feat: add protoc plugin for generating json BigQuery schemas

encoding/jsonbq/marshal.go

@@ -22,11 +22,29 @@ type Field struct {
 	Fields      []*Field           `json:"fields,omitempty"`
 }
 
-// MarshalSchema marshals a BigQuery schema to a valid BigQuery JSON schema.
-// The JSON conforms to the format in this example:
+// MarshalSchema marshals a BigQuery schema to a valid BigQuery JSON schema based on
+// default MarshalOptions. The JSON conforms to the format in this example:
 // https://cloud.google.com/bigquery/docs/nested-repeated#example_schema
 func MarshalSchema(schema bigquery.Schema) ([]byte, error) {
+	return MarshalOptions{}.MarshalSchema(schema)
+}
+
+// MarshalOptions is a configurable BigQuery schema json marshaller.
+type MarshalOptions struct {
+	// Indent specifies the set of indentation characters to use in a formatted
+	// output such that every entry is preceded by Indent and terminated by a newline.
+	// Leave Indent empty to disable formatted output.
+	Indent string
+}
+
+// MarshalSchema marshals a BigQuery schema to a valid BigQuery JSON schema based on
+// the given MarshalOptions. The JSON conforms to the format in this example:
+// https://cloud.google.com/bigquery/docs/nested-repeated#example_schema
+func (o MarshalOptions) MarshalSchema(schema bigquery.Schema) ([]byte, error) {
 	fields := convertSchema(schema)
+	if o.Indent != "" {
+		return json.MarshalIndent(fields, "", o.Indent)
+	}
 	return json.Marshal(fields)
 }
 

encoding/protobq/schema.go

@@ -29,11 +29,11 @@ type SchemaOptions struct {
 // InferSchema infers a BigQuery schema for the given proto.Message using options in
 // MarshalOptions.
 func (o SchemaOptions) InferSchema(msg proto.Message) bigquery.Schema {
-	return o.inferMessageSchema(msg.ProtoReflect().Descriptor())
+	return o.InferMessageSchema(msg.ProtoReflect().Descriptor())
 }
 
-// inferMessageSchema infers the BigQuery schema for the given protoreflect.MessageDescriptor.
-func (o SchemaOptions) inferMessageSchema(msg protoreflect.MessageDescriptor) bigquery.Schema {
+// InferMessageSchema infers the BigQuery schema for the given protoreflect.MessageDescriptor.
+func (o SchemaOptions) InferMessageSchema(msg protoreflect.MessageDescriptor) bigquery.Schema {
 	schema := make(bigquery.Schema, 0, msg.Fields().Len())
 	for i := 0; i < msg.Fields().Len(); i++ {
 		fieldSchema := o.inferFieldSchema(msg.Fields().Get(i))
@@ -64,7 +64,7 @@ func (o SchemaOptions) inferFieldSchema(field protoreflect.FieldDescriptor) *big
 		Repeated: field.IsList(),
 	}
 	if fieldSchema.Type == bigquery.RecordFieldType && fieldSchema.Schema == nil {
-		fieldSchema.Schema = o.inferMessageSchema(field.Message())
+		fieldSchema.Schema = o.InferMessageSchema(field.Message())
 		if len(fieldSchema.Schema) == 0 {
 			return nil
 		}

go.mod

@@ -7,6 +7,7 @@ require (
 	cloud.google.com/go/bigquery v1.22.0
 	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
 	github.com/google/go-cmp v0.5.6
+	github.com/spf13/pflag v1.0.3
 	golang.org/x/net v0.0.0-20210825183410-e898025ed96a // indirect
 	golang.org/x/oauth2 v0.0.0-20210819190943-2bc19b11175f
 	golang.org/x/text v0.3.7 // indirect

go.sum

@@ -173,6 +173,7 @@ github.com/prometheus/client_model v0.0.0-20190812154241-14fe0d1b01d4/go.mod h1:
 github.com/rogpeppe/fastuuid v1.2.0/go.mod h1:jVj6XXZzXRy/MSR5jhDC/2q6DgLz+nrA6LYCDYWNEvQ=
 github.com/rogpeppe/go-internal v1.3.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/spaolacci/murmur3 v0.0.0-20180118202830-f09979ecbc72/go.mod h1:JwIasOWyU6f++ZhiEuf87xNszmSA2myDM2Kzu9HwQUA=
+github.com/spf13/pflag v1.0.3 h1:zPAT6CGy6wXeQ7NtTnaTerfKOsV6V6F8agHXFiazDkg=
 github.com/spf13/pflag v1.0.3/go.mod h1:DYY7MBk1bdzusC3SYhjObp+wFpr4gzcvqqNjLnInEg4=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
 github.com/stretchr/testify v1.4.0/go.mod h1:j7eGeouHqKxXV5pUuKE4zz7dFj8WfuZ+81PSLYec5m4=

protoc-gen-bq-json-schema/genjson/config.go

@@ -0,0 +1,16 @@
+package genjson
+
+import "github.com/spf13/pflag"
+
+type Config struct {
+	Format bool
+}
+
+func (c *Config) AddToFlagSet(flags *pflag.FlagSet) {
+	flags.BoolVar(
+		&c.Format,
+		"format",
+		false,
+		"Set to true to get a formatted json output.",
+	)
+}

protoc-gen-bq-json-schema/genjson/message.go

@@ -0,0 +1,24 @@
+package genjson
+
+import (
+	"fmt"
+
+	"go.einride.tech/protobuf-bigquery/encoding/jsonbq"
+	"go.einride.tech/protobuf-bigquery/encoding/protobq"
+	"google.golang.org/protobuf/compiler/protogen"
+)
+
+func GenerateSchemaFile(gen *protogen.Plugin, msg *protogen.Message, config Config) error {
+	g := gen.NewGeneratedFile(fmt.Sprintf("%s.json", msg.Desc.Name()), "")
+	schema := protobq.SchemaOptions{}.InferMessageSchema(msg.Desc)
+	jsonOpt := jsonbq.MarshalOptions{}
+	if config.Format {
+		jsonOpt.Indent = "  "
+	}
+	out, err := jsonOpt.MarshalSchema(schema)
+	if err != nil {
+		return err
+	}
+	g.P(string(out))
+	return nil
+}

protoc-gen-bq-json-schema/genjson/run.go

@@ -0,0 +1,19 @@
+package genjson
+
+import (
+	"google.golang.org/protobuf/compiler/protogen"
+)
+
+func Run(gen *protogen.Plugin, config Config) error {
+	for _, f := range gen.Files {
+		if !f.Generate {
+			continue
+		}
+		for _, msg := range f.Messages {
+			if err := GenerateSchemaFile(gen, msg, config); err != nil {
+				return err
+			}
+		}
+	}
+	return nil
+}

protoc-gen-bq-json-schema/main.go

@@ -0,0 +1,16 @@
+package main
+
+import (
+	"github.com/spf13/pflag"
+	"go.einride.tech/protobuf-bigquery/protoc-gen-bq-json-schema/genjson"
+	"google.golang.org/protobuf/compiler/protogen"
+)
+
+func main() {
+	flagSet := pflag.NewFlagSet("protoc-gen-bq-json-schema", pflag.ContinueOnError)
+	var config genjson.Config
+	config.AddToFlagSet(flagSet)
+	protogen.Options{ParamFunc: flagSet.Set}.Run(func(gen *protogen.Plugin) error {
+		return genjson.Run(gen, config)
+	})
+}