page so one may need to temporarily remove any front matter differently numbered.

Am I the only one here feeling as though I've stepped into a Tardis that's going backwards????? The Cave Tardis?

Hitch

[DNSB]

Originally Posted by Hitch

Am I the only one here feeling as though I've stepped into a Tardis that's going backwards????? The Cave Tardis?

Hitch

Nope. Sheesh, next we will be arguing about which edition of the pbook we need to duplicate the page numbers for.