Re: machine-readable explain output v4
От | Robert Haas |
---|---|
Тема | Re: machine-readable explain output v4 |
Дата | |
Msg-id | 603c8f070908051948o10832925paae2db13e0e5eb15@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: machine-readable explain output v4 (Andres Freund <andres@anarazel.de>) |
Ответы |
Re: machine-readable explain output v4
|
Список | pgsql-hackers |
On Sun, Aug 2, 2009 at 7:29 PM, Andres Freund<andres@anarazel.de> wrote: >> Well, the whole explain output format is pretty idiosyncratic, and I >> had to work pretty hard to beat it into submission. I think that it >> would not be totally trivial to do what you're suggesting here because >> it would require adding code to manage es->indent outside of >> ExplainPrintPlan(), which we currently don't. I'm not sure whether >> that works out to a net win. > Thats why I suggested doing it for JSON/XML only. E.g. like in the attached > patch. The result looks simpler for my eyes. I looked at this some more. I think it's a mess. It's probably right to do what you're suggesting here, but this patch only changes pieces of it without making the whole thing consistent. report_triggers(), for example, kludges a value into es->indent but then ignores it when determining how much to indent <Trigger>, etc. I can't even figure out why this works now, let alone being able to maintain it down the line. I'm working on trying to fix this. ...Robert
В списке pgsql-hackers по дате отправления: