Fixing ODT document extraction (#17700)

services/docextractor/docextractor_test.go

@@ -105,6 +105,14 @@ func TestExtract(t *testing.T) {
 			[]string{},
 			false,
 		},
+		{
+			"Odt file",
+			"sample-doc.odt",
+			ExtractSettings{},
+			[]string{"simple", "document", "contains"},
+			[]string{},
+			false,
+		},
 		{
 			"Pptx file",
 			"sample-doc.pptx",

services/docextractor/documents.go

@@ -5,10 +5,7 @@ package docextractor
 
 import (
 	"errors"
-	"fmt"
 	"io"
-	"io/ioutil"
-	"os"
 	"path"
 	"strings"
 
@@ -41,19 +38,7 @@ func (de *documentExtractor) Extract(filename string, r io.ReadSeeker) (string,
 		return "", errors.New("unknown converter")
 	}
 
-	f, err := ioutil.TempFile(os.TempDir(), "docconv")
+	text, _, err := converter(r)
-	if err != nil {
-		return "", fmt.Errorf("error creating temporary file: %v", err)
-	}
-	defer f.Close()
-	defer os.Remove(f.Name())
-
-	_, err = io.Copy(f, r)
-	if err != nil {
-		return "", fmt.Errorf("error copying data into temporary file: %v", err)
-	}
-
-	text, _, err := converter(f)
 	if err != nil {
 		return "", err
 	}