Files
query-orchestration/api/docTextRunner/runner_test.go
T
Michael McGuinness fee71e7740 Merged in feature/postprocessing (pull request #114)
Feature/postprocessing

* tests

* passtest

* fixshorttests

* mosttests

* improvingbasedockerfile

* testspeeds

* testing

* host

* canparallel

* clean

* passfullsuite

* singlepagemax

* test

* findfeatures

* findstables

* tbls

* tablestoo

* tablestoo

* lateraltests

* tableloc

* cleanup

* inlinetable

* childids

* cleanup

* tests
2025-04-22 14:40:16 +00:00

211 lines
4.9 KiB
Go

package doctextrunner_test
import (
"context"
"encoding/json"
"fmt"
"io"
"os"
"strings"
"testing"
"time"
doctextrunner "queryorchestration/api/docTextRunner"
"queryorchestration/internal/database/repository"
documenttext "queryorchestration/internal/document/text"
"queryorchestration/internal/server/runner"
"queryorchestration/internal/serviceconfig/aws"
"queryorchestration/internal/serviceconfig/objectstore"
"queryorchestration/internal/serviceconfig/queue/querysync"
"queryorchestration/internal/serviceconfig/textract"
objectstoremock "queryorchestration/mocks/objectstore"
queuemock "queryorchestration/mocks/queue"
textractmock "queryorchestration/mocks/textract"
"github.com/aws/aws-sdk-go-v2/service/s3"
"github.com/aws/aws-sdk-go-v2/service/sqs"
awstextract "github.com/aws/aws-sdk-go-v2/service/textract"
"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 DocTextConfig struct {
runner.BaseConfig[doctextrunner.Body]
aws.AWSConfig
querysync.QuerySyncConfig
textract.TextractConfig
objectstore.ObjectStoreConfig
}
func TestDocCleanRunner(t *testing.T) {
ctx := context.Background()
pool, err := pgxmock.NewPool()
require.NoError(t, err)
cfg := &DocTextConfig{}
cfg.DBPool = pool
cfg.DBQueries = repository.New(pool)
mockStore := objectstoremock.NewMockS3Client(t)
cfg.StoreClient = mockStore
mockTextract := textractmock.NewMockTextractClient(t)
cfg.TextractClient = mockTextract
mockQueue := queuemock.NewMockSQSClient(t)
cfg.QueueClient = mockQueue
cfg.QuerySyncURL = "ex"
runner := doctextrunner.New(&doctextrunner.Services{
Text: documenttext.New(cfg),
})
assert.NotNil(t, runner)
doc := doctextrunner.Body{
DocumentID: uuid.New(),
}
cleanId := uuid.New()
bucket := "bucket"
hash := `"a06d5e0e795adeb6d0b3732330dbcc15"`
textId := uuid.New()
key := objectstore.BucketKey{
ClientID: "aa",
Location: objectstore.Import,
CreatedAt: time.Now().UTC(),
EntityID: cleanId,
}
keyStr := key.String()
filename := "../../assets/sampleGeneration/generated/helloWorld.gen"
file, err := os.Open(filename)
require.NoError(t, err)
defer file.Close()
decoder := json.NewDecoder(file)
var textractBaseOut map[string][]awstextract.AnalyzeDocumentOutput
err = decoder.Decode(&textractBaseOut)
require.NoError(t, err)
mimetype := repository.NullCleanmimetype{
Valid: true,
Cleanmimetype: repository.CleanmimetypeApplicationPdf,
}
pool.ExpectQuery("name: IsDocumentTextExtracted :one").WithArgs(doc.DocumentID).WillReturnRows(
pgxmock.NewRows([]string{"isextracted"}).
AddRow(false),
)
pool.ExpectQuery("name: GetCleanEntryByDocId :one").WithArgs(doc.DocumentID).
WillReturnRows(
pgxmock.NewRows([]string{"id", "documentId", "bucket", "key", "version", "hash", "mimetype", "fail", "clientId"}).
AddRow(cleanId, doc.DocumentID, &bucket, &keyStr, int64(123), &hash, mimetype, nil, key.ClientID),
)
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)
sent := 0
mockTextract.EXPECT().
AnalyzeDocument(
mock.Anything,
mock.MatchedBy(func(in *awstextract.AnalyzeDocumentInput) bool {
if sent == 0 {
sent++
return true
}
return false
}),
mock.Anything,
).
Return(&textractBaseOut["base"][0], nil).Once()
pool.ExpectBegin()
pool.ExpectQuery("name: GetDocumentTextExtractionByHash :one").WithArgs(cleanId, hash).WillReturnRows(
pgxmock.NewRows([]string{"id"}).
AddRow(textId),
)
pool.ExpectExec("name: AddDocumentTextEntry :exec").WithArgs(textId, pgxmock.AnyArg()).
WillReturnResult(pgxmock.NewResult("", 1))
pool.ExpectCommit()
mockQueue.EXPECT().
SendMessage(
mock.Anything,
mock.MatchedBy(func(in *sqs.SendMessageInput) bool {
return *in.QueueUrl == cfg.QuerySyncURL && *in.MessageBody == fmt.Sprintf("{\"id\":\"%s\"}", doc.DocumentID.String())
}),
mock.Anything,
).
Return(&sqs.SendMessageOutput{}, nil)
assert.True(t, runner.Process(ctx, doc))
}
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`