How to export a poorly flowed ID file to ePub?

[dkudler]

I'm converting four long (200K word+), moderately complex books from ID 2021 to ebooks. The problem is that the designer did a lovely job designing for print, but the way the file is put together makes it a complete mess when I try to export to reflowable ePub:

• Each book is a single INDD file — not chapters bound into an INDB file
• There's no XML or article structure applied
• The body text is a single flow, from beginning to end
• However, the part and chapter heads are placed in free-floating text blocks
• Some of the images (50–100 per volume) are anchored, but many are not
• For those images that are anchored, the designer added blank paragraphs to create space rather than flowing the text (unfortunately, the text features breaks that the designer coded as simple double returns, so I can't just find-and-replace all of those)
• Some of the captions flow with the text, while others are, like the chapter heads, placed in free-floating text blocks
• Frequently, text is hand-formatted instead of having styles applied (ie, hard hairspaces added rather than using tracking, there are forced line breaks, or the paragraph style is just "Normal" with the typeface/font/tracking set by hand, rather than using an existing paragraph style)
• Also, there are a whole bunch of overly specific character styles
• All lists are manually formatted rather than having numbers/bullets applied
• All of the endnotes are unlinked — just superscripted numbers.

Again, the print editions are beautiful. But exporting to ePub? Not so pretty.

Most of this I can deal with easily enough. I've got an existing stylesheet that I've developed for this client's ebooks, so I'm going in and making sure that all of the style names and export tagging link up properly. Before export, I can apply lists styles and get rid of hairspaces and forced breaks. I can go through after export and use REGEX to create hyperlinks between the endnotes and the  references, etc.

What's been the most frustrating are the free-floating image, caption, and header blocks, because when I export, they don't appear "based on the page layout" — they end up in completely different locations in unpredictable order. This means going through and placing them manually.

Now, I guess I don't mind doing that — it's just time, right? But there are four of these monsters, and I don't want to waste my time and my client's money.

I tried breaking the first book into chapters to make the search for missing art, captions, and heads a bit easier, but of course, the text simply reflowed — each "chapter" contained the whole book's body text.

I consider myself a moderately experienced InDesign user and a quite experienced ePub designer — but I can't think of a simple way to make sure that all of the elements/objects for each book actually export to where they're supposed to. 

I have looked, but I can't seem to find any kind of a solution here or elsewhere. I spent some time looking at scripts, but couldn't find anythinig that seemed as if it would make those floating objects stay put.

Help? I'm sure I'm missing a simple solution, but can't think what it is.

[James Gifford—NitroPress · Accepted solution]

No, there is no simple solution. The fundamental need is for the text to be in one, continuous flow, with all elements like images anchored to the appropriate point in the text. Until that's accomplished, there's no good way foward to EPUB.

INDD files bound together as an INDB book will work fine once each component file is fixed.

ETA: Keep in mind that you can rebuild then quite roughly in ID, with only modest attention to formatting, page layouts, etc. You don't need to replicate the original layout nor try to make a perfect EPUB layout in the source.

[dkudler]

Ah, well. I was afraid you were going to say that.

Yeah, I'm butchering the lovely layout I was given. I'll just copy and paste the chapter heads into the main text. 

I guess I'll drop the captions into the main text as well — it's what I have to do, basically, in the ePub file — and then anchor the images to the captions.

Whew. That's more work than I'd like it to be, but if that's how it has to be, that's how it has to be.

[James Gifford—NitroPress]

It might be easiest to create a new doc with any convenient page format (simple) and then cut and paste the old material into it, one block at a time. Develop simple styles with matching names so you end up with a vaguely e-booky look in the new one. That will make it easy to proof and edit. (And clean up the styles, hierarchy and name, as you go.)

Many details, but you may want to follow the practice of anchoring each illustration to a text paragraph and then placing the caption as its own style paragraph below. Otherwise, the captions have to go over the illos. Use a named and defined graphics frame for the illustrations to allow precise control in the conversion. And if you haven't gotten into CSS styling... this might be the project on which to learn.

And FWIW, I just did much this same project... but the designer who hacked together the lovely but badly structured print version was, uh, me. 🙂 There were reasons.

[dkudler]

LOL

Hey, I have no judgment about the designer's work! It's just creating a huge headache for me at the moment. (I will also say, it's laid out very much as if by someone who learned to design books by literally cutting and pasting blocks of film. Which it was, of course.)

I have in fact been going through and getting rid of extraneous styles, merging them with ones that line up with the CSS stylesheet. And since most of the book was in fact a single text flow, I've been working in a duplicate version of the original INDD file, just copying and pasting the floating bits into the flow. I know that some of the image blocks will still end up in weird places — and the whole thing will need to be refined once I have a conversion I'm happy with — but at least I'm confident that this is how it has to be done.

Again, thank you!

[James Gifford—NitroPress]

Without seeing it, I can't judge, but putting everything in separate text frames is... not the most professional approach. I started on blueline boards but very quickly learned to appreciate the more sophisticated options of publication software. (Even if, throwback Thursday here, all the original PageMaker offered was a digitized pasteboard!)

You can do this in one or two passes; it might be more productive to go as you are, just get all the stuff into the new file in the right structure and then clean it up in a second pass. (I know I get bored and frazzled trying to do such combinations of repetition and precision at the same time.) Or, if you have the styles increasingly sorted out (and a bit more patience) it might be worth fixing each chunk as you put it into place—apply the right graphics frame style and optimize the framing, for example.

Anything that works... as long as it doesn't drive you crazy in the process. 🙂

[dkudler]

Yeah, well, as I said, it's not how I'd lay the books out.

And yes, the combination of needing to be very careful with so many moving pieces — not my favorite.

I've got the first volume cleaned up and exported — looks like everything is where it's supposed to be. A couple of tables that got converted to images... but I'm actually okay with that, since they were going to be diffcult to scale properly as HTML tables..

Now I get to go through and link up all of the endnotes. Luck me. 😉

(I remember the early days of PM! Actually, my first experience with page layout software was ReadySetGo — which I was actually pretty fond of at the time.)