I think that should be the case for any methods of the file analysis interface that don’t use a pre-computed file id, but only if the file handles returned from the script layer “get_file_handle” function end up differing between calls to the file analysis interface. So the question I’d try to answer would be “are the file handles returned from my get_file_handle function differing unintentionally?”
Ultimately, moving to the pre-computed file id interface like you did makes sense for this code since you know the two calls to the file analysis interface right next to each other are associated w/ the same file.
- Jon