Re: Patch for column properties
От | Guillaume Lelarge |
---|---|
Тема | Re: Patch for column properties |
Дата | |
Msg-id | 45EAA045.6020609@lelarge.info обсуждение исходный текст |
Ответ на | Re: Patch for column properties ("Florian G. Pflug" <fgp@phlo.org>) |
Ответы |
Re: Patch for column properties
|
Список | pgadmin-hackers |
Florian G. Pflug a écrit : > Dave Page wrote: >> Guillaume Lelarge wrote: >>> Hi, >>> >>> When someone set a default value for a column and set NOT NULL at the >>> same time, the query could fail if the table already contains NULL >>> values for this column. This patch adds an UPDATE statement to put >>> the DEFAULT value in this column for all NULL values (but only when >>> you SET NULL and SET DEFAULT the column at the same time). >>> >>> I've done this patch for a friend but I don't quite like this patch >>> because of its behaviour. It doesn't warn the user. I think it would >>> be better to show a dialog telling that an UPDATE statement will be >>> executed because otherwise the whole action could fail. >>> >>> So here are some questions : >>> * do you think this patch is worth it ? >> >> I'm not keen on the idea.I prefer to leave that sort of behaviour to >> the server, *if* thats the way it gets written. > > You could first execute "alter table .. set default ...; alter table > ..set not null" first, and only if the "set not null" fails ask the user > if he wants > the update to be performed. I personally would like the feature I think, > though I can easily live without it ;-) > Perhaps we can do this with a better error message. There's something I really like with EMS SQL Manager. When the user change a property, the SQL is shown and can be changed. If the COMMIT failed, they show you the SQL and it also can be modified. Perhaps we just need this ? When there is a PostgreSQL related error message, pgAdmin shows the SQL it tried to execute and let the user change it. I think it would be a better behavior than what my patch implied. Would something like this be possible ? And one more question, I wonder why the SQL tab is not editable ? I can think of one reason but I'm not sure :) Regards. -- Guillaume. <!-- http://abs.traduc.org/ http://lfs.traduc.org/ http://docs.postgresqlfr.org/ -->
В списке pgadmin-hackers по дате отправления: