provide hash function for the go/python/java/c++ client

[whitewum]

Is your feature request related to a problem? Please describe.

Ther is no explanation about which hash function used in console. I've to look up the Nebula C++ code, and guess for my language.

> yield hash("mike")

So when I use golang client to insert vertex id(vid), I don't know which hash function to use (I think in golang, it is murmurhash2A)

Describe the solution you'd like

I hope Nebula can provide a nebula::util::hash

Describe alternatives you've considered

You can tell user the hash function (murmurhash, m, r, seed... ), and let user search for their own language packages. But that's a really really bad idea. It takes me half hour to check my golang hash is consistent with nebula hash ( I don't think anyone else wants to do this search).

Additional context

Add any other context or screenshots about the feature request here.

[whitewum]

I hope you can provide it in the next release. It's simple and useful.

[rexdouglass]

I second this. I'm currently just looking for documentation on which hash algorithm Nebula uses for IDs.

I'm doing my data prep in spark which now has the 64-bit xxhash function. I was thinking I could keep everything in sync if I generated IDs in spark, imported them through the Nebula spark importer, and then later refered to them as plain text on the fly as hash('original plain text name')

So for example, I'm working with RDF data, I want to create a vertex for each URI, and I want the option of easily refering to them as either their 64 bit int or original plain text names.

[wey-gu]

This issue/feature request is valid and should be created in the corresponding repos. :) @jamieliu1023

[CPWstatic]

We had discussed this early this year, we would not support this feature.

[whitewum]

https://docs.nebula-graph.com.cn/2.5.0/3.ngql-guide/6.functions-and-expressions/12.hash/
https://docs.nebula-graph.com.cn/2.5.0/1.introduction/3.nebula-graph-architecture/4.storage-service/#_6
How time flies....