a cryptographic hash would make it only possible to know what hash to look for to find the number
you are simply going to need more entropy than that, that's the short answer
there might be some potential to use proximity hashes like used in AI neural networks to get a partial match based on the combination of pubkey and phone number, and thin down candidates, but simply, the phone number is too small to be resistant to a white pages index attack
nostr:npub1xtscya34g58tk0z605fvr788k263gsu6cy9x0mhnm87echrgufzsevkk5s https://pypi.org/project/proximityhash/ geohashes are one form of proximity hash scheme, there probably is others
i think if you look closely at this you may find a way whereby you can get a close match that massively thins down from a set of publicly visible hashes that are partially generated from phone numbers, such that they don't facilitate reverse indexing but enable those with the number to find them
there is probably a heap of papers you can find and maybe someone has already posed the question of how to make a forward matching phone number search resistant to reverse matching
it's a problem that is very old, i remember in the old days when there started to be reverse phone number databases, which is the problem you are trying to solve, so probably someone has already got a scheme that is data-type-specific
I was trying to google this, I would be surprised if this hasn’t been thought of before.
Yeah, first thought is to add pubkey.
