How To Serialize Information Successfully with Protocol Buffers – Grape Up

In a world of microservices, we frequently need to cross data between purposes. We serialize knowledge right into a format that may be retrieved by each side. One of many serialization options is Protocol Buffers (Protobuf) – Google’s language-neutral mechanism. Messages might be interpreted by a receiver utilizing the identical or completely different language than a producer. Many languages are supported, reminiscent of Java, Go, Python, and C++.

An information construction is outlined utilizing impartial language by way of .proto recordsdata. The file is then compiled into code for use in purposes. It’s designed for efficiency. Protocol Buffers encode knowledge into binary format, which reduces message dimension and improves transmission pace.

Defining Message Format

This .proto the file represents geolocation data for a given car.

1 syntax = "proto3";
3 package deal com.grapeup.geolocation;
5 import "google/kind/latlng.proto";
6 import "google/protobuf/timestamp.proto";
8 message Geolocation 
9  string vin = 1;
10  google.protobuf.Timestamp occurredOn = 2;
11  int32 pace = 3;
12  google.kind.LatLng coordinates = 4;
1 syntax = "proto3";

Syntax refers to Protobuf model, it may be proto2 or proto3.

1package com.grapeup.geolocation;

Bundle declaration prevents naming conflicts between completely different initiatives.

1 message Geolocation 
2  string vin = 1;
3  google.protobuf.Timestamp occurredOn = 2;
4  int32 pace = 3;
5  google.kind.LatLng coordinates = 4;

Message definition comprises a reputation and a set of typed fields. Easy knowledge sorts can be found, reminiscent of bool, int32, double, string, and so on. You too can outline your personal sorts or import them.

1google.protobuf.Timestamp occurredOn = 2;

The = 1, = 2 markers establish the distinctive tag. Tags are a numeric illustration for the sphere and are used to establish the sphere within the message binary format. They need to be distinctive in a message and shouldn’t be modified as soon as the message is in use. If a discipline is faraway from a definition that’s already used, it have to be reserved.

Subject sorts

Except for scalar sorts, there are various different kind choices when defining messages. Listed below are few, however you’ll find all of them within the Language Information Language Guide (proto3)  |  Protocol Buffers  |  Google Developers .

Properly Recognized Sorts

1 import "google/kind/latlng.proto";
2 import "google/protobuf/timestamp.proto";
4 google.protobuf.Timestamp occurredOn = 2;
5 google.kind.LatLng coordinates = 4;

There are predefined sorts accessible to make use of Overview  |  Protocol Buffers  |  Google Developers . They’re referred to as Properly Know Sorts and need to be imported into .proto .

LatLng represents a latitude and longitude pair.

Timestamp is a particular cut-off date with nanosecond precision.

Customized sorts

1 message SingularSearchResponse 
2  Geolocation geolocation = 1;

You should use your custom-defined kind as a discipline in one other message definition.


1 message SearchResponse 
2  repeated Geolocation geolocations = 1;

You’ll be able to outline lists through the use of repeated key phrase.


It may occur that in a message there’ll all the time be just one discipline set. On this case, TelemetryUpdate will include both geolocation, mileage, or gasoline stage data.

This may be achieved through the use of oneof. Setting worth to one of many fields will clear all different fields outlined in oneof.

1 message TelemetryUpdate 
2  string vin = 1;
3  oneof replace 
4    Geolocation geolocation = 2;
5    Mileage mileage =3;
6    FuelLevel fuelLevel = 4;
10 message Geolocation 
11  ...
14 message Mileage 
15  ...
18 message FuelLevel 
19  ...

Take into accout backward-compatibility when eradicating fields. If you happen to obtain a message with oneof that has been faraway from .proto definition, it is not going to set any of the values. This conduct is similar as not setting any worth within the first place.

You’ll be able to carry out completely different actions primarily based on which worth is ready utilizing the getUpdateCase() methodology.

1 public Elective<Object> getTelemetry(TelemetryUpdate telemetryUpdate) 
2        Elective<Object> telemetry = Elective.empty();
3        change (telemetryUpdate.getUpdateCase()) 
4            case MILEAGE -> telemetry = Elective.of(telemetryUpdate.getMileage());
5            case FUELLEVEL -> telemetry = Elective.of(telemetryUpdate.getFuelLevel());
6            case GEOLOCATION -> telemetry = Elective.of(telemetryUpdate.getGeolocation());
7            case UPDATE_NOT_SET -> telemetry = Elective.empty();
9        return telemetry;

Default values

In proto3 format fields will all the time have a price. Due to this proto3 can have a smaller dimension as a result of fields with default values are omitted from payload. Nevertheless this causes one concern – for scalar message fields, there isn’t any manner of telling if a discipline was explicitly set to the default worth or not set in any respect.

In our instance, pace is an non-obligatory discipline – some modules in a automobile would possibly ship pace knowledge, and a few may not. If we don’t set pace, then the geolocation object can have pace with the default worth set to 0. This isn’t the identical as not having pace set on messages.

In an effort to cope with default values you should utilize official wrapper sorts protobuf/wrappers.proto at main · protocolbuffers/protobuf . They permit distinguishing between absence and default. As a substitute of getting a easy kind, we use Int32Value, which is a wrapper for the int32 scalar kind.

1 import "google/protobuf/wrappers.proto";
3 message Geolocation 
4  google.protobuf.Int32Value pace = 3;

If we don’t present pace, it is going to be set to nil.

Configure with Gradle

When you’ve outlined your messages, you should utilize protoc, a protocol buffer compiler, to generate courses in a selected language. The generated class can then be used to construct and retrieve messages.

In an effort to compile into Java code, we have to add dependency and plugin in construct.gradle

1 plugins 
2    id '' model '0.8.18'
5 dependencies 
6    implementation ''

and setup the compiler. For Mac customers an osx particular model needs to be used.

1 protobuf 
2    protoc 
3        if (osdetector.os == "osx") 
4            artifact = "$protobuf_version:osx-x86_64"
5         else 
6            artifact = "$protobuf_version"

Code will probably be generated utilizing generateProto process.

The code will probably be positioned in construct/generated/supply/proto/principal/java in a package deal as laid out in .proto file.

We additionally want to inform gradle the place the generated code is positioned

1 sourceSets 
2    principal 
3        java 
4            srcDirs 'construct/generated/supply/proto/principal/grpc'
5            srcDirs 'construct/generated/supply/proto/principal/java'

The generated class comprises all the required strategies for constructing the message in addition to retrieving discipline values.

1 Geolocation geolocation = Geolocation.newBuilder()
2            .setCoordinates(LatLng.newBuilder().setLatitude(1.2).setLongitude(1.2).construct())
3            .setVin("1G2NF12FX2C129610")
4            .setOccurredOn(Timestamp.newBuilder().setSeconds(12349023).construct())
5            .construct();
7 LatLng coordinates = geolocation.getCoordinates();
8 String vin = geolocation.getVin();

Protocol Buffers – Abstract

As proven protocol buffers are simply configured. The mechanism is language agnostic, and it’s simple to share the identical .proto definition throughout completely different microservices.

Protobuf is definitely paired with gRPC, the place strategies might be outlined in .proto recordsdata and generated with gradle.

There’s official documentation accessible Protocol Buffers  |  Google Developers and guides Language Guide (proto3)  |  Protocol Buffers  |  Google Developers .