Implementation:Datahub project Datahub ProtobufField

Knowledge Sources	Datahub_project_Datahub
Domains	Protobuf_Integration
Last Updated	2026-02-10 00:00 GMT

Overview

Represents a single field within a protobuf message, providing type mapping to DataHub SchemaFieldDataType, field path generation, and source comment extraction from protobuf source code locations.

Description

ProtobufField implements the ProtobufElement interface and models a field within a protobuf message definition. It is built using Lombok's @Builder pattern and wraps a FieldDescriptorProto along with its parent ProtobufMessage.

Key responsibilities:

• Type Mapping -- The schemaFieldDataType() method maps protobuf types (TYPE_DOUBLE, TYPE_STRING, TYPE_MESSAGE, TYPE_ENUM, etc.) to DataHub schema types (NumberType, StringType, RecordType, EnumType, etc.). Repeated fields are mapped to ArrayType.
• Field Path Generation -- The fieldPathType() method produces DataHub field path type tokens (e.g., [type=string], [type=array].[type=int]) used in the schema field path specification.
• Native Type Resolution -- The nativeType() method extracts the protobuf type name, returning the type keyword for primitives (e.g., int32) or the fully qualified message name for message types.
• Comment Extraction -- The comment() method traverses protobuf SourceCodeInfo.Location entries to find documentation comments associated with this field, supporting both top-level and nested type fields.
• Enum Support -- Methods isEnum(), getEnumDescriptor(), getEnumValues(), and getEnumValuesWithComments() provide access to enum type information and associated comments.
• OneOf Support -- The oneOfProto() method returns the OneofDescriptorProto if the field participates in a oneof group.

Equality is based on the fully qualified field name (parentMessage.fullName + "." + fieldName).

Usage

Use ProtobufField as a vertex in the ProtobufGraph to represent individual fields. It is created during graph construction and visited by SchemaFieldVisitor and ProtobufExtensionFieldVisitor to produce DataHub SchemaField entries.

Code Reference

Source Location

• Repository: Datahub_project_Datahub
• File: metadata-integration/java/datahub-protobuf/src/main/java/datahub/protobuf/model/ProtobufField.java

Signature

@Builder(toBuilder = true)
@Getter
@AllArgsConstructor
public class ProtobufField implements ProtobufElement {

    // Core accessors
    public String name();
    public String fullName();
    public String nativeType();
    public String fieldPathType();
    public boolean isMessage();
    public int sortWeight();
    public int getNumber();

    // Type information
    public SchemaFieldDataType schemaFieldDataType();
    public OneofDescriptorProto oneOfProto();
    public boolean isEnum();
    public Optional<EnumDescriptorProto> getEnumDescriptor();
    public List<EnumValueDescriptorProto> getEnumValues();
    public Map<String, String> getEnumValuesWithComments();

    // Source code metadata
    public String comment();
    public Stream<SourceCodeInfo.Location> messageLocations();

    // Visitor pattern
    public <T> Stream<T> accept(ProtobufModelVisitor<T> visitor, VisitContext context);
}

Import

import datahub.protobuf.model.ProtobufField;

I/O Contract

Input	Type	Description
protobufMessage	ProtobufMessage	Parent message containing this field
fieldProto	FieldDescriptorProto	The raw protobuf field descriptor
isNestedType	Boolean	Whether this field belongs to a nested message type

Output	Type	Description
schemaFieldDataType	SchemaFieldDataType	DataHub schema field data type mapping
fieldPathType	String	Field path type token (e.g., [type=string])
comment	String	Extracted documentation comment from proto source

Usage Examples

ProtobufField field = ProtobufField.builder()
    .protobufMessage(parentMessage)
    .fieldProto(fieldDescriptorProto)
    .isNestedType(false)
    .build();

// Get DataHub schema field type
SchemaFieldDataType dataType = field.schemaFieldDataType();

// Get field path type for schema field path specification
String pathType = field.fieldPathType(); // e.g., "[type=string]"

// Check if field is an enum and extract values
if (field.isEnum()) {
    Map<String, String> enumValues = field.getEnumValuesWithComments();
}