You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'd like to implement the hash function for BigQuery (specifically, I'd like to use the BigQuery SHA256 hash function), but I see a couple of questions that need to be answered before implementing.
Currently, hash is a ValueExpr, but in BigQuery only BYTES / STRING are acceptable as inputs. Should we implement it in BigQuery as a string expression instead?
Currently, hash outputs int64, but BigQuery outputs BYTES (which makes sense since SHA256 generates a 256-bit hash). Would it make sense to change the impala return type to match? Or is it okay to be inconsistent?
Note: other databases tend to align with how BigQuery handles hashes. See:
This will allow the SHA256 function to run and should be flexible
enough to support other hash functions and backends. `hashbytes` was
chosen as the function name because `hash` conflicts with the existing
function that has a different output type. See:
ibis-project/ibis#2310
I'd like to implement the hash function for BigQuery (specifically, I'd like to use the BigQuery SHA256 hash function), but I see a couple of questions that need to be answered before implementing.
Note: other databases tend to align with how BigQuery handles hashes. See:
The text was updated successfully, but these errors were encountered: