hi, unfortunately there is no other way. this is a known problem with re-indexing. I presume only in the database version of logseq (which is still not finished and released) will this be solved.
Is there any more info, how this database version would look like? This direction sounds very useful with regards to performance and data consistency like indexing.
I would be in favor of making the database “single source of truth”, storing every content in it. To accommodate for the need of having Markdown files on the filesystem and reflect current behavior, we could create a virtual filesystem mapper, which maps current page content in database as files for external processing.