LFC

ping @iwinux , any updates?

It's ok, I'll handle it.

@qstommyshu feel free to submit PRs to this issue. You can do it progressively, modify part of the codes at a time.

@qstommyshu `table` crate should be logically lower level than `meta-client`. Let's reuse the `TableReference` struct we already have, and forget about `TableFullName`. All in all, we just want to unify...

@SSebo I think "prepared" stmt should be handled here: https://github.com/GreptimeTeam/greptimedb/blob/57979c9d3dba13beef5d63b4dc73fa1b3281e7e6/src/servers/src/mysql/handler.rs#L111 ane then "execute" here: https://github.com/GreptimeTeam/greptimedb/blob/57979c9d3dba13beef5d63b4dc73fa1b3281e7e6/src/servers/src/mysql/handler.rs#L123

@SSebo yes, part of it. The session crate was introduced in [this PR](https://github.com/GreptimeTeam/greptimedb/pull/672) already, I think prepared stmt can be stored in the `QueryContext` now.

@SSebo It seems you have to use [COM_STMT_PREPARE](https://dev.mysql.com/doc/dev/mysql-server/latest/page_protocol_com_stmt_prepare.html) to trigger the callback of `on_prepare`. It is not text protocol.

@SSebo prepare stmt in mysql could be executed in mysql-cli like [this](https://dev.mysql.com/doc/refman/8.0/en/sql-prepared-statements.html), and this feature request is tracked in #769 . I think in this issue, we can focus on...

@SSebo Yes, it's not COM_STMT_PREPARE. It's not using binary protocol, in the mysql command cli. It's the other form of prepare stmt.

ping @SSebo , any updates?