Incorrect text extraction

[AndyTheFactory]

Issue by somnathrakshit Fri Jul 21 14:40:55 2017 Originally opened as https://github.com/codelucas/newspaper/issues/405

---

The incorrect text is being extracted from these two links at http://newspaper-demo.herokuapp.com for these web pages:

1. http://www.independent.co.uk/news/business/news/uk-budget-deficit-brexit-interest-rates-economy-government-bank-england-a7852436.html
2. http://www.independent.co.uk/news/business/news/tsb-bank-customers-bank-account-eye-scanner-login-access-android-samsung-galaxy-s8-a7851416.html

Please check the reason.