Re: gram.y comment issue

hello,

The gram.y includes the line(line:1965):
/* ALTER TABLE <name> ALTER [COLUMN] <colname> SET ( column_parameter =
value [, ... ] ) */
ALTER opt_column ColId RESET reloptions

The comment has the word "SET", but I think that's should be "RESET".

and

src/include/nodes/parsenodes.h line:141
List *distinctClause; /* a list of SortGroupClause's */
The comment has the word "SortGroupClause", but that's should be
"distinctClause".

I apologize if this is too small of an error to be considered a bug, but I
figured that we want to be the best we can.

regards,wensheng zhang

--
----------------------
张文升 | PostgreSQL DBA

On Wed, Mar 9, 2016 at 3:04 PM, 张文升 <wensheng(dot)zhang(at)qunar(dot)com> wrote:
> The gram.y includes the line(line:1965):
> /* ALTER TABLE <name> ALTER [COLUMN] <colname> SET ( column_parameter =
> value [, ... ] ) */
> ALTER opt_column ColId RESET reloptions
>
> The comment has the word "SET", but I think that's should be "RESET".

This one needs to be fixed.

> src/include/nodes/parsenodes.h line:141
> List *distinctClause; /* a list of SortGroupClause's */
> The comment has the word "SortGroupClause", but that's should be
> "distinctClause".

This one is fine. The name distinctClause here is parser's way of
referring to the SQL construct DISTINCT ON (<list-of-items>) in a
SELECT query, where each member of the list is a SortGroupClause item.

Thanks,
Amit

Amit Langote <amitlangote09(at)gmail(dot)com> writes:
> On Wed, Mar 9, 2016 at 3:04 PM, <wensheng(dot)zhang(at)qunar(dot)com> wrote:
>> The gram.y includes the line(line:1965):
>> /* ALTER TABLE <name> ALTER [COLUMN] <colname> SET ( column_parameter =
>> value [, ... ] ) */
>> ALTER opt_column ColId RESET reloptions
>> The comment has the word "SET", but I think that's should be "RESET".

> This one needs to be fixed.

Agreed, done.

>> src/include/nodes/parsenodes.h line:141
>> List *distinctClause; /* a list of SortGroupClause's */
>> The comment has the word "SortGroupClause", but that's should be
>> "distinctClause".

> This one is fine. The name distinctClause here is parser's way of
> referring to the SQL construct DISTINCT ON (<list-of-items>) in a
> SELECT query, where each member of the list is a SortGroupClause item.

Yes. It might be surprising if you don't know that ORDER BY, GROUP BY,
and DISTINCT all have a common parsetree representation ... but the
comment is correct as written.

regards, tom lane

thanks !

On 2016年03月09日 23:31, Tom Lane wrote:
> Amit Langote

--
----------------------
张文升 | PostgreSQL DBA