There are rules that define what zserio fundamental type is mapped to what c++/java/python type.
For code maintainability, it is better that the mapped types are not used directly. Instead there should be a typedef for every mapped type in the proper namespace in the emitted code.
See https://github.com/ndsev/zserio/blob/master/doc/ZserioTypesMapping.md#base-types
The idea would be to create a type definition in the emitted code for every mentioned Zserio type (first column of the table).
And use these type definitions in the generated code (for example in the getXXX() and setXXX() methods for a field called XXX in a zserio struct).
There are rules that define what zserio fundamental type is mapped to what c++/java/python type. For code maintainability, it is better that the mapped types are not used directly. Instead there should be a typedef for every mapped type in the proper namespace in the emitted code.
See https://github.com/ndsev/zserio/blob/master/doc/ZserioTypesMapping.md#base-types The idea would be to create a type definition in the emitted code for every mentioned Zserio type (first column of the table). And use these type definitions in the generated code (for example in the getXXX() and setXXX() methods for a field called XXX in a zserio struct).
For example:
Zserio should emit a typedef for 'varint':
And given the following struct definition:
We should get something like the following:
instead of: