Re: Support logical replication of DDLs

On Mon, Mar 13, 2023 at 2:24 AM Zheng Li <zhengli10@gmail.com> wrote:
>
> Thanks for working on the test coverage for CREATE and ALTER TABLE.
> I've made fixes for some of the failures in the v79 patch set (0002,
> 0003 and 0004 are updated). The changes includes:
> 1. Fixed a syntax error caused by ON COMMIT clause placement in
> deparse_CreateStmt.
> 2. Fixed deparse_Seq_As and start using it in deparse_CreateSeqStmt,
> this issue is also reported in [1].
> 3. Fixed a bug in append_not_present: the 'present: false' element
> can't be omitted even in non-verbose mode. It will cause syntax error
> on reformed command if 'present: false' element is missing but the fmt
> string indicates the corresponding object must be present.
> 4. Replaced if_not_exists with if_exists in deparse of
> AT_DropConstraint and AT_DropColumn.
> 5. Added missing CASCADE clause for AT_DropConstraint deparse.
> 6. Enabled the fixed test cases.
>

I found out that the option ONLY was not parsed in the "CREATE INDEX" command,
for eg: CREATE UNIQUE INDEX ... ON ONLY table_name ...

I've fixed this in patch 0002.

regards,
Ajin Cherian
Fujitsu Australia