Why Do Some Historical 1m-Kline Spot Data Files Not Have 1440 Entries for Non-Listing Days?

I attempted to download spot historical minute-level kline data from Binance Data Collection, and I noticed that for some dates and some symbols, the data file count for days other than the listing day is not 1440.
For example:

  • XLMBUSD-1m-2020-12-21.csv: 1188 rows, start from 2019-11-21
  • ADABTC-1m-2020-12-21.csv: 1188 rows, start from 2017-11-30
  • NEARBUSD-1m-2020-12-21.csv: 1188 rows, start from 2020-10-14
  • HOTETH-1m-2020-12-21.csv: 1188 rows, start from 2018-07-24
  • XTZUSDT-1m-2021-04-25.csv: 1157 rows, start from 2019-09-24

    (nearly 898 unusual symbols on 2020-12-21; 488 unusual symbols on 2021-04-25; on other dates, there are also a few symbols with an abnormal number of data entries scattered around.)

What could be the reason for this?

Most likely it’s because the trading was halted. In this case no candlesticks are produced, leaving a gap. For example, on 2020-12-21 the XLMBUSD stopped trading at 1608558442949 (13:47:22) and was trading again at 1608573600000 (18:00:00), skipping 252 minutes out of 1440 in the day.

I would like to inquire where I can find information on trading halts. I want to verify and confirm if the insufficient number of data entries in other examples is also due to trading halts.

@1067147135 I don’t think there is an endpoint for trading status history. That’s a presumption on my part.

Current trading status is available via /api/v3/exchangeInfo endpoint: https://api.binance.com/api/v3/exchangeInfo?symbol=XLMBUSD

{
  // ...
  "symbols": [
    {
      "symbol": "XLMBUSD",
      "status": "BREAK",
      // ...
    }
  ],
  // ...
}

No trades happen when symbol is on BREAK, and no new klines are recorded though.

The timestamps I referenced before were from the kline data.