DE upgrade ePub problems

[Steven Lyle Jordan]

I've recently come across a problem in DE handling new ePub files. Has this happened to anyone else?

Ever since the new mandatory upgrade to DE (v1.6.934), I've discovered that any new ePub files I create using WinZip will not open. Any attempt to open a new file gives the following message:

"The document appears to have minor errors that might cause it to be displayed incorrectly
--- errorListChange ---"

I've tried opening new ePubs created through Book Glutton, and they open fine. However, all I have to do is convert the .epub file extension to .zip, convert back to .epub (with or without actually opening the .zip file), and the ePub file will not open, and I get the message above.

The only thing that has changed in this equation is DE. I'm using WinZip 8, which I've already used successfully to create 2 ePub files before the DE upgrade.

As I have a series of books that I'd like to be able to offer for sale in ePub, and Book Glutton does not provide the full format options I'm looking for, any help would be appreciated.

[wallcraft]

ADE might be inforcing the mimetype file first requirement (but it is hard to see why). See Explode and Implode an ePub?, the 2nd post contains a simple way to do this properly under Windows (once I worked out that "explorer" just means right clicking on a folder). Perhaps WinZip is using non-standard compression, which is again fixed/tested by trying a different Zip program

However, if renaming the file and then renaming it back is enough to cause a problem I am out of ideas.

[kovidgoyal]

what are you generationg your epub files from?

[Steven Lyle Jordan]

Originally Posted by kovidgoyal

what are you generationg your epub files from?

My ePub creation has been a strictly manual process so far. I'm working in WinXP, have WinZip 8, and use Dreamweaver for any HTML editing.

I used the Jedisaber ePub files as a beginning template. I start by manually copying and pasting existing (and working) ePub files into a new folder, and editing as needed (titles, dates, chapter names, etc). Then I manually add each chapter as an HTML file, and images as JPG files. I use the same filetype descriptions that were in the original (working) ePub files.

The Book Glutton files use a slightly different formatting than the Jedisaber template, but both of their files open in DE, and after converting to .zip and back to .epub, will no longer open.

I haven't upgraded WinZip yet, but it is the only other obvious thing I can try. Still, I don't see how simply changing the extension could cause this problem.

[Steven Lyle Jordan]

Originally Posted by wallcraft

ADE might be inforcing the mimetype file first requirement (but it is hard to see why).

I did find and try this out, making sure the mimetype went in first, and uncompressed. The ePub file still did not open.

[kovidgoyal]

Originally Posted by Steve Jordan

My ePub creation has been a strictly manual process so far. I'm working in WinXP, have WinZip 8, and use Dreamweaver for any HTML editing.

I used the Jedisaber ePub files as a beginning template. I start by manually copying and pasting existing (and working) ePub files into a new folder, and editing as needed (titles, dates, chapter names, etc). Then I manually add each chapter as an HTML file, and images as JPG files. I use the same filetype descriptions that were in the original (working) ePub files.

The Book Glutton files use a slightly different formatting than the Jedisaber template, but both of their files open in DE, and after converting to .zip and back to .epub, will no longer open.

I haven't upgraded WinZip yet, but it is the only other obvious thing I can try. Still, I don't see how simply changing the extension could cause this problem.

Wow that's a lot of work. Since you already have your chapters as HTML files, create a simple cover HTML file (one that has a table of contents) that looks like this:

.
.
.
<h2>Table of contents</h2>
<a href="chapter1.htm">Title of chapter 1</a>
.
.
.

In the top part of the file you can have title, dedication etc. The important thing is that it should have links to all the chapter files, in the order that you want them.

Now make sure that in all your chapter files, you mark the tags that have the chapter titles with class="chapter_title", so for e.g. in a chapter file

.
.
.
<h2 class="chapter_title">Chapter title</h2>
.
.
.

You can choose any class you want, instead of chapter_title, so long as it unique to chapter titles.

Then just run html2epub (part of calibre) on the table of contents file (lets call it index.html)

html2epub --chapter '//*[@class = "chapter_title"]' index.html -o mybook.epub --title booktitle --author "Steve Jordan"

And voila! You have a nice EPUB file called mybook.epub that automatically has all your chapter files, your images, stylesheets etc included.

If you dont want the table of contents to be displayed in the final epub, you can enclose it like so:

<div style="display:none">
<h2>Table of contents</h2>
<a href="chapter1.htm">Title of chapter 1</a>
</div>

[Steven Lyle Jordan]

Good news: I tried a test of a Calibre-generated file from basic HTML, and it did open in DE (with some minor issues, which are mostly from the original HTML file, I believe). At any rate, I'll have to try again with the full setup you described above, and see if I can organize the entire file as desired.

One note, Kovid: For some reason, Calibre runs veerrrryyyyy sslllooooowwwwly on my PC, which is why I've spent very little time using it in the past. (I get impatient... what can I say?) But my PC also doesn't run multimedia well, so I'm pretty sure it's more a problem with my HW than your SW.

[kovidgoyal]

Glad it worked. Do the command line tools also run slowly?

Note that if you already have lit/mobi versions you can run:

any2epub file.it

instead and not have to worry about creating the setup.

[Steven Lyle Jordan]

Originally Posted by kovidgoyal

Glad it worked. Do the command line tools also run slowly?

Haven't tried them yet (my command-line usage is rusty, too!).

Originally Posted by kovidgoyal

Note that if you already have lit/mobi versions you can run:

any2epub file.it

instead and not have to worry about creating the setup.

I'll try a few variations, and see what comes out best. Thanks!

[=X=]

Originally Posted by kovidgoyal

.
<h2 class="chapter_title">Chapter title</h2>
.

html2epub --chapter '//*[@class = "chapter_title"]'

.
<p class=2CHAPTERTITLE>HASTIE LANYON</P>
.

So what would be the XPath syntax for the --chapter switch the sample code above?

=X=

[kovidgoyal]

Originally Posted by =X=

.
<p class=2CHAPTERTITLE>HASTIE LANYON</P>
.

So what would be the XPath syntax for the --chapter switch the sample code above?

=X=

'//p[@class = "2CHAPTERTITLE"]'

[=X=]

@kovidgoyal thanks, Last question is it possible to look for two tags I'm looking for 2CHAPTERTITLE and 2CHAPTERSUBTITLE (notice the SUB in the latter attribute)


=X=

[kovidgoyal]

XPath supports the OR operator |

It also support regular expression based matches.

[Peter Sorotokin]

Steve,

Any chance you can post the files that do not open in Digital Editions, so I can take a look?

Peter Sorotokin
Adobe Systems Inc.
Digital Publishing

[Steven Lyle Jordan]

Originally Posted by Peter Sorotokin

Steve,

Any chance you can post the files that do not open in Digital Editions, so I can take a look?

I can do that... unfortunately, I'm at work now, so it will have to be this evening, when I get back home.