BeautifulSoup: cleaning article text further - python-2.7

Question

I need to count characters in a news article. Some pages have lots of stuff I don't need (navigation, footer etc.). I managed to get rid of all these but I still have a couple of things like images copyright, image and video captions and adverts I struggle to remove. Could anyone suggest how to improve the code below to get only useful text from the article?
import requests
from bs4 import BeautifulSoup
r = requests.get("http://www.bbc.co.uk/news/world-europe-39612562")
soup = BeautifulSoup(r.content)
for s in soup.findAll("div", {"class":"story-body__inner"}):
article = ''.join(s.findAll(text=True))
print(article)
print (len(article))
The code for this particular url yields this (top part just to illustrate the problem):
Image copyright
AFP
Image caption
Erdogan supporters began celebrating early outside party headquarters in Ankara
Turks have backed President Recep Tayyip Erdogan's call for sweeping new presidential powers, partial official results of a referendum indicate.With about 98% of ballots counted, "Yes" was on about 51.3% and "No" on about 48.7%.Erdogan supporters say replacing the parliamentary system with an executive presidency would modernise the country. Opponents have attacked a decision to accept unstamped ballot papers as valid unless proven otherwise.The main opposition Republican People's Party (CHP) is already demanding a recount of 60% of the votes.
/**/
(function() {
if (window.bbcdotcom && bbcdotcom.adverts && bbcdotcom.adverts.slotAsync) {
bbcdotcom.adverts.slotAsync('mpu', [1,2,3]);
}
})();
/**/

Answer 1

it seems that you don't need the script nor the figure tags, so:
import requests
from bs4 import BeautifulSoup
r = requests.get("http://www.bbc.co.uk/news/world-europe-39612562")
soup = BeautifulSoup(r.content)
# delete unwanted tags:
for e in soup(['figure', 'script']):
e.decompose()
article_soup = [e.get_text() for e in soup.find_all(
'div', {'class': 'story-body__inner'})]
article = ''.join(article_soup)
print(article)
print (len(article))