query-orchestration/internal/document/clean/clean_test.go

package documentclean

import (
	"context"
	"io"
	"queryorchestration/internal/database"
	"queryorchestration/internal/database/repository"
	"queryorchestration/internal/document"
	cleanversion "queryorchestration/internal/document/clean/version"
	objectstoremock "queryorchestration/mocks/objectstore"
	"strings"
	"testing"

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

func TestClean(t *testing.T) {
	ctx := context.Background()
	pool, err := pgxmock.NewPool()
	if err != nil {
		t.Fatalf("failed to open pgxmock database: %v", err)
	}

	cfg := &DocCleanConfig{}
	cfg.DBPool = pool
	cfg.DBQueries = repository.New(pool)
	mockS3 := objectstoremock.NewMockS3Client(t)
	cfg.StoreClient = mockS3

	svc := Service{
		cfg: cfg,
		svc: &Services{
			Version: cleanversion.New(cfg),
		},
	}

	doc := document.Document{
		ID:    uuid.New(),
		JobID: uuid.New(),
		Hash:  "example_hash",
	}
	inloc := document.Location{
		Bucket: "bucket_name",
		Key:    "/i/am/here",
	}
	dbid := database.MustToDBUUID(doc.ID)

	pool.ExpectQuery("name: GetDocument :one").WithArgs(dbid).
		WillReturnRows(
			pgxmock.NewRows([]string{"id", "jobId", "hash"}).
				AddRow(dbid, database.MustToDBUUID(doc.JobID), doc.Hash),
		)
	pool.ExpectQuery("name: GetDocumentEntry :one").WithArgs(dbid).WillReturnRows(
		pgxmock.NewRows([]string{"documentId", "bucket", "key"}).
			AddRow(dbid, inloc.Bucket, inloc.Key),
	)
	mimeType := "application/pdf"
	dbmimetype := repository.NullCleanmimetypes{
		Valid:          true,
		Cleanmimetypes: repository.Cleanmimetypes(mimeType),
	}
	pool.ExpectExec("name: AddDocumentCleanEntry :exec").WithArgs(dbid, int32(1), &inloc.Bucket, &inloc.Key, dbmimetype, repository.NullCleanfailtype{}).
		WillReturnResult(pgxmock.NewResult("", 1))

	var length int64 = 10
	mockS3.EXPECT().
		HeadObject(
			mock.Anything,
			mock.MatchedBy(func(in *s3.HeadObjectInput) bool {
				return *in.Bucket == inloc.Bucket && *in.Key == inloc.Key
			}),
			mock.Anything,
		).
		Return(&s3.HeadObjectOutput{
			ContentType:   &mimeType,
			ContentLength: &length,
		}, nil)

	body := io.NopCloser(strings.NewReader(pdfHelloWorld))
	mockS3.EXPECT().
		GetObject(
			mock.Anything,
			mock.MatchedBy(func(in *s3.GetObjectInput) bool {
				return *in.Bucket == inloc.Bucket && *in.Key == inloc.Key
			}),
			mock.Anything,
		).
		Return(&s3.GetObjectOutput{
			Body: body,
		}, nil)

	err = svc.clean(ctx, doc.ID)
	assert.NoError(t, err)
}

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

	cfg := &DocCleanConfig{}
	mockS3 := objectstoremock.NewMockS3Client(t)
	cfg.StoreClient = mockS3
	svc := Service{
		cfg: cfg,
	}

	id := uuid.New()
	hash := "example_hash"
	inloc := document.Location{
		Bucket: "bucket_name",
		Key:    "/i/am/here",
	}

	mimeType := "application/pdf"
	mockS3.EXPECT().
		HeadObject(
			mock.Anything,
			mock.MatchedBy(func(in *s3.HeadObjectInput) bool {
				return *in.Bucket == inloc.Bucket && *in.Key == inloc.Key
			}),
			mock.Anything,
		).
		Return(&s3.HeadObjectOutput{
			ContentType: &mimeType,
		}, nil)

	body := io.NopCloser(strings.NewReader(pdfHelloWorld))
	mockS3.EXPECT().
		GetObject(
			mock.Anything,
			mock.MatchedBy(func(in *s3.GetObjectInput) bool {
				return *in.Bucket == inloc.Bucket && *in.Key == inloc.Key
			}),
			mock.Anything,
		).
		Return(&s3.GetObjectOutput{
			Body: body,
		}, nil)

	outloc, err := svc.executeCleanTasks(ctx, &CleanParams{
		ID:       id,
		Hash:     hash,
		Location: inloc,
	})
	assert.NoError(t, err)
	assert.EqualExportedValues(t, ExecuteCleanResponse{
		location: &inloc,
	}, *outloc)
}

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`