data-discovery

Currently we only support db store publisher (e.g neo4j, mysql,neptune). But it would be pretty easy to support message queue publisher using the interface (e.g SQS, kinesis, Eventhub, kafka) which allows push ETL model support.

There is a pr (amundsen-io/amundsendatabuilder#431) which unfortunately isn't get merged. The pr could be used as an example on how to support t

@cantzakas created the SQL query necessary to pull metadata in (hyperqueryhq/whale#140) -- we just have to make the greenplum extractor scaffolding. This should just follow the exact same shape as the Postgres extractor.

Specification must have an order parameter for fields. ODD Platform must take it into account after it's implemented on the ODD Spec side

Promoted from related-sciences/articat#5

More information in the parent issue

Collector generates invalid oddrn for fields that have a parent field. It's doesn't take into accounts those parents' fields oddrns which causes merging issues.

For example, a valid case for BigQuery would be:

{
    field: {
        name: "..."
    },
    another_field: {
        name: "..."
    }
}

which causes collector to generate same oddrns for name fields

Depends on #5