These are the json formats for The Project Open Corpus.

books.json

{
  "at": 1356360832,
  "authors": [
    "J. M. Barrie"
  ],
  "name": "Peter Pan",
  "text": "All children, except one, grow up. They soon know that they will grow up, and the way Wendy knew ...",
  "raw": {} // bibliographic record
}

news.json

{
  "at": 1356360832,
  "source": "http://www.bbc.co.uk/turkce/haberler/2012/11/121107_obama_new_term.shtml",
  "title": "Turkey grows fast.",
  "text": "IMF has just reported that Turkey is the fastest count....",
  "raw": {} // raw news source object without text
}

twitter.json

{
  "at": 1356360832,
  "userId": "muatik2",
  "text": "Hello, this is a tweet.",
  "raw": {} // raw twitter object
}

Every language will have its own .json file.