You need a special parser for Reuters headlines. They are to be taken exactly as written, since they are designed to be displayed as one liners in a financial

You need a special parser for Reuters headlines. They are to be taken exactly as written, since they are designed to be displayed as one liners in a financial terminal, like in this image: https://i.insider.com/4bc623be7f8b9a2b3a870500 https://i.insider.com/4bc623be7f8b9a2b3a870500 They actually put a lot of work into their titles.