Re: CREATE TABLE ( .. STORAGE ..)

2022-07-13 Thread Aleksander Alekseev
Hi, > Committed. > > I thought the removal of the documentation details of SET COMPRESSION > and SET STORAGE from the ALTER TABLE ref page was a bit excessive, since > that material actually contained useful information about what happens > when you change compression or storage on a table with ex

Re: CREATE TABLE ( .. STORAGE ..)

2022-07-13 Thread Peter Eisentraut
On 12.07.22 12:10, Aleksander Alekseev wrote: Hi Peter, The "safety check: do not allow toasted storage modes unless column datatype is TOAST-aware" could be moved into GetAttributeStorage(), so it doesn't have to be repeated. (Note that GetAttributeCompression() does similar checking.) Good

Re: CREATE TABLE ( .. STORAGE ..)

2022-07-12 Thread Aleksander Alekseev
Hi Peter, > The "safety check: do not allow toasted storage modes unless column > datatype is TOAST-aware" could be moved into GetAttributeStorage(), so > it doesn't have to be repeated. (Note that GetAttributeCompression() > does similar checking.) Good point. Fixed. > ATExecSetStorage() curre

Re: CREATE TABLE ( .. STORAGE ..)

2022-07-11 Thread Peter Eisentraut
On 11.07.22 11:27, Aleksander Alekseev wrote: Here is a patch updated according to all the recent feedback, except for two suggestions: In v4 I forgot to list possible arguments for STORAGE in alter_table.sgml, similarly as it is done for other subcommands. Here is a corrected patch. Here is

Re: CREATE TABLE ( .. STORAGE ..)

2022-07-11 Thread Aleksander Alekseev
Hi hackers, > > Here is a patch updated according to all the recent feedback, except > > for two suggestions: > > In v4 I forgot to list possible arguments for STORAGE in > alter_table.sgml, similarly as it is done for other subcommands. Here > is a corrected patch. Here is the rebased patch. --

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-24 Thread Aleksander Alekseev
Hi hackers, > Here is a patch updated according to all the recent feedback, except > for two suggestions: In v4 I forgot to list possible arguments for STORAGE in alter_table.sgml, similarly as it is done for other subcommands. Here is a corrected patch. - SET STORAGE + SET STORAGE { PLAIN | EXT

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-24 Thread Aleksander Alekseev
Hi hackers, Many thanks for the review! Here is a patch updated according to all the recent feedback, except for two suggestions: > This adds support for "ADD COLUMN SET STORAGE" but it is not described > in the doc. COMPRESSION is not described, too. Shouldn't we add the > both this time? Or

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-22 Thread Peter Eisentraut
On 29.03.22 22:28, Matthias van de Meent wrote: As I'm new to the grammar files; would you know the difference between `name` and `ColId`, and why you would change from one to the other in ALTER COLUMN STORAGE? The grammar says name: ColId { $$ = $1; };

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-22 Thread Peter Eisentraut
On 15.06.22 16:51, Aleksander Alekseev wrote: I noticed that cfbot is not entirely happy with the patch, so I rebased it. I see that COMPRESSION and STORAGE now are handled slightly differently in the grammar. Maybe we could standardize that a bit more; so that we have only one `STORAGE [kind]`

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-16 Thread Kyotaro Horiguchi
Thanks! I have been annoyed sometimes by the lack of this feature. At Thu, 16 Jun 2022 16:40:55 +0300, Aleksander Alekseev wrote in > Hi Matthias, > > > Apart from this comment on the format of the patch, the result seems solid. > > Many thanks. > > > When updating a patchset generally we tr

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-16 Thread Aleksander Alekseev
Hi Matthias, > Apart from this comment on the format of the patch, the result seems solid. Many thanks. > When updating a patchset generally we try to keep the patches > self-contained, and update patches as opposed to adding incremental > patches to the set. My reasoning was to separate my cha

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-16 Thread Matthias van de Meent
On Wed, 15 Jun 2022 at 16:51, Aleksander Alekseev wrote: > > Hi hackers, > > I noticed that cfbot is not entirely happy with the patch, so I rebased it. > > > I see that COMPRESSION and STORAGE now are handled slightly > > differently in the grammar. Maybe we could standardize that a bit > > more;

Re: CREATE TABLE ( .. STORAGE ..)

2022-06-15 Thread Aleksander Alekseev
Hi hackers, I noticed that cfbot is not entirely happy with the patch, so I rebased it. > I see that COMPRESSION and STORAGE now are handled slightly > differently in the grammar. Maybe we could standardize that a bit > more; so that we have only one `STORAGE [kind]` definition in the > grammar?

Re: CREATE TABLE ( .. STORAGE ..)

2022-03-29 Thread Matthias van de Meent
On Wed, 2 Feb 2022 at 11:13, Teodor Sigaev wrote: > > Hi! > > > Are they both set to name or ColId? Although they are the same. > > > > Thank you, fixed, that was just an oversight. > > > 2 For ColumnDef new member storage_name, did you miss the function > > _copyColumnDef() _equalColumnDef()? >

Re: CREATE TABLE ( .. STORAGE ..)

2022-02-02 Thread Teodor Sigaev
Hi! Are they both set to name or ColId? Although they are the same. Thank you, fixed, that was just an oversight. 2 For ColumnDef new member storage_name, did you miss the function _copyColumnDef() _equalColumnDef()? Thank you, fixed Regards Wenjing 2021年12月27日 15:51，Teodor Sigae

Re: CREATE TABLE ( .. STORAGE ..)

2022-01-20 Thread wenjing zeng
HI For patch create_table_storage-v1 1 +ALTER opt_column ColId SET STORAGE name +opt_column_storage: + STORAGE ColId { $$ = $2; } Are they both set to name or ColId? Although they are the same. 2 For ColumnDef new member

CREATE TABLE ( .. STORAGE ..)

2021-12-26 Thread Teodor Sigaev
Hi! Working on pluggable toaster (mostly, for JSONB improvements, see links below) I had found that STORAGE attribute on column is impossible to set in CREATE TABLE command but COMPRESS option is possible. It looks unreasonable. Suggested patch implements this possibility. [1] http://www.sa