Recipe problem on Kindle DX

[jbambridge]

I have been trying to hack together an updated recipe for the Guardian (guardian.co.uk) and have had some success (ie images are now appearing in the feeds as well as headings).

I still have a problem however with the display of the TOC on the Kindle DX. Only the first feed is compete, with only one article being shown for the next feed. All other feeds are missing. However, when I page through the mobi document on the Kindle DX, all the articles are actually there.

Any ideas what I should look for in the download HTML files from ebook-converter?

John

[GRiker]

Display of the TOC has been modified in version 0.6 for the Kindle - it now uses the Kindle's native TOC and navigation features. Look at the the bottom of the screen and you should see 'View Articles List' or 'View Sections List'. Click the joystick to go to the Kindle TOC, and then you can select articles from that interface.

G

[elmoglick]

Originally Posted by GRiker

Display of the TOC has been modified in version 0.6 for the Kindle - it now uses the Kindle's native TOC and navigation features. Look at the the bottom of the screen and you should see 'View Articles List' or 'View Sections List'. Click the joystick to go to the Kindle TOC, and then you can select articles from that interface.

G

That's a great feature. I have a related question - Is there a TOC "flag" of some sort in mobi or azw files? I had to remove the DRM from a few magazines in order to transfer them to a replacement K2 and they lost that handy TOC in the process. Is there a way to restore it? Any idea what might have happened?

Thanks,
El

[jbambridge]

Originally Posted by GRiker

Display of the TOC has been modified in version 0.6 for the Kindle - it now uses the Kindle's native TOC and navigation features. Look at the the bottom of the screen and you should see 'View Articles List' or 'View Sections List'. Click the joystick to go to the Kindle TOC, and then you can select articles from that interface.

G

Thanks, but the Kindle TOC is working fine, just the data from the feed is not using it.

Other feeds (eg The Telegraph) all display n the TOC OK, but the output from my recipe for the Guardian only shows one topic in the Kindle TOC.

Are there specific HTML tags that the Kindle TOC uses to create the structure?

John

[GRiker]

Lots of flags and many data structures in a properly structured MOBI periodical. The process to remove the DRM probably ignored some necessary elements. I don't think it's a simple process to simply turn it back on.

G

Originally Posted by elmoglick

That's a great feature. I have a related question - Is there a TOC "flag" of some sort in mobi or azw files? I had to remove the DRM from a few magazines in order to transfer them to a replacement K2 and they lost that handy TOC in the process. Is there a way to restore it? Any idea what might have happened?

Thanks,
El

[GRiker]

Originally Posted by jbambridge

Thanks, but the Kindle TOC is working fine, just the data from the feed is not using it.

Other feeds (eg The Telegraph) all display n the TOC OK, but the output from my recipe for the Guardian only shows one topic in the Kindle TOC.

If you post your recipe I'll take a look to see if I can spot the problem.

Are there specific HTML tags that the Kindle TOC uses to create the structure?
John

In calibre, TOC structure is derived from the OPF and NCX files, which specify the section(s) and details like Title, Author and Description for the TOC. For recipes, these files are created dynamically. It is possible to create your own structured documents by providing properly constructed OPF and NCX files describing your HTML content, then wrapping everything up in an ePub for calibre to process. It's non-trivial, but not difficult.

G

[jbambridge]

If you post your recipe I'll take a look to see if I can spot the problem.

Thanks G. Here is my recipe:

PHP Code:

#!/usr/bin/env  python
__license__   = 'GPL v3'
__copyright__ = '2008, Kovid Goyal kovid@kovidgoyal.net'
__docformat__ = 'restructuredtext en'

'''
www.guardian.co.uk
'''

from calibre.web.feeds.news import BasicNewsRecipe

class Guardian(BasicNewsRecipe):

    title = u'My Guardian'
    language = _('English')
    encoding = 'utf-8'
    oldest_article = 7
    max_articles_per_feed = 20
    remove_javascript     = True
    simultaneous_downloads = 1
    use_embedded_content = False
    recursions = 0
    filter_regexps = [r'\.g\.doubleclick\.net']

    timefmt = ' [%a, %d %b %Y]'

    keep_only_tags = [dict(id=['article-wrapper', 'main-article-info'])]




    no_stylesheets = True
    extra_css = 'h2 {font-size: medium;} \n h1 {text-align: left;}'


    feeds = [
       ('Front Page',     'http://feeds.guardian.co.uk/theguardian/rss'),
#        ('UK',             'http://feeds.guardian.co.uk/theguardian/uk/rss'),
#        ('Business',       'http://www.guardian.co.uk/business/rss'),
#        ('Politics',       'http://feeds.guardian.co.uk/theguardian/politics/rss'),
#        ('Culture',        'http://feeds.guardian.co.uk/theguardian/culture/rss'),
#        ('Money',          'http://feeds.guardian.co.uk/theguardian/money/rss'),
#        ('Life & Style',   'http://feeds.guardian.co.uk/theguardian/lifeandstyle/rss'),
#        ('Travel',         'http://feeds.guardian.co.uk/theguardian/travel/rss'),
#        ('Environment',    'http://feeds.guardian.co.uk/theguardian/environment/rss')
        ]

    def print_version(self, url):
        return url + '/print' 


Any help very welcome.

John

[GRiker]

The hash marks (#) in front of the feeds after 'Front Page' cause the rest of the line to be ignored. Remove the hash marks for the feeds you want to receive, and you should be good to go.

If you want to limit the number of articles in each section, change the '20' in 'max_articles_per_feed = 20' to a lower number.

G

[kiklop74]

No need for all that complication since guardian feeds contain complete text of every article:

class Guardian_uk(BasicNewsRecipe):
    title                 = 'The Guardian'
    oldest_article        = 2
    max_articles_per_feed = 100
    no_stylesheets        = True
    use_embedded_content  = True
    encoding              = 'utf-8'
    language              = _('English')
    
    feeds              = [(u'News: Main Section', u'http://www.guardian.co.uk/theguardian/mainsection/rss')]

    remove_tags        = [dict(name='div', attrs={'class':['related','guRssAdvert','terms']})]

Just add more feeds you need

[jbambridge]

That's certainly a lot simpler.

Is there any way to retain the images referred to in the RSS feed under <media:content>?

John