Sigil, the ePub editor

[Agama]

Originally Posted by ksng

What's the best and efficient way to convert a book to epub file that it's readable? I'm using Indesign to publish epub file and then use text editor to edit some values in the epub file.

It depends a bit on what your source format is, but if you use plain text files then you can use calibre with markdown processing to obtain excellent results. Calibre also does a great job with html input.

[Oldpilot]

I have a book in html newly uploaded to Amazon DTP. I came to the Sigil forum just now because I find it more intuitive than Calibre, and I needed some guidance. Is Calibre really the go-to software to translate my html file to epub for publication on Barnes & Noble?

[DaleDe]

Originally Posted by Oldpilot

I have a book in html newly uploaded to Amazon DTP. I came to the Sigil forum just now because I find it more intuitive than Calibre, and I needed some guidance. Is Calibre really the go-to software to translate my html file to epub for publication on Barnes & Noble?

Actually HTML is a special case where Sigil can be used directly. Just zip up the HTML file(s) and any images and then rename the extension to .epub. At this point Sigil can open the file directly.

Dale

[ksng]

Once I use Sigil, my ePUB will not pass the ePUB check so make sure you test it.

[Valloric]

Originally Posted by ksng

Once I use Sigil, my ePUB will not pass the ePUB check so make sure you test it.

A fairly broad statement. Are you sure that the source you're bringing into Sigil is error-free and that Sigil is the one introducing problems?

Here's something I've noted during the last two years of Sigil development: when someone complains that Sigil is the one introducing validation errors into their source files, it is practically always either:

1. a result of user error;
2. a result of user ignorance;
3. just plain false;
4. all of the above.

Then again, if you can produce a test case that consistently makes Sigil introduce validation errors, feel free to create an issue about it on the tracker. And I say this without even the tiniest shred of sarcasm: I'd truly love to see it.

But do bear in mind that the number of such reported issues that are not a result of one of the four things listed above is very, very small.

In any case, making broad statements along the lines of "Sigil invalidates my epub files so you need to be careful" is, at the very least, unproductive.

[ksng]

I have tested a few ePUB files with Sigil. I validated the files first to make sure they passed the epub check and then i opened on of them in Sigil and tried to change the bullet points. I saved the file and tried to validate again and now it failed.

I'm not sure what's wrong with my Sigil but this obviously showed that Sigil was causing the errors? I tried to install the 32 and 64 version but both have the same problem. Could you please advise if there is something that i have done wrong that caused the errors? I also wish that i can use sigil to edit my file without errors as the tool is very easy to use.

thanks!

[Valloric]

Originally Posted by ksng

Could you please advise if there is something that i have done wrong that caused the errors? I also wish that i can use sigil to edit my file without errors as the tool is very easy to use.

What errors are you seeing? You haven't said. You can create a new thread here and list the errors you are seeing; chances are someone will quickly point out the problem.

Then again, you can create an issue on the tracker with your epub files.

[Oldpilot]

I ran my Calibre epub through epubcheck at Threepress and got 24 errors. I am reasonably sure these are brain-dead things like not closing out a paragraph tag--but how do I find them? Is there a check that displays errors?

Interestingly, Barnes & Noble accepted the flawed epub without a murmur. Alas, there was a glaring error evidently introduce by Calibre: the first chapter was truncated, ending with the word Calibre. The balance of that chapter appeared as an untitled chapter three. So now I plan to start again in Sigil, using the rename route kindly mentioned above, whereby the bookname.zip file is renamed bookname.epub and imported. (Thanks for that!)

[Fabe]

Originally Posted by Oldpilot

I ran my Calibre epub through epubcheck at Threepress and got 24 errors. ... how do I find them? Is there a check that displays errors?

...the first chapter was truncated...

Threepress displays the errors. Open the file in Sigil and go to the code lines.

My experience with truncated text is the ampersand. "&" appears in the text instead of "&amp"

- Fabe

[ksng]

I will try to recapture the error text for you.

[Oldpilot]

Originally Posted by Fabe

Threepress displays the errors. Open the file in Sigil and go to the code lines.

My experience with truncated text is the ampersand. "&" appears in the text instead of "&amp"

- Fabe

Right! I went back and did it again with a clear head, and found that this was indeed the case. Ninety percent of the "errors" were actually just the absence of an alt= text within the images. Another recurring problem was my use of <small> to go down a size in the text, which we routinely do when writing html for Kindle books, but which evidently is a no-no in epub.

Thank you!

[Jellby]

Originally Posted by Oldpilot

Another recurring problem was my use of <small> to go down a size in the text, which we routinely do when writing html for Kindle books, but which evidently is a no-no in epub.

<small> is perfectly valid in ePub, it must go inside a <p>, <div> or other block element, though.

[Oldpilot]

Thanks for the news on <small>. I'll use <p> in the future.

Now I have another problem! (In another book.) This time I avoided most of the errors beforehand, but I have one that baffles me. It comes from Validate Epub Documents:

ERROR: OLGAEPUB.epub/OEBPS/content.opf(7): unfinished element

When I look at Sigil, I see a bunch of xhtml files under Text and a bunch of jpegs under Images, but I don't see the content.opf file. Where is it hiding?

(The epub file passes the Preflight check on the same Threepence website.)

Thanks again!

[Dave_S]

Originally Posted by Oldpilot

When I look at Sigil, I see a bunch of xhtml files under Text and a bunch of jpegs under Images, but I don't see the content.opf file. Where is it hiding?

Don't think that Sigil explicitly shows that one. Just open the EPUB with an un-zip tool like 7-ZIP and you will find the content.opf inside the OEBPS folder.

[Oldpilot]

Thanks! Found it. Still mystified. I can't find any unclosed element:

<?xml version="1.0" encoding="UTF-8"?>

<package xmlns="http://www.idpf.org/2007/opf" unique-identifier="BookID" version="2.0">

<metadata xmlns:dc="http://purl.org/dc/elements/1.1/" xmlnspf="http://www.idpf.org/2007/opf">

<dc:title></dc:title>

<dc:creator opf:role="aut"></dc:creator>

<dc:language>en</dc:language>

<dc:identifier id="BookID" opf:scheme="UUID">a0d1fd34-da99-4a38-bdca-773188174932</dc:identifier>

<meta name="cover" content="olgacover800.jpg"/>

<meta name="Sigil version" content="0.2.4"/>

</metadata>

(I spaced them out in order not to get them all run together, and I removed the book title and author. I make dc:identifier as line 7.)