query-orchestration/api/docCleanRunner/runner_test.go

package doccleanrunner_test

import (
	"context"
	"fmt"
	"io"
	"strings"
	"testing"
	"time"

	doccleanrunner "queryorchestration/api/docCleanRunner"
	"queryorchestration/internal/database/repository"
	"queryorchestration/internal/document"
	documentclean "queryorchestration/internal/document/clean"
	"queryorchestration/internal/server/runner"
	"queryorchestration/internal/serviceconfig/objectstore"
	"queryorchestration/internal/serviceconfig/queue/documenttexttrigger"
	objectstoremock "queryorchestration/mocks/objectstore"
	queuemock "queryorchestration/mocks/queue"

	"github.com/aws/aws-sdk-go-v2/service/s3"
	"github.com/aws/aws-sdk-go-v2/service/sqs"
	"github.com/google/uuid"
	"github.com/pashagolub/pgxmock/v3"
	"github.com/stretchr/testify/assert"
	"github.com/stretchr/testify/mock"
	"github.com/stretchr/testify/require"
)

type DocCleanConfig struct {
	runner.BaseConfig[doccleanrunner.Body]
	documenttexttrigger.DocTextTriggerConfig
	objectstore.ObjectStoreConfig
}

func TestDocCleanRunner(t *testing.T) {
	ctx := context.Background()

	pool, err := pgxmock.NewPool()
	require.NoError(t, err)

	cfg := &DocCleanConfig{}
	cfg.DBPool = pool
	cfg.DBQueries = repository.New(pool)
	mockStore := objectstoremock.NewMockS3Client(t)
	cfg.StoreClient = mockStore
	mockSQS := queuemock.NewMockSQSClient(t)
	cfg.QueueClient = mockSQS
	cfg.DocumentTextTriggerURL = "/i/am/here"

	runner := doccleanrunner.New(&doccleanrunner.Services{
		Clean: documentclean.New(cfg),
	})

	t.Run("valid", func(t *testing.T) {
		bod := doccleanrunner.Body{
			DocumentID: uuid.New(),
		}

		doc := document.DocumentSummary{
			ID:       bod.DocumentID,
			ClientID: "hello",
			Hash:     "example_hash",
		}
		bucket := "hello"
		key := objectstore.BucketKey{
			CreatedAt: time.Now().UTC(),
			ClientID:  "hi",
			EntityID:  uuid.New(),
			Location:  objectstore.Import,
		}
		keyStr := key.String()
		dbid := doc.ID

		pool.ExpectQuery("name: HasDocumentCleanEntry :one").WithArgs(dbid).WillReturnRows(
			pgxmock.NewRows([]string{"isclean"}).
				AddRow(false),
		)
		pool.ExpectQuery("name: GetDocumentSummary :one").WithArgs(dbid).
			WillReturnRows(
				pgxmock.NewRows([]string{"id", "clientId", "hash"}).
					AddRow(dbid, doc.ClientID, doc.Hash),
			)
		pool.ExpectQuery("name: GetDocumentEntry :one").WithArgs(dbid).WillReturnRows(
			pgxmock.NewRows([]string{"documentId", "bucket", "key"}).
				AddRow(dbid, bucket, keyStr),
		)
		mimeType := "application/pdf"
		dbmimetype := repository.NullCleanmimetype{
			Valid:         true,
			Cleanmimetype: repository.Cleanmimetype(mimeType),
		}
		pool.ExpectBegin()
		cleanid := uuid.New()
		pool.ExpectQuery("name: GetMostRecentDocumentCleanEntry :one").WithArgs(dbid).
			WillReturnRows(
				pgxmock.NewRows([]string{"id", "documentId", "bucket", "key", "version", "hash", "mimetype", "fail"}),
			)
		pool.ExpectQuery("name: AddDocumentClean :one").WithArgs(dbid, &bucket, &keyStr, &doc.Hash, dbmimetype, repository.NullCleanfailtype{}).
			WillReturnRows(
				pgxmock.NewRows([]string{"id"}).
					AddRow(cleanid),
			)
		pool.ExpectExec("name: AddDocumentCleanEntry :exec").WithArgs(cleanid, pgxmock.AnyArg()).
			WillReturnResult(pgxmock.NewResult("", 1))
		pool.ExpectCommit()

		mockSQS.EXPECT().
			SendMessage(
				mock.Anything,
				mock.MatchedBy(func(in *sqs.SendMessageInput) bool {
					return *in.QueueUrl == cfg.DocumentTextTriggerURL && *in.MessageBody == fmt.Sprintf("{\"id\":\"%s\"}", doc.ID.String())
				}),
				mock.Anything,
			).
			Return(&sqs.SendMessageOutput{}, nil)

		mockStore.EXPECT().
			HeadObject(
				mock.Anything,
				mock.MatchedBy(func(in *s3.HeadObjectInput) bool {
					return *in.Bucket == bucket && *in.Key == keyStr
				}),
				mock.Anything,
			).
			Return(&s3.HeadObjectOutput{
				ContentType: &mimeType,
			}, nil)

		bodyMsg := io.NopCloser(strings.NewReader(pdfHelloWorld))
		mockStore.EXPECT().
			GetObject(
				mock.Anything,
				mock.MatchedBy(func(in *s3.GetObjectInput) bool {
					return *in.Bucket == bucket && *in.Key == keyStr
				}),
				mock.Anything,
			).
			Return(&s3.GetObjectOutput{
				Body: bodyMsg,
			}, nil)

		assert.True(t, runner.Process(ctx, bod))
	})
}

const pdfHelloWorld = `%PDF-1.4
%����
1 0 obj
<<
  /Type /Catalog
  /Pages 2 0 R
  /Version /1.4
>>
endobj
2 0 obj
<<
  /Type /Pages
  /Kids [3 0 R]
  /Count 1
>>
endobj
3 0 obj
<<
  /Type /Page
  /Parent 2 0 R
  /MediaBox [0 0 612 792]
  /Resources <<
    /Font <<
      /F1 <<
        /Type /Font
        /Subtype /Type1
        /BaseFont /Helvetica
      >>
    >>
  >>
  /Contents 4 0 R
>>
endobj
4 0 obj
<<
  /Length
  44
>>
stream
BT
/F1 24 Tf
100 700 Td
(Hello World!) Tj
ET
endstream
endobj
xref
0 5
0000000000 65535 f
0000000015 00000 n
0000000086 00000 n
0000000151 00000 n
0000000376 00000 n
trailer
<<
  /Size 5
  /Root 1 0 R
>>
startxref
472
%%EOF`