Because the dataset contains so many symbols and the message bus relies on binary encoding, we need to map textual symbols to their numeric IDs.
The import tool already generated numeric ids for each symbol and for good reason because most of the conversion between test symbols and numeric IDs happens automatically behind the scenes. Please note, the symbol to id mapping is generated once during import and is stored in the database without any updates.
The symbol manager provides methods to convert symbols and numeric IDs back and forth, which helps to implement clients efficiently.
The data importer generates symbol id's on the fly to map a symbol name to an unique symbol id, which means when you import data on your system, your symbol mapping will differ from the example code in this repo.
For example, when XBTEUR maps to 202 in the examples, you will almost certainly see a different number on your system after the data import.
This is deemed acceptable for research and back testing, but clearly insufficient for production systems.
If you want to figure out the which symbol corresponds to a certain ID, or the other way around, you can query the database. Suppose you want to know which symbol belongs to ID 42:
SELECT symbol, symbol_id
FROM kraken_symbols
WHERE symbol_id=42;
Now we know that ID 42 corresponds to the Energy Web (EWT) quoted in EUR. By convention, the first currency within a currency pair is called the base, while the second currency is called the quote.
Conversely, when you want to know the ID corresponding to the symbol pair "ADAAUD", you can query the database:
SELECT symbol, symbol_id
FROM kraken_symbols
WHERE symbol='adaaud';
As you see, the symbol pair "ADAAUD" (ADA / Cardano, quoted in Australian Dollars AUD) corresponds to ID 49.