kcinnick / pyredictit Goto Github PK

Thank you for writing this tool! I was trying to set up the repo and get it to work, but it appears that PredictIt has nontrivially changed its API. I tried updating your code to use the new API endpoints, but I ran into some trouble. I don't know much about web development, but it looks like as a prerequisite to putting in a buy order, the browser makes a GET request to https://www.predictit.org/api/Trade//OrderBook. Python returns a 401 when I try to hit this URL; and even when I try to hit it directly in the browser, on a session where I'm logged in, I get a 401. Don't know why that's happening, but presumably there's some sort of authentication that happens on the contracts page where you can only proceed if you're on that page? Would appreciate if you could take a look and see if you can tell what's going on.

Here's my updated function that attempts to use the new API:

def buy_shares(self, api, number_of_shares, buy_price):
    # this URL is wrong, it's not self.cid, it's some other id, not sure what it's an id for. but it doesn't
    # work with the correct id either (I tried it directly in the browser)
    order_book_page = api.browser.get(f'https://www.predictit.org/api/Trade/{self.cid}/OrderBook')
    all_bids_and_offers = order_book_page.json()

    confirm = api.browser.post(f'https://www.predictit.org/api/Trade/ConfirmTrade', {
        'quantity': number_of_shares,
        'pricePerShare': buy_price,
        'contractId': self.cid,
        'tradeType': 1,
    })
    if str(confirm.status_code) != '200':
        print(f"Request returned an invalid {confirm.status_code} code. Please make sure you're using valid login credentials.")

I'm wondering if the point is to better emulate the browser as opposed to reverse engineering their API endpoints.