package doccleanrunner_test import ( "fmt" "io" "regexp" "strings" "testing" "time" doccleanrunner "queryorchestration/api/docCleanRunner" "queryorchestration/internal/database/repository" documentclean "queryorchestration/internal/document/clean" "queryorchestration/internal/server/runner" "queryorchestration/internal/serviceconfig/objectstore" "queryorchestration/internal/serviceconfig/queue/documenttext" "queryorchestration/internal/test" objectstoremock "queryorchestration/mocks/objectstore" "github.com/aws/aws-sdk-go-v2/service/s3" "github.com/google/uuid" "github.com/stretchr/testify/assert" "github.com/stretchr/testify/mock" "github.com/stretchr/testify/require" ) type DocCleanConfig struct { runner.BaseConfig[doccleanrunner.Body] documenttext.DocTextConfig objectstore.ObjectStoreConfig } func TestDocCleanRunner(t *testing.T) { cfg := &DocCleanConfig{} test.CreateDB(t, cfg) acfg := test.CreateAWSContainer(t, cfg) test.SetQueueClient(t, t.Context(), cfg, acfg.ExternalEndpoint) cfg.DocumentTextURL = test.CreateQueue(t, cfg, test.DocTextRunnerName) mockStore := objectstoremock.NewMockS3Client(t) cfg.StoreClient = mockStore runner := doccleanrunner.New(&doccleanrunner.Services{ Clean: documentclean.New(cfg), }) err := cfg.GetDBQueries().CreateClient(t.Context(), &repository.CreateClientParams{ Clientid: "clientid", Name: "client_name", }) require.NoError(t, err) docId, err := cfg.GetDBQueries().CreateDocument(t.Context(), &repository.CreateDocumentParams{ Clientid: "clientid", Hash: "hash", }) require.NoError(t, err) part := uint16(1) filetype := "pdf" key := objectstore.BucketKey{ CreatedAt: time.Now().UTC(), ClientID: "clientid", EntityID: uuid.New(), Location: objectstore.Import, Part: &part, FileType: &filetype, } err = cfg.GetDBQueries().AddDocumentEntry(t.Context(), &repository.AddDocumentEntryParams{ Documentid: docId, Bucket: "bucket", Key: key.String(), }) require.NoError(t, err) mimeType := "application/pdf" mockStore.EXPECT(). HeadObject( mock.Anything, mock.MatchedBy(func(in *s3.HeadObjectInput) bool { return *in.Bucket == "bucket" && *in.Key == key.String() }), 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 == key.String() }), mock.Anything, ). Return(&s3.GetObjectOutput{ Body: bodyMsg, }, nil) bod := doccleanrunner.Body{ DocumentID: docId, } assert.True(t, runner.Process(t.Context(), bod)) test.AssertMessageBody(t, cfg, cfg.GetDocumentTextURL(), regexp.MustCompile(fmt.Sprintf("{\"id\":\"%s\"}", docId.String()))) } 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`