Add third_party/collectorpayload module

go.mod

@@ -48,6 +48,7 @@ require (
 )
 
 require (
+	github.com/apache/thrift v0.17.0
 	github.com/davecgh/go-spew v1.1.1
 	github.com/dop251/goja v0.0.0-20220722151623-4765a9872229
 	github.com/hashicorp/hcl/v2 v2.13.0

third_party/snowplow/collectorpayload/README.md

@@ -0,0 +1,12 @@
+# collectorpayload
+
+This module is forked from the [Scala equivalent](https://github.com/snowplow/snowplow/tree/master/2-collectors/thrift-schemas/collector-payload-1).
+
+To generate the `gen-go/model1` directory:
+
+1. Install `thrift`: https://thrift.apache.org/download
+2. From this directory run `thrift -r --gen go collector_payload_1.thrift`
+
+_NOTE_: Running `make format` will fix formatting issues in the auto-generated code.
+
+To see how to use the library have a look at `collector_payload_test.go` which contains helper functions for seralizing and deserializing payloads.

third_party/snowplow/collectorpayload/collector_payload.go

@@ -0,0 +1,62 @@
+//
+// Copyright (c) 2023-present Snowplow Analytics Ltd. All rights reserved.
+//
+// This program is licensed to you under the Snowplow Community License Version 1.0,
+// and you may not use this file except in compliance with the Snowplow Community License Version 1.0.
+// You may obtain a copy of the Snowplow Community License Version 1.0 at https://docs.snowplow.io/community-license-1.0
+
+package collectorpayload
+
+import (
+	"context"
+	"encoding/base64"
+	"encoding/json"
+
+	thrift "github.com/apache/thrift/lib/go/thrift"
+
+	model1 "github.com/snowplow/snowbridge/third_party/snowplow/collectorpayload/gen-go/model1"
+)
+
+// BinarySerializer serializes a CollectorPayload into a byte array ready for transport
+func BinarySerializer(ctx context.Context, collectorPayload *model1.CollectorPayload) ([]byte, error) {
+	t := thrift.NewTMemoryBufferLen(1024)
+	p := thrift.NewTBinaryProtocolFactoryDefault().GetProtocol(t)
+
+	serializer := &thrift.TSerializer{
+		Transport: t,
+		Protocol:  p,
+	}
+
+	return serializer.Write(ctx, collectorPayload)
+}
+
+// BinaryDeserializer deserializes a CollectorPayload byte array back to a struct
+func BinaryDeserializer(ctx context.Context, collectorPayloadBytes []byte) (*model1.CollectorPayload, error) {
+	var inputBytes []byte
+
+	// Attempt to decode from base64 as most payloads will arrive with the thrift string re-encoded
+	base64DecodedCollectorPayload, base64Err := base64.StdEncoding.DecodeString(string(collectorPayloadBytes))
+	if base64Err != nil {
+		inputBytes = collectorPayloadBytes
+	} else {
+		inputBytes = []byte(base64DecodedCollectorPayload)
+	}
+
+	t := thrift.NewTMemoryBufferLen(1024)
+	p := thrift.NewTBinaryProtocolFactoryDefault().GetProtocol(t)
+
+	deserializer := &thrift.TDeserializer{
+		Transport: t,
+		Protocol:  p,
+	}
+
+	collectorPayload := model1.NewCollectorPayload()
+	err := deserializer.Read(ctx, collectorPayload, inputBytes)
+
+	return collectorPayload, err
+}
+
+// ToJSON converts the collector payload struct to a JSON representation for simpler portability
+func ToJSON(collectorPayload *model1.CollectorPayload) ([]byte, error) {
+	return json.Marshal(collectorPayload)
+}

third_party/snowplow/collectorpayload/collector_payload_1.thrift

@@ -0,0 +1,33 @@
+//
+// Copyright (c) 2023-present Snowplow Analytics Ltd. All rights reserved.
+//
+// This program is licensed to you under the Snowplow Community License Version 1.0,
+// and you may not use this file except in compliance with the Snowplow Community License Version 1.0.
+// You may obtain a copy of the Snowplow Community License Version 1.0 at https://docs.snowplow.io/community-license-1.0
+
+namespace go model1
+
+struct CollectorPayload {
+  31337: string schema
+
+  // Required fields which are intrinsic properties of HTTP
+  100: string ipAddress
+
+  // Required fields which are Snowplow-specific
+  200: i64 timestamp
+  210: string encoding
+  220: string collector
+
+  // Optional fields which are intrinsic properties of HTTP
+  300: optional string userAgent
+  310: optional string refererUri
+  320: optional string path
+  330: optional string querystring
+  340: optional string body
+  350: optional list<string> headers
+  360: optional string contentType
+
+  // Optional fields which are Snowplow-specific
+  400: optional string hostname
+  410: optional string networkUserId
+}

third_party/snowplow/collectorpayload/collector_payload_test.go

@@ -0,0 +1,36 @@
+//
+// Copyright (c) 2023-present Snowplow Analytics Ltd. All rights reserved.
+//
+// This program is licensed to you under the Snowplow Community License Version 1.0,
+// and you may not use this file except in compliance with the Snowplow Community License Version 1.0.
+// You may obtain a copy of the Snowplow Community License Version 1.0 at https://docs.snowplow.io/community-license-1.0
+
+package collectorpayload
+
+import (
+	"context"
+	"testing"
+
+	model1 "github.com/snowplow/snowbridge/third_party/snowplow/collectorpayload/gen-go/model1"
+
+	"github.com/stretchr/testify/assert"
+)
+
+// TestBinarySerializer
+func TestBinarySerializerAndDeserializer(t *testing.T) {
+	ctx := context.Background()
+
+	assert := assert.New(t)
+
+	payload := model1.NewCollectorPayload()
+	payload.IpAddress = "192.168.0.1"
+
+	res, err := BinarySerializer(ctx, payload)
+	assert.Nil(err)
+	assert.NotNil(res)
+
+	res1, err1 := BinaryDeserializer(ctx, res)
+	assert.Nil(err1)
+	assert.NotNil(res1)
+	assert.Equal("192.168.0.1", res1.IpAddress)
+}