The inventors' current setup is that the retrieval system database has only a few large currencies, requires that the custom data source must return a specific format, and provides some code for the transaction collector, which I do not understand. I have to retest a coin, I have to build a server myself, or I have to collect data myself. Is it possible that the inventor's retrieval system directly calls the API interface to retrieve the data? There is no need to set up a server, and no need to collect the data yourself. If the format fails, the user can adjust it himself. If you want to get tick-level data, it's easy to understand with a market gatherer, and if it's just a normal retest, it's really not necessary to do this.
For example: Binance now offers up to 1500 k-strings per call, and can be used to get 7500 k-strings through 5 looping calls, which is enough retesting.
The Little DreamThere is a transaction collector in the library, which automatically provides the data source, just by running a physical disk, it can not be easier than this. Because the retrieval system is on the web side of the platform, you must be able to access the data source to do it, or you have to get a server, you have data on your own computer, so that the public network can be accessed. Or as you say: > Example: Binance now offers up to 1500 k-strings per call, and can get 7500 k-strings through 5 looping calls, which is enough retesting. This is also the case, with python requests for exchange K-lines, and then through K-line retrieval, reference libraries, and several useful articles.