EPUB output

[JeffElkins]

Originally Posted by kovidgoyal

0.4.91

Thanks Kovid,

epub-meta seems to be working, but unfortunately any2epub does not change meta info

any2epub -a "zzz zzzz" -t "XXXXXX" xxx.epub

epub-meta xxx.epub


Title    : Table of Contents
Author   : xxx xxxx
Producer : calibre (0.4.91) [http://calibre.kovidgoyal.net]
Series   : XXXXX #1
Language : Unknown

[kovidgoyal]

yeah any2epub doesn't. You have to call epub-meta to do that. any2epub is meant more for stuff like automatic splitting, font size control etc.

[JeffElkins]

Originally Posted by kovidgoyal

yeah any2epub doesn't. You have to call epub-meta to do that. any2epub is meant more for stuff like automatic splitting, font size control etc.

OK, thanks.

[=X=]

CSS is not working in 4.9.0

It looks like embedded style in an html are not used. In the creation of the ePUB HTML the styles are not included in the ePUB html.

=X=

[kovidgoyal]

the styles are extracted and put into a separate CSS file

[sconnolly]

First off, thank you for this great piece of software. Just what I was looking for!. I am attempting to standardize all my books to epub . I then use mobigen_linux to convert to a palm format. However it fails with an invalid language error. Inspection of the generated metadata.opf shows language as "Unknown".

metadata:

<metadata xmlns:dc="http://purl.org/dc/elements/1.1/" xmlnspf="http://www.idpf.org/2007/opf">
<dc:title>A Spell for Chameleon</dc:title>
<dc:creator opf:role="aut">Piers Anthony</dc:creator>
<dc:contributor opf:role="bkp" opf:files-as="calibre">calibre (0.4.91) [http://calibre.kovidgoyal.net]</dc:contributor>
<dc:identifier opf:scheme="calibre" id="calibre_id">33a0bfae-ae76-4efa-9aaf-df9c9d842eac</dc:identifier>
<dc:language>Unknown</dc:language>
<dcublisher>DELREY</dcublisher>
<series>Xanth</series>
<series-index>1</series-index>
<rating>5</rating>
<dc:subject>fantasy</dc:subject>
</metadata>

Is there a way to specify a language? I looked but found nothing. 2 items I noticed with the generated opf. I believe the elements series, series-index and rating should be placed inside meta tags as follows:

<meta name="series" content="Xanth" />
<meta name="series-index" content="1" />
<meta name="rating" content="5" />

per http://www.openebook.org/2007/opf/OP...ec.html#TOC2.2

and in the element contributor the attribute files-as should be file-as.

I am a COBOL programmer (please, no dinosaur jokes!) but this project has me looking into python. Keep up the good work.

[Hadrien]

Originally Posted by sconnolly

Is there a way to specify a language?

Use the language attribute in DublinCore (dc:language) and the xml:lang attribute in your XML flows.

[JeffElkins]

I opened a ticket with two books attached. In book 1, there is underlining of paragraphs every three or four pages and in book 2 each line of text is only three or four characters wide. Previously, these books converted w/o these errors.

[JeffElkins]

Pagination is different with these epubs. It generally takes at least two page forwards to advance the page count. Also I'm seeing the page number in extremely tiny print embedded about mid-page.

[=X=]

Originally Posted by kovidgoyal

the styles are extracted and put into a separate CSS file

Right I assume so, but that is not what is happening. Here I've attached a HTML and an ePUB to illistrate the problem


BTW the --level1-toc and --level2-toc work great.

[kovidgoyal]

Originally Posted by JeffElkins

Pagination is different with these epubs. It generally takes at least two page forwards to advance the page count. Also I'm seeing the page number in extremely tiny print embedded about mid-page.

That's a feature of Adobe DE, it counts a page as a certain font and resolution independent size. The number of screens that make a logical page therefore varies. As for the page number on the right, another annoying "feature" of DE.

[kovidgoyal]

Originally Posted by =X=

Right I assume so, but that is not what is happening. Here I've attached a HTML and an ePUB to illistrate the problem


BTW the --level1-toc and --level2-toc work great.

Oops, will be fixed in the next release.

[JeffElkins]

Originally Posted by kovidgoyal

That's a feature of Adobe DE, it counts a page as a certain font and resolution independent size. The number of screens that make a logical page therefore varies. As for the page number on the right, another annoying "feature" of DE.

I figured as much, FBReader does something similar.

[Selzernium]

I am having trouble converting three books [lrf conversion is OK] they multi page/file html[ziped]. I have opened a ticket and attached the files hope you can find a fix.

[kovidgoyal]

released version 0.4.92 with various fixes.