Single page vs. multiple pages (epub creation)

[electrotype]

I'm trying to create an epub. The original source of the text is a PDF, but I downloaded an "epub" version generated by archive.org.

Unfortunately, the generated epub is far from perfect! The content is not split where it should be, etc. So I'm trying to clean everything up and create the final epub myself.

I have already read about epub format and I have successfully created one. But I still have one question.

In the process of collecting all the original text (I am a developer so I created some code for this), I ended up with a single large piece of text. I could split the text into multiple pages, but I'm wondering if this is really useful?

Can't I just have one big "content.html" file, with "id" attributes on some "<p>" elements so that the "toc.ncx" can point to them? Are there advantages to actually splitting the content into several separate files instead of a big one?

[pdurrant]

The advantage of having multiple files is that more ereaders will handle the ePub successfully. ereaders tend to be more limited in memory than most computers.

[electrotype]

I see! Thanks for the information.

[pdurrant]

There are disadvantages too. Most ereaders enforce a page break between files.

[electrotype]

Yes, I guessed that. You have to split the files at the right positions. Thanks again.

[Quoth]

My experience is that the Internet Archive "ebooks" are worthless. They are generated automatically from un-proofed OCR text only marginally good for searching.

So I deleted them all and only download PDFs (after checking they are really PD) and read them on a tablet.

You are better doing your own OCR of the PDF and proofing it. Do put page breaks at chapters, sections or other natural breaks in your wordprocessor. Later those will start new files in the epub. A new file is the only reliable page break and works for epub converted to mobi, azw3/KF8, dual mobi and KFX.

[electrotype]

Originally Posted by Quoth

You are better doing your own OCR of the PDF and proofing it.

Good to know. The text of the one I downloaded isn't too bad, it's just not formatted properly. I do find some OCR errors, but not too many.

Anyway, my opinion is that if you really want to do a good job, you need to read the final epub from A to Z once it's created, fix mistakes, and then create a final version. Some books really deserve that work.

I plan to create epubs for other books too, so I might try what you suggest! I'll test some OCR applications to see how they work. Thanks.

[slm]

Originally Posted by Quoth

My experience is that the Internet Archive "ebooks" are worthless. They are generated automatically from un-proofed OCR text only marginally good for searching.

So I deleted them all and only download PDFs (after checking they are really PD) and read them on a tablet.

You are better doing your own OCR of the PDF and proofing it. Do put page breaks at chapters, sections or other natural breaks in your wordprocessor. Later those will start new files in the epub. A new file is the only reliable page break and works for epub converted to mobi, azw3/KF8, dual mobi and KFX.

Just for the record--and to offset the quoted view--my experience with reading many Internet Archive "epubs" OCR'd from pdfs over many years is that about two-thirds of them are perfectly OK and about 10% are completely unusable. Note that this is just for casual reading.

[Quoth]

About 100% of ones I downloaded were poor and more than 1/2 unusable.

Perhaps it depends on the source and content and when the scanned book was printed. Also check they aren't "pirated".

Many other people have remarked that you are better off downloading the PDF.

[Karellen]

Originally Posted by electrotype

I'll test some OCR applications to see how they work. Thanks.

Might be of use...
https://www.mobileread.com/forums/sh...93#post4341993

[electrotype]

Originally Posted by Karellen

Might be of use...
https://www.mobileread.com/forums/sh...93#post4341993

Thanks, bookmarked.