Either for comparisons or initialization of a new variable, does it make a difference which one of these you use?
I know that BigDecimal.ZERO is a 1.5 feature, so that's a concern, but assuming I'm using 1.5 does it matter?
Thanks.
Either for comparisons or initialization of a new variable, does it make a difference which one of these you use?
I know that BigDecimal.ZERO is a 1.5 feature, so that's a concern, but assuming I'm using 1.5 does it matter?
Thanks.
BigDecimal.ZERO
is a predefined constant and therefore doesn't have to be evaluated from a string at runtime as BigDecimal("0")
would be. It will be faster and won't require creation of a new object.
If your code needs to run on pre-1.5, then you can use the (much maligned) Singleton pattern to create an object equivalent to BigInteger.ZERO
. The first time it is used, it would call BigInteger("0")
to create a zero object, and return that object on subsequent calls. Otherwise, if your code is running on a 1.5 system, your singleton object can just return BigInteger.ZERO
with no runtime penalty.
Using ZERO doesn't create a new object or require any parsing. Definitely the way to go.
Out of curiosity I checked to constructor for BigDecimal and it doesn't have any optimizations for the "0" string. So definitely yes, there's a difference.
Before talking about runtime penalties make sure that this piece of code matters. Setup profiling and measure the complete use case.
Nevertheless prefer Bigdecimal.ZERO
as it's checked at compile time wheres you can accidently type new Bigdecimal("9")
which the compile will eat but like will cause bugs into your system.