-
Notifications
You must be signed in to change notification settings - Fork 15.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Protobuf is really missing inheritance #5645
Comments
FWIW To quote the Protocol Buffers tutorial: Instead of inheritance, Protocol Buffers tends to favor composition as a pattern. |
GRPC is not doing "favor composition over inheritance" pattern, it's applying "remove inheritance" pattern. :D There are cases when inheritance is more suitable. Just because it's a good practice to "favor somehing", it doesn't mean it should be removed completely (or not implemented at all). Real life example
Yes, most of the time we should use public transport because it's better for Earth. On the other hand, you have cases when driving a car is a better option, e.g. if there is no public transport in your surroundings. |
How are you supposed to implement inheritance in a language that doesn't support inheritance like Rust or Go which only support struct-like data types? |
For proto3, use Any. |
I want to have a base type for all server messages, request and response, then I want to have different types of requests/responses. |
For one project I did something like this to minimize redundancy: message Request {
uint64 magic = 1; // Optional unique identifier for this request
oneof _Request_oneof {
ProfileRequest profile_request = 3;
// ...other kinds of requests
}
}
message Response {
uint64 magic = 1; // Your optional unique identifier echoed back so you may match our responses with your requests
oneof _Response_oneof {
string error = 2; // Defined instead of your expected response below, on failure
ProfileResponse profile_response = 3;
// ...other kinds of responses
}
}
message Message {
// ... auth stuff
repeated Request requests = 4;
repeated Response responses = 5;
string status = 6; // Protocol-level status or error message
} So a Russian doll, basically: a |
…tion (#1413) ## What does this PR do? This PR standardizes fury cross-language serialization specification. It comes with following changes: - Remove type tag from the protocol since it introduce space and performance overhead to the implementation. The `type tag` version can be seen in https://github.com/apache/incubator-fury/blob/6ea2e0b83d5449d63ca62296ff0dfd67b96c5bc5/docs/protocols/xlang_object_graph_spec.md . - Fury preserves `0~63` for internal types, but let users register type by id from `0`(added by 64 automatically) to setup type mapping between languages. - Streamline the type systems, only `bool/byte/i16/i32/i64/half-float/float/double/string/enum/list/set/map/Duration/Timestamp/decimal/binary/array/tensor/sparse/tensor/arrow/record/batch/arrow/table` are allowed. - Formulized the binary format for above types. - Add type disambiguation: the deserialization are determined by data type in serialized binary and target type jointly. - Introduce meta string encoding algorithm for field name to reduce space cost by 3/8. - Introduce schema consist mode format for struct. - Introduce schema envolution mode for struct: - this mode can embeed meta in the data or share across multiple messages, - it can avoid the cost of type tag comparison in frameworks like protobuf This protocol also supports object inheriance for xlang serializaiton. This is a feature request that users has been discussed for a long time in protobuf/flatbuffer: - google/flatbuffers#4006 - protocolbuffers/protobuf#5645 Although there are some languages such as `rust/golang` doesn't support inheriance, there are many cases only langauges like `java/c#/python/javascript` are involved, and the support for inheriance is not complexed in the protocol level, so we added the inheriance support in the protocol. And in languages such as `rust/golang`, we can use some annotation to mark composition field as parent class for serialization layout, or we can disable inheriance foor such languages at the protocol level. The protocol support polymorphic natively by type id, so I don't include types such as `OneOf/Union`. With this protocol, you can even serialize multiple rust `dyn trait` object which implement same trait., and get exactly the same objects when deserialization. ## Related issue This PR Closes #1418 --------- Co-authored-by: Twice <[email protected]>
What language does this apply to?
proto3 OR proto-future
Describe the problem you are trying to solve.
Lot's of DRY code in statically typed languages.
Describe the solution you'd like
I'd like you to add inheritance to current proto3 or at least the future protobuf versions. An example is:
Alternatives you've considered
I know about composition and it's really not a cure.
Additional context
I know that protobuf favors composition over inheritance but it's really not a cure to remove inheritance completely. Statically typed languages like Java or C# suffer from composition only approach. These languages cannot use polymorphism because all the messages have the same parent
GeneratedMessageV3
. If inheritance was available, we could use polymorphism. Instead of having to write:we could simply write
void method(Parent arg)
. Consequently, the missing inheritance generates a lot of DRY code is hardly or not at all DRYable. It leads to wider test suites that take more time to execute. I could go on but I hope I explained myself. (If not, just ask and I'll clarify it.)The text was updated successfully, but these errors were encountered: