Warn on reusing common subqueries across multiple QueryBuilders

[OneRaynyDay]

For beginners and library writers who work atop pypika, there is a common pattern of reusing Pypika objects for queries:

subquery = Query.from_(table).select(table.col1.as_("col"), table.col2)
query1 = Query.from_(subquery).select(Avg(subquery.col)
...
query2 = Query.from_(subquery).select(Sum(subquery.col)

The nested subqueries currently fail this task because the alias for query is changed in-place when it's Query.from_()'d. In general, these unintended, subtle side-effects should at least have a warning for the users. In fact, even more subtle scenarios emerge when trying to join multiple tables which have been assigned an alias on accident and resulted in ambiguous names:

 > SELECT "sq0"."a","sq0"."c"
    -> FROM
    ->   (SELECT "a","b"
    ->    FROM
    ->      (SELECT *
    ->       FROM (
    ->             VALUES (1,3), (2,4), (3,5)) AS "yy14" ("a","b"))) "sq0"
    -> JOIN
    ->   (SELECT "c","b"
    ->    FROM
    ->      (SELECT *
    ->       FROM (
    ->             VALUES (5,3), (6,5), (7,6)) AS "7a4d" ("c","b"))) "sq0" ON "sq0"."b"="sq0"."b";
...
Column 'sq0.b' is ambiguous

As a result, I propose we add a warning to any user who is encountering this subtle caveat from pypika:

ReusedNestedQueryWarning: Query SELECT "name","customer" FROM "a" is used as a nested subquery elsewhere. Consider deepcopying the query to avoid unexpected aliasing.
    warn(f"Query {selectable} is used as a nested subquery elsewhere. "

I found that solving this problem but providing backwards compatibility is impractical, since we'd need to deepcopy every subquery that comes in to prevent mutating state, but also understand that Terms deriving from subqueries belong to the deepcopied version of its table parent. The surface area for Terms is large enough that I think having a warning would be a good first step if any for this issue.