Change justification sometimes does not work

[TomasHnyk]

I try out my recipe with

ebook-convert my_recipe.recipe output.mobi --password pass --username username --test 10 --output-profile kindle --change-justification justify -vv --debug-pipeline debug

and the justification is changed to justify (even though the original is not justified).

When I load the recipe into gui via Fetch News>Add a custom news source and I set "Text Justitifaction" to "Justify text" in Preferences>Conversion>Common Options>Look and Feel and I run the recipe, justification is not changed.

If I set no_stylesheets = True in my recipe, justification is changed. When I run conversion from mobi to mobi, the justification is changed to justify as well.

to me, it seems to be a bug unless --change-justification justify is intended to mean something else then the corresponding option in the gui.

This was also discussed here:
http://www.mobileread.com/forums/showthread.php?p=2672301

(The recipe is here. It normally needs an account but this is visible even without one.)

#!/usr/bin/python
# -*- coding: utf-8 -*-
# License: GNU General Public License v3 - http://www.gnu.org/copyleft/gpl.html
# Copyright: <email address hidden>

__license__ = 'GNU General Public License v3 - http://www.gnu.org/copyleft/gpl.html'
__copyright__ = '<email address hidden>'

import re
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import BeautifulSoup,Tag
#This imports the version bundled with Calibre
import lxml
from lxml.builder import E

class respektRecipe(BasicNewsRecipe):
    __author__ = 'Tomáš Hnyk'
    title = u'Respekt'
    publisher = u'Respekt Publishing a. s.'
    description = u'Articles from the printed edition without translations from The Economist that are not available online'
    encoding = 'cp1250'
    language = 'cs'
    remove_javascript = True
    extra_css = 'ul {color:black} .image_caption {font-size:50%;font-style:italic;}, .author {text-align:left;}'
    remove_tags_before = dict(name='div',attrs={'class':['l']})
    remove_tags_after = dict(id='text')
    remove_tags = [dict(name='ul', attrs={'class':['tabs-d'],'id':['comm']}), \
    dict(name='div',attrs={'class':['slot','reklama','date']}), \
    dict(name='span', attrs={'class':['detail-vykrik']}), \
    dict(name='p', attrs={'class':['detail-vykrik']}), \
    dict(name='div', attrs={'id':['col123d-video','col123d-infographic','col123d-gallery','col12d-discussion']}), # soup>lxml>soup in prprocess requires this
    dict(name='strong', attrs={'class':['detail-vykrik']})]
    # this makes authors left-aligned by not using the author class)
    preprocess_regexps = [(re.compile(r'<div class="author">', re.DOTALL|re.IGNORECASE), lambda match: '<div class="">')]
    # remove empty tags
    preprocess_regexps.append((re.compile(r'<strong> </strong>', re.DOTALL|re.IGNORECASE), lambda match: ' '))
    preprocess_regexps.append((re.compile(r'<strong>&nbsp;</strong>', re.DOTALL|re.IGNORECASE), lambda match: '&nbsp;'))
    preprocess_regexps.append((re.compile(r'<p></p>', re.DOTALL|re.IGNORECASE), lambda match: ''))

    def get_cover_url(self):
        soup = self.index_to_soup('http://respekt.ihned.cz/')
        cover = soup.findAll('div', attrs={'class':'cover'})[0].find('img')['src']
        return cover

    #needs_subscription = True

    """
    def get_browser(self):
        br = BasicNewsRecipe.get_browser(self)
        if self.username is not None and self.password is not None:
            br.open('http://muj-ucet.ihned.cz/')
            br.select_form(name='login')
            br['login[nick]'] = self.username
            br['login[pass]'] = self.password
            br.submit()
        return br
    """

    def parse_index(self):
        raw = self.index_to_soup('http://respekt.ihned.cz/aktualni-cislo/', raw=True)
        root = lxml.html.fromstring(raw)
        ans = []
        for article in root.xpath("//div[@class='ow-enclose']/div[@class='ow']"):
            section_title = article.xpath(".//span[text()='(rubrika: ']")[0].find("a").text
            date = article.xpath("//span[@class='date-author']")[0].text[:-3]
            author = article.xpath("//span[@class='date-author']")[0].find("a").text
            title = article.find("h2").find("a").text
            url = article.find('h2').find('a').get('href')
            link = {'title':title,'url':url,'date':date,'author':author}
            for section in ans:
                if section[0] == section_title:
                    section[1].append(link)
                    break
            else:
                ans.append((section_title,[link]))
        return ans

    def cleanup(self):
        self.browser.open('http://muj-ucet.ihned.cz/?login[logout]=1')

    def preprocess_html(self,soup):
        raw = u''.join(unicode(a) for a in soup.contents)
        root = lxml.html.fromstring(raw)

        # Make image captions visible
        body = root.xpath("//div[@id='text']")[0]
        add = 0
        for index, element in enumerate(body):
            try:
                if element.tag == 'img':
                    body.insert(index+add+1,E.p(element.get('title'),{"class":"image_caption"}))
                    add += 1
            except:
                pass

        # Add length in words after author
        article_length = str(len(body.text_content().split(' '))) + ' slov'
        root.xpath("//div[@class='author-image']/div[@class='']/ul")[0].append(E.li(article_length))

        # Make perex (subheading) start on a new line
        root.xpath("//h1")[0].append(E.br(''))

        return(BeautifulSoup(lxml.etree.tostring(root,encoding=unicode)))

[TomasHnyk (sup)]

Eh, the line is this:
ebook-convert my_recipe.recipe output.mobi --test 1 --output-profile kindle --change-justification justify -vv --debug-pipeline debug

And this is on Calibre 1.10, Ubuntu 13.10.

[Kovid Goyal (kovid)]

THe GUI does not apply conversion options (with a few specific
exceptions) to news downloads. If you want to apply a conversion option
unconditionally to the recipe, use conversion_options in the recipe
itself.

 status invalid