Remove mailing names and address associated with suppresion requests from address view

[wrridgeway]

This PR creates a new table ccao.hidename which is fed by a list of PINs stored on the O drive. Any PIN from this list will not have any data from iasworld.maildat surfaced in default.vw_pin_address.

select mail_address_name is null as "mailing address is null",
	count() as rows,
	'new' as source
from z_ci_hidenames_default.vw_pin_address
group by mail_address_name is null
union all
select mail_address_name is null as "mailing address is null",
	count() as rows,
	'old' as source
from default.vw_pin_address
group by mail_address_name is null

mailing_address_is_null	rows	source
false	48003634	old
true	2667468	old
true	2668055	new
false	48003047	new

the difference in null rows is 587, which is exactly what we expect:

select count()
from default.vw_pin_address vpa
	inner join ccao.hidename hide on vpa.pin = hide.pin

gives us 587.

with pins as (
	select pin
	from ccao.hidename
)
select vpa.*
from z_ci_hidenames_default.vw_pin_address as vpa
	inner join pins on vpa.pin = pins.pin
where mail_address_name is not null

returns no rows.

[jeancochrane]

Thanks for picking this up so quickly! I have one question about high-level behavior, but otherwise this looks fine to me.

[jeancochrane]

[Question, non-blocking] Just confirming that this behavior will remove the entire row for every year for every PIN in the source list. Is that the behavior we want @ccao-jardine? Or do we want to preserve the row and replace the non-PIN/year data with indicator values, or nulls?

[wrridgeway]

The PIN will still have a row in the view, just NULLs for columns from maildat. Could totally change the NULLs to a filler value.

[jeancochrane]

Oops, that's right, my bad! The current behavior seems like the best option to me, then.

[ccao-jardine]

Thanks for this bandaid and thorough review! Confirming that I do want the PIN to have a row, replacing the contents of the MAILDAT columns with nulls.