Add maxLength to discovered schema

[techtangents]

Add the JSON Schema "maxLength" property when encountering varchars with a max size. e.g. a varchar(256) would have "maxLength": 256 in the schema.

This would give information to the target that's useful when creating fields. In the above example, a Postgresql or Redshift target could create a varchar(256) field.

This is particularly important in Redshift. Redshift docs advise to use the smallest possible column size for data (see here and here).

For my situation, it appears that this is causing excessive disk usage in my Redshift warehouse.

The behaviour to emit maxLength values could be a config setting. I'd probably keep it off by default.

Note: corresponding target-redshift issue here: TicketSwap/target-redshift#105

[techtangents]

First raised at https://meltano.slack.com/archives/C069CQNHDNF/p1722991360991679

[edgarrmondragon]

This should be supported once ship meltano/sdk#2618.

[edgarrmondragon]

@techtangents this should've been addressed by

1. feat: Bump to singer-sdk v0.41.0 #503
2. (upstream) feat(taps): SQL taps now emit schemas with maxLength when applicable meltano/sdk#2651

Those changes were shipped with v0.0.15. Let me know if you find any problems with it.

Thanks!