Improve document extraction and including a document extraction command (#17183)

* Add extract documents content command

* Adding the extraction command and making the pure go pdf library as secondary option

* Improving the memory usage and docextractor interface

* Enable content extraction by default in all the instances

* Tiny improvement on archive indexing

* Adding App interface generation and the opentracing layer

* Fixing linter errors

* Addressing PR review comments

* Addressing PR review comments

app/upload.go

@@ -14,7 +14,6 @@ import (
 
 	"github.com/mattermost/mattermost-server/v5/model"
 	"github.com/mattermost/mattermost-server/v5/plugin"
-	"github.com/mattermost/mattermost-server/v5/services/docextractor"
 	"github.com/mattermost/mattermost-server/v5/shared/mlog"
 	"github.com/mattermost/mattermost-server/v5/store"
 )
@@ -303,15 +302,9 @@ func (a *App) UploadData(us *model.UploadSession, rd io.Reader) (*model.FileInfo
 	if *a.Config().FileSettings.ExtractContent && a.Config().FeatureFlags.FilesSearch {
 		infoCopy := *info
 		a.Srv().Go(func() {
-			text, err := docextractor.Extract(infoCopy.Name, file, docextractor.ExtractSettings{
-				ArchiveRecursion: *a.Config().FileSettings.ArchiveRecursion,
-			})
+			err := a.ExtractContentFromFileInfo(&infoCopy)
 			if err != nil {
-				mlog.Error("Failed to extract file content", mlog.Err(err))
-				return
-			}
-			if storeErr := a.Srv().Store.FileInfo().SetContent(infoCopy.Id, text); storeErr != nil {
-				mlog.Error("Failed to save the extracted file content", mlog.Err(storeErr))
+				mlog.Error("Failed to extract file content", mlog.Err(err), mlog.String("fileInfoId", infoCopy.Id))
 			}
 		})
 	}