[Feature]: Add sql-datatype to the SDK discovery and catalog.

[BuzzCutNorman]

Feature scope

Taps (catalog, state, stream maps, etc.)

Description

It would be very helpful if SDK based taps during the discovery process of a database grabbed column data types and placed them into the catalog metadata area for the column in the field of sql-datatype. The sql-datatype field is a reserved keyword mentioned in the singer discovery documentation.

https://github.com/singer-io/getting-started/blob/master/docs/DISCOVERY_MODE.md#metadata

Keyword	Tap Type	Discoverable?	Description
sql-datatype	database	discoverable	Represents the datatype of the database column

        {
          "breadcrumb": [
            "properties",
            "name"
          ],
          "metadata": {
            "sql-datatype": "varchar(255)",
            "selected-by-default": true
          }
        }

[stale]

This has been marked as stale because it is unassigned, and has not had recent activity. It will be closed after 21 days if no further activity occurs. If this should never go stale, please add the evergreen label, or request that it be added.

[edgarrmondragon]

Related: #1903

[edgarrmondragon]

Copying here my comment from #1872 (comment):

One thing that unfortunately isn't very clear to me is how are targets supposed to consume this information since a target isn't aware of the tap's metadata. It might be better if this metadata lived in the schema, at least that way it would be emitted with SCHEMA messages and could be used by the target.

The official Singer docs only explain the field as Represents the datatype of a database column, and pipelinewise's taps implement but the target don't seem to use it in any way:

https://github.com/search?q=org%3Atransferwise+sql-datatype+language%3APython&type=code&l=Python

Otherwise, who is the consumer of this metadata?

[BuzzCutNorman]

I am guessing the consumer might be the Stich platform. During this conversation I had in the Meltano Slack:

https://discuss.meltano.com/t/16288098/i-was-wondering-if-i-someone-could-point-me-in-the-proper-di

We came to the same conclusion that this would need to be a part of the SCHEMA message to become useful. The alternative of using JSON Shema attributes to describe enough of the specifics to determine what the types are was mentioned. This has been done in countless other taps and targets just there is not a standardization of it.

knowing the database type (posgresql, sql server, oracle) and sql data type ( integer(), varchar(10) ) would be easier and maybe more accurate than the JSON Schema fingerprint method.

I find myself looking back on this and almost thinking this feature request should be closed. Please let me know wdyt?