Performance of reading every time from a concurrent hashmap vs. assigning a value to a variable then using it

Question

Code 1:

listModel.getStockList().forEach((contractName, v) -> {
    if (listModel.getContractDetails().get(contractName).getAskPrice() >
            listModel.getArchieveContractDetails().get(contractName).getHighPrice()
            && validatorUtils.checkOrderValidity1(Action.BRAC_ORDER_1, contractName)) {
        method1(listModel.getContractDetails().get(contractName).getAskPrice());
        method2(0, listModel.getContractDetails().get(contractName).getAskPrice());
        syso(listModel.getContractDetails().get(contractName).getAskPrice());
    }

});

Code 2:

listModel.getStockList().forEach((contractName, v) -> {
    int askPrice = listModel.getContractDetails().get(contractName).getAskPrice();
    if (askPrice > listModel.getArchieveContractDetails().get(contractName).getHighPrice()
            && validatorUtils.checkOrderValidity1(Action.BRAC_ORDER_1, contractName)) {
        method1(askPrice);
        method2(0, askPrice);
        syso(askPrice);
    }
});

In code 1, I am reading price directly from getaskPrice() ; and in code 2, I used an askPrice variable.

Which approach will give me better performance and why?

Answer 1

Well, storing the result of get in a variable and then using the variable multiple times should definitely never be any worse, performance-wise, than repeating the get multiple times. Whereas the Hashmap implementation might, or might not , perform comparably well on multiple get s of the same key.

But consider that you are in complete control of how you write your Java code, but you are not necessarily in control of how good the HashMap implementation is where your Java code runs.

Seems better to err on the side of caution and take control over the things you have control over (ie, do write your Java code to save the result of get to be reused multiple times), rather than hope for the best in things you have little or no control over (ie, don't expect you'll always be running on a JDK with a "good" HashMap implementation).