EPUB output

[kaytuss]

I havebeen usin Calibre for months with no issues. After updating to 4.137, I can no longe read NY Times subscription on the PRS-505. I see thetable of contents ad it shows the total number of pages but when i change to the page after the TOC, the screen goes blank.

Can yo direct m to where your previus files are so that I can go back to where I was before?

[kovidgoyal]

Originally Posted by kaytuss

I havebeen usin Calibre for months with no issues. After updating to 4.137, I can no longe read NY Times subscription on the PRS-505. I see thetable of contents ad it shows the total number of pages but when i change to the page after the TOC, the screen goes blank.

Can yo direct m to where your previus files are so that I can go back to where I was before?

This will likely have been caused by a change on the NYTimes website as there was no change associated with the feed fetching code in calibre, so returning to a previous version is unlikely to solve your problem. Open a ticket about it and I will fix it before the next release.

[ddavtian]

Originally Posted by kaytuss

I can no longe read NY Times subscription on the PRS-505. I see thetable of contents ad it shows the total number of pages but when i change to the page after the TOC, the screen goes blank.

I have the same problem on the reader, but Calibre viewer shows it just fine.
The second NY Times (not subscription) works fine.

[Student1]

Hi,

found a small problem when converting from html to epub, tried first changing all marging settings and page break settings, but when viewed on my sony 505 the epub format has missing letters on the right end. Is there a way to force the text to overflow rather than be be cut if it doesn t fit on the resolution of the device?


thanks

[kovidgoyal]

This happends when you source HTML uses non reflowable markup like <pre> tags of <table> tags. First try the --linearize-tables option in calibre

[Student1]

Thanks going to give it a try once caliber finishes importing a ton of books !

[Student1]

Was also wondering, when converting html to opf and then to epub is there a way to remove the header and footer of each html page without having to do it manually ?

thanks

[kovidgoyal]

If the headers and footers are in some sort of regular markup you can use override css to remove them, fo example if headers are always in markup of the form

<div class="header">some text</div>

you can use

.header { display:none }

in override css

[Student1]

Thanks that will work perfectly! Thanks for all your help and this great soft! Would be hell working with sony s software... and they are getting payed for it... amazing!

[klx14]

I can confirm the bug reported in #346 and #348. With the attached self-developed recipe I have exactly the same problem: Page count and TOCs are OK, but following a link gives a blank page only. Reverting to 0.4.136 solved the problem for me - so there is definitely a regression in 0.4.137. My suspicion is that there is a bug in the EPUB conversion routines, not the fetching.

Here is the recipe I am using:

class AdvancedUserRecipe1231145686(BasicNewsRecipe):
    title          = u'Tagesspiegel'
    encoding = 'cp1252'
    oldest_article = 7
    max_articles_per_feed = 100
    no_stylesheets = True
    remove_tags = [dict(attrs={'class':['related', 'themen_alarm',
                                        'tso_tool', 'postings ps_swus_nowrap',
                                        'anzeige_spacer', 'tool2',
                                        'anzeige_top_double_border',
                                        'ISI_IGNORE ps_rs_versteckt',
                                        'ISI_IGNORE tso_tools',
                                        'login_box', 'tso_small',
                                        'aufmacher_content',
                                        'tso_breadcrumb', 'tso_headline']}),
                   dict(id=['footer', 'toolsRight', 'articleInline',
                            'navigation', 'archive', 'side_search',
                            'blog_sidebar', 'side_tool', 'thumblupe',
                            'thumbbig_var', 'linkbox_spacer',
                            'navbar', 'header', 'search', 'masthead',
                            'banner', 'sky', 'sky_move']),
                   dict(name=['script', 'noscript', 'style',
                              'pinboardbtn'])]
    remove_tags_after  = dict(id='ar_text')
    import re
    preprocess_regexps = [
        (re.compile(r'float:left;', re.DOTALL|re.IGNORECASE),
         lambda match: ''),
        ]

    extra_css = '#ar_text {font-family:times; font-size:9pt}'
    cover_url = 'file:///home/peterh/eBooks/tagesspiegel.png'
    html2lrf_options = ['--base-font-size', '9']
    
    feeds          = [(u'Aktuelle Nachrichten aus dem Ressort Politik', u'http://www.tagesspiegel.de/rss/politik.xml'), (u'Aktuelle Nachrichten aus Berlin und Brandenburg', u'http://www.tagesspiegel.de/rss/berlin.xml'), (u'Aktuelle Nachrichten aus dem Ressort Wirtschaft', u'http://www.tagesspiegel.de/rss/wirtschaft.xml'), (u'Aktuelle Nachrichten aus dem Ressort Kultur', u'http://www.tagesspiegel.de/rss/kultur.xml'), (u'Buch-Rezensionen aus dem gedruckten Tagesspiegel', u'http://www.tagesspiegel.de/rss/literatur.xml'), (u'Aktuelle Nachrichten aus aller Welt', u'http://www.tagesspiegel.de/rss/welt.xml')]

BTW on Linux you can revert by issuing:

easy_install -U calibre==0.4.136
calibre_postinstall

[Student1]

Originally Posted by kovidgoyal

If the headers and footers are in some sort of regular markup you can use override css to remove them, fo example if headers are always in markup of the form

<div class="header">some text</div>

you can use

.header { display:none }

in override css

Tried it but it didn t work, words are still cut on the right side, well one letter or two so its not so bad, would there be a css overide to force the code to cut from the left so the right can show? Would enable to get that missing letter and loose some white space on the left side .

thanks!

[kovidgoyal]

The roblem with the NYT (and hopefully the other feeds as well) will be fixed in the next release (basically it was caused by Adobe DE not being able to handle HTML comments with code in them. calibre now strips the comments, so it should be fine.

[kaytuss]

Originally Posted by kovidgoyal

The roblem with the NYT (and hopefully the other feeds as well) will be fixed in the next release (basically it was caused by Adobe DE not being able to handle HTML comments with code in them. calibre now strips the comments, so it should be fine.

Thanks for your help. 4.138 corrected the NY Times issue. It now works like a gem.

[yacoob]

A small question: is there a way to set 'collection' the book prepared with any2epub is going to end up with? After looking at the options, --subjects looks related, but looks like it's not working.

[kovidgoyal]

Use the --tags option of the epub-meta command