ANY and ALL contains their operators

[SeanTroyUWO]

Changes Expr::AllOp and Expr::AnyOp to contain the operator which precedes them, rather than be contained by them. This fixes an issue I had as a polars user: pola-rs/polars#10081.

The rational for this is that despite the order the syntax demands, the conditional operator modifies the behavior of the ANY command, and not vice versa. Similar to NOT as with NOT IN, the negation modifies the behavior of the IN, and the IN is the real operation. In both cases, its more convenient to parse the real operation than the modifier(s).

I did not find any uses the the old ANY or ALL commands in any of DataFusion, LocustDB, Ballista, GlueSQL or Opteryx. This might still briefly break their builds however.

If more unit tests are required, please let me know.

Fixes: #936

[alamb]

Thank you very much for the contribution @SeanTroyUWO -- I apologize for the delay in reviewing.

This change also appears to be consistent with postgres any mysql (where the ANY/ALL is part of the comparison expression):

https://www.postgresql.org/docs/current/functions-comparisons.html

9.24.3. ANY/SOME (array)
expression operator ANY (array expression)
expression operator SOME (array expression)

https://dev.mysql.com/doc/refman/8.0/en/any-in-some-subqueries.html

The ANY keyword, which must follow a comparison operator, ...

Another alternative might be to add it directly to Expr::BinaryOp but I think that would be far more disruptive to downstream users so this looks good to me

[SeanTroyUWO]

Thank you very much for the contribution @SeanTroyUWO -- I apologize for the delay in reviewing.

This change also appears to be consistent with postgres any mysql (where the ANY/ALL is part of the comparison expression):

https://www.postgresql.org/docs/current/functions-comparisons.html

9.24.3. ANY/SOME (array)
expression operator ANY (array expression)
expression operator SOME (array expression)

https://dev.mysql.com/doc/refman/8.0/en/any-in-some-subqueries.html

The ANY keyword, which must follow a comparison operator, ...

Another alternative might be to add it directly to Expr::BinaryOp but I think that would be far more disruptive to downstream users so this looks good to me

Great, I'll look forward to the next release!

[alamb]

Next release is tracked by #955

ETA in about 2 weeks