Historic market cap data is incorrect

Describe the bug
Historic market cap data is incorrect

To Reproduce

client = RESTClient(api_key=api_key)
date = ["2004-03-30", "2004-04-01", "2017-03-30", "2017-04-01", "2026-03-27"]
ticker = "AAPL"
for d in date:
    details = client.get_ticker_details(ticker, date=d)
    print(f"date: {d}  market_cap: {int(details.market_cap)/1e9:.2f}B")

Output:
AAPL:
date: 2004-03-30 market_cap: 458.07B
date: 2004-04-01 market_cap: 444.78B
date: 2017-03-30 market_cap: 2361.37B
date: 2017-04-01 market_cap: 2356.94B
date: 2026-03-27 market_cap: 3652.67B

MSFT:
date: 2004-03-30 market_cap: 189.20B
date: 2004-04-01 market_cap: 188.30B
date: 2017-03-30 market_cap: 493.35B
date: 2017-04-01 market_cap: 494.48B
date: 2026-03-27 market_cap: 2649.24B

Expected behavior
AAPL is off by almost 2 magnitude on 2004-03-30 according to https://companiesmarketcap.com/apple/marketcap/

MSFT is off by ~80B on 2004-03-30 according to https://companiesmarketcap.com/microsoft/marketcap/

2017 data also off

Additional context
Additionally, is there other ways of querying for market cap and weighted_shares_outstanding of a ticker for a range of dates, other than to query get_ticker_details for each date separately?

Commencez par reproduire l’exemple Python fourni via RESTClient.get_ticker_details, puis comparez les valeurs historiques renvoyées avec les données citées de CompaniesMarketCap. Déterminez si l’écart provient du client ou du service, et établissez si une requête couvrant une plage de dates pour market cap et weighted_shares_outstanding est déjà prise en charge ou si une requête distincte est nécessaire.