We should be careful about backporting this one.

1. It's not a bug as such, but rather an improvement.
2. It's the first time in more than two decades anyone has ever asked for this.

I suppose it's pretty low risk, but it's even lower risk to ask anyone who wants this to update to seven.

This used to work in 6.5, which is why see this as a bug, or at least a new side effect of some recent changes.

Are you saying it worked before HHH-18215?

Yes, i think that was the reason

I guess this commit introduced this bug: 6435137

I guess this commit introduced this bug

Exactly as I suggested above:

Are you saying it worked before HHH-18215?

But, and yes, I'm going to keep hammering on this forever: it's not a "bug" if we never documented it anywhere and never had any tests for it.

The spec requires us to support SQL. It doesn't say anything about supporting SQL with weird JDBC escapes. The only place the spec mentions JDBC escapes is in this line:

The JDBC escape syntax may be used for the specification of date, time, and timestamp literals.

And that is referring to JPQL, not to SQL.

So if this worked, it worked by accident.

That said, I guess it's OK to backport this, as long as we're sure it doesn't break something else.

Good enough for me :)

It seems that I don't have the rights to merge this:

Merging is blocked
You're not authorized to push to this branch.

@beikov could you merge this please?