Please have a look at BinarySchema and it's usages. Essentially this is a fingerprint of object fields in the order they were serialized, represented as [field name] -> [position] hash map. Consider and object with 3 fields - A, B, C. There could be 6 different schemas for it: [A, B, C], [A, C, B], [B, A, C], [B, C, A]. [C, A, B], [C, B, A].
We use schemas to achieve O(1) field lookup time. On Mon, Feb 13, 2017 at 11:05 AM, ALEKSEY KUZNETSOV < [email protected]> wrote: > What is schema in your context? > > пн, 13 февр. 2017 г. в 10:50, Vladimir Ozerov <[email protected]>: > > > Alex, > > > > FieldAccessor stores field positions for the given schema to allow for > fast > > field value lookup. > > > > On Fri, Feb 10, 2017 at 7:06 PM, ALEKSEY KUZNETSOV < > > [email protected] > > > wrote: > > > > > Guys, i have a question.Why do we store fields position information in > > > FieldAccessor before hashcode evaluation? Test example is in > > > : org.apache.ignite.internal.binary.BinaryFieldIdentityResolverSel > > > fTest#testHashCode > > > -- > > > > > > *Best Regards,* > > > > > > *Kuznetsov Aleksey* > > > > > > -- > > *Best Regards,* > > *Kuznetsov Aleksey* >
