EPUB output

[mikeywilliams25]

Thanks for the information.
Please expand on what I need to put into the level1-toc and --level2-toc settings.

[Icarusbop]

Originally Posted by kovidgoyal

All that means is that your source file references a missing stylesheet. Most probably your book will look fine even without the stylesheet.

The book appears empty (blank pages) in Adobe DE (but the TOC shows up OK), If I try to view it with Calibre's eBook viewer I get the following error:-

'Traceback (most recent call last):
File "calibre\gui2\viewer\documentview.pyo", line 326, in load_finished
File "main.py", line 347, in load_finished
AttributeError: 'NoneType' object has no attribute 'spine''

I checked the original OPF file - there is indeed a spine area of the file, I have attached the OPF file for reference.

[kovidgoyal]

@icarusbop: open a ticket and attach the file.
@mikeywilliams: try h2 for --level1-toc

[scottsan]

Originally Posted by kovidgoyal

Oops, looks like I forgot to propagate the fix. Will be in 97

Very frustrating!
Downloaded 0.4.97, then downloaded the ny times only to receive this error:

cannot import name NewMagickWand
Detailed traceback:
Traceback (most recent call last):
File "/Ebooks/A-SOFTWARE/calibre.app/Contents/Resources/lib/python2.5/site-packages.zip/calibre/ebooks/metadata/epub.py", line 123, in render_html
ImportError: cannot import name NewMagickWand

After quitting and restarting calibre the book was in the library.

Things haven't worked well since v.0.4.94.

What's up?

[=X=]

Originally Posted by mikeywilliams25

Thanks for the information.
Please expand on what I need to put into the level1-toc and --level2-toc settings.

Are you using html2epub and it's an HTML file created from BookCreator use

--level1-toc "//p[@class = '2CHAPTERTITLE']"

--level2-toc "//p[@class = '21CHAPTERSUBTITLE']"

=X=

[kovidgoyal]

Originally Posted by scottsan

Very frustrating!
Downloaded 0.4.97, then downloaded the ny times only to receive this error:

cannot import name NewMagickWand
Detailed traceback:
Traceback (most recent call last):
File "/Ebooks/A-SOFTWARE/calibre.app/Contents/Resources/lib/python2.5/site-packages.zip/calibre/ebooks/metadata/epub.py", line 123, in render_html
ImportError: cannot import name NewMagickWand

After quitting and restarting calibre the book was in the library.

Things haven't worked well since v.0.4.94.

What's up?

Hmm, forgot that ImportError is not asubclass of Exception. Will be fixed in 98 (and this time I mean it

[scottsan]

v.0.4.98 is working in a most excellent fashion.

Thanks!

[mikeywilliams25]

Originally Posted by =X=

Are you using html2epub and it's an HTML file created from BookCreator use

--level1-toc "//p[@class = '2CHAPTERTITLE']"

--level2-toc "//p[@class = '21CHAPTERSUBTITLE']"

=X=

No, I am simply loading the rtf file into calibre and then creating the epub from within calibre.
No luck yet with any entry in the toc slots.

[sahlberg]

Hi,

I opened a ticket for a book I have where html2epub 0.4.98 does not work :
http://calibre.kovidgoyal.net/ticket/1186

I have linked to a tgz file containing both the book (html + images) as well as the resulting epub file.
http://www.samba.org/~sahlberg/indarkestafrica.tgz


There are three issues,
1, during the conversion html2epub generates a lot of warnings
2, the output that is generated has become "stuck" in center alignment once the first chapter starts
3, the book can not be read at all in the sony reader and only gives a cryptic page error.


regards
ronnie sahlberg

[astrodad]

Lately I have been having problems converting .lit books with cover images (not the ones downloaded from LibraryThing). The cover images just disappear. I am using 0.4.98 right now. Any one else having this problem.

[kovidgoyal]

there were some changes to cover handling, perhaps a regression has been introduced. Open a ticket and attach one of the problem files.

[Thomas_]

Hi,

Is there a way to easily convert to epub with images? Like if you try to read a geopolitics review without the maps, it gets way less tasty. Can Calibre do that?

Thanks.

Thomas.

[kovidgoyal]

calibre's conversion preserves images

[thawk]

I have a html, which has a CSS style contains CJK font-face. I received the following errors with html2epub:
Traceback (most recent call last):
File "/usr/bin/html2epub", line 8, in <module>
load_entry_point('calibre==0.4.100', 'console_scripts', 'html2epub')()
File "/usr/lib/python2.5/site-packages/calibre/ebooks/epub/from_html.py", line 320, in main
convert(args[1], opts)
File "/usr/lib/python2.5/site-packages/calibre/ebooks/epub/from_html.py", line 261, in convert
parse_content(filelist, opts, tdir)
File "/usr/lib/python2.5/site-packages/calibre/ebooks/epub/from_html.py", line 121, in parse_content
hp.save()
File "/usr/lib/python2.5/site-packages/calibre/ebooks/epub/from_html.py", line 93, in save
Processor.save(self)
File "/usr/lib/python2.5/site-packages/calibre/ebooks/html.py", line 579, in save
open(path, 'wb').write(getattr(sheet, 'cssText', sheet).encode('utf-8'))
UnicodeDecodeError: 'ascii' codec can't decode byte 0xe6 in position 34: ordinal not in range(128)

The result of self.log_info('attr=%s' % getattr(sheet, 'cssText', sheet) is:
attr=.calibre_class_2 {
font-face: 楷体_GB2312
}
.calibre_class_3 {
color: #ff0000
}
.calibre_class_0 {
color: #000000
}

[kovidgoyal]

will be fixed in next release