Files
query-orchestration/api/docCleanRunner/runner_test.go
T
Michael McGuinness bbd86fb4ea Merged in feature/CollectorBuildVersion (pull request #100)
Collector Build Version

* lint

* fixtests
2025-03-11 18:15:49 +00:00

224 lines
5.2 KiB
Go

package doccleanrunner_test
import (
"context"
"encoding/json"
"fmt"
"io"
"strings"
"testing"
doccleanrunner "queryorchestration/api/docCleanRunner"
"queryorchestration/internal/database"
"queryorchestration/internal/database/repository"
"queryorchestration/internal/document"
documentclean "queryorchestration/internal/document/clean"
"queryorchestration/internal/serviceconfig"
"queryorchestration/internal/serviceconfig/objectstore"
"queryorchestration/internal/serviceconfig/queue/documenttext"
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/aws/aws-sdk-go-v2/service/sqs/types"
"github.com/go-playground/validator/v10"
"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 {
serviceconfig.BaseConfig
documenttext.DocTextConfig
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.DocumentTextURL = "/i/am/here"
runner := doccleanrunner.New(validator.New(), &doccleanrunner.Services{
Clean: documentclean.New(cfg),
})
assert.NotNil(t, runner)
t.Run("valid", func(t *testing.T) {
bod := doccleanrunner.Body{
ID: uuid.New(),
}
bodyBytes, err := json.Marshal(bod)
assert.NoError(t, err)
body := string(bodyBytes)
msg := &types.Message{
Body: &body,
}
doc := document.Document{
ID: bod.ID,
ClientID: uuid.New(),
Hash: "example_hash",
}
inloc := document.Location{
Bucket: "bucket_name",
Key: "/i/am/here",
}
dbid := database.MustToDBUUID(doc.ID)
pool.ExpectQuery("name: HasDocumentCleanEntry :one").WithArgs(dbid).WillReturnRows(
pgxmock.NewRows([]string{"isclean"}).
AddRow(false),
)
pool.ExpectQuery("name: GetDocument :one").WithArgs(dbid).
WillReturnRows(
pgxmock.NewRows([]string{"id", "clientId", "hash"}).
AddRow(dbid, database.MustToDBUUID(doc.ClientID), 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.NullCleanmimetype{
Valid: true,
Cleanmimetype: repository.Cleanmimetype(mimeType),
}
pool.ExpectBegin()
cleanid := database.MustToDBUUID(uuid.New())
pool.ExpectQuery("name: GetMostRecentDocumentCleanEntry :one").WithArgs(dbid).
WillReturnRows(
pgxmock.NewRows([]string{"id", "documentId", "bucket", "key", "version", "mimetype", "fail"}),
)
pool.ExpectQuery("name: AddDocumentClean :one").WithArgs(dbid, &inloc.Bucket, &inloc.Key, 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.DocumentTextURL && *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 == inloc.Bucket && *in.Key == inloc.Key
}),
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 == inloc.Bucket && *in.Key == inloc.Key
}),
mock.Anything,
).
Return(&s3.GetObjectOutput{
Body: bodyMsg,
}, nil)
assert.True(t, runner.Process(ctx, msg))
})
t.Run("invalid body", func(t *testing.T) {
body := "definitely invalid"
msgId := "id"
msg := &types.Message{
Body: &body,
MessageId: &msgId,
}
assert.True(t, runner.Process(ctx, msg))
})
}
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`