India Seminar Magazine

[unkn0wn]

This is a really old interface website but with really good content.

https://www.india-seminar.com/

This link - https://www.india-seminar.com/semframe.html will take you directly to the present issue

the article links contain only text [example] for which auto_cleanup will work.

All i want is for the recipe to parse article links to feed.

from calibre.web.feeds.news import BasicNewsRecipe, classes
        
class Seminar(BasicNewsRecipe):
    title = 'India-Seminar Magazine'
    __author__ = 'unkn0wn'
    description = 'Seminar - attempts a departure from the usual journal. Problems, national and international, are posed and discussed. Each issue deals with a single problem. Those who hold different and at times opposing viewpoints express their thoughts'
    language = 'en_GB'
    encoding = 'utf-8'
    use_embedded_content = False
    no_stylesheets = True
    remove_javascript = True
    masthead_url = 'https://www.india-seminar.com/semlogo/semlogo_top_1.jpg'
    ignore_duplicate_articles = {'url'}
    
    auto_cleanup = True
        
    def parse_index(self):
        soup = self.index_to_soup('https://www.india-seminar.com/semframe.html')
        ans = []

        for a in soup.findAll('a', href=lambda x: x and x.startswith(''751/'')):
            url = a['href']
            title = self.tag_to_string(a)
            self.log(title, ' at ', url)
            ans.append({'title': title, 'url': url})
        return [('Articles', ans)]
# 751/ would need to be changed to 752/ in the next month.

hrefs' can be found here

<frame src="2022/751.htm" target="_self" name="action">
#document
html body div.WordSection1 p.MsoNormal b span a

I've tried so many things, nothing worked. I thought this would be really easy.

It wont even give me an error.. just empty epub.

[kovidgoyal]

You dont load https://www.india-seminar.com/semframe.html instead yo load the inner document 2022/751.htm and parse that.

[unkn0wn]

from calibre.web.feeds.news import BasicNewsRecipe
from datetime import datetime
     
class Seminar(BasicNewsRecipe):
    title = 'India-Seminar Magazine'
    __author__ = 'unkn0wn'
    description = 'Seminar - attempts a departure from the usual journal. Problems, national and international, are posed and discussed. Each issue deals with a single problem. Those who hold different and at times opposing viewpoints express their thoughts'
    language = 'en_GB'
    use_embedded_content = False
    remove_javascript = True
    masthead_url = 'https://www.india-seminar.com/semlogo/semlogo_top_1.jpg'
    
    
    
    ignore_duplicate_articles = {'url'}
        
    def parse_index(self):
        d = datetime.today()
        soup = self.index_to_soup('https://www.india-seminar.com/' + d.strftime('%Y') + '/751.htm')
        ans = []
        
        for a in soup.findAll('a', href=lambda x: x):
            url = a['href']
            if url.startswith('751/'):
                url = 'https://www.india-seminar.com/' + d.strftime('%Y') + '/' + url
            title = self.tag_to_string(a)
            self.log(title, ' at ', url)
            ans.append({'title': title, 'url': url})
        return [('Articles', ans)]

This works great.. but is there's a way to get the number to change with months..

like %m (is month number) and add it to 748

so that jan will be 748+1
dec will become 748+12

[kovidgoyal]

You shouldnt need to calculate the month, just get it from the href in the page containing the frame.

[unkn0wn]

oh.. Thanks.

It works good now.

[kovidgoyal]

https://github.com/kovidgoyal/calibr...43e4072c0c71d6