Merged in feature/clean (pull request #76)

Clean Set Up

* fail

* starteddb

* constraint

* cleantest

* fixqueries

* fixtests

* storemimetype

* buffer

* tests

* setup

* passingtests

* pdfHElloWorld

* rm

* test

* notodos

* tests

* clean

* testpdf
This commit is contained in:
Michael McGuinness
2025-02-28 13:11:53 +00:00
parent a49084ebce
commit f11f4def43
198 changed files with 71722 additions and 146 deletions
+112 -6
View File
@@ -4,6 +4,7 @@ import (
"context"
"encoding/json"
"fmt"
"io"
doccleanrunner "queryorchestration/api/docCleanRunner"
"queryorchestration/internal/database"
"queryorchestration/internal/database/repository"
@@ -15,8 +16,10 @@ import (
"queryorchestration/internal/serviceconfig/queue/documenttext"
objectstoremock "queryorchestration/mocks/objectstore"
queuemock "queryorchestration/mocks/queue"
"strings"
"testing"
"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"
@@ -56,30 +59,46 @@ func TestDocCleanRunner(t *testing.T) {
})
assert.NotNil(t, runner)
doc := doccleanrunner.Body{
bod := doccleanrunner.Body{
ID: uuid.New(),
}
bodyBytes, err := json.Marshal(doc)
bodyBytes, err := json.Marshal(bod)
assert.NoError(t, err)
body := string(bodyBytes)
msg := &types.Message{
Body: &body,
}
doc := document.Document{
ID: bod.ID,
JobID: uuid.New(),
Hash: "example_hash",
}
inloc := document.Location{
Bucket: "bucket_name",
Key: "/i/am/here",
}
dbid := database.MustToDBUUID(doc.ID)
pool.ExpectQuery("name: IsDocumentClean :one").WithArgs(database.MustToDBUUID(doc.ID)).WillReturnRows(
pool.ExpectQuery("name: HasDocumentCleanEntry :one").WithArgs(dbid).WillReturnRows(
pgxmock.NewRows([]string{"isclean"}).
AddRow(false),
)
pool.ExpectQuery("name: GetDocumentEntry :one").WithArgs(database.MustToDBUUID(doc.ID)).WillReturnRows(
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(database.MustToDBUUID(doc.ID), inloc.Bucket, inloc.Key),
AddRow(dbid, inloc.Bucket, inloc.Key),
)
pool.ExpectExec("name: AddDocumentCleanEntry :exec").WithArgs(database.MustToDBUUID(doc.ID), int32(1), 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))
mockSQS.EXPECT().
@@ -92,6 +111,93 @@ func TestDocCleanRunner(t *testing.T) {
).
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)
err = runner.Process(ctx, msg)
assert.NoError(t, err)
}
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`