ne 24. 7. 2022 v 21:39 odesílatel Fabien COELHO <coelho@cri.ensmp.fr> napsal:
>> Attached v4 simplifies the format and fixes this one. > > I think this goes way way overboard in terms of invasiveness. There's no > need to identify individual call sites of PSQLexec. [...]
ISTM that having the information was useful for the user who actually asked for psql to show hidden queries, and pretty simple to get, although somehow invasive.
> It also looks like a mess from the translatibility standpoint. > You can't expect "%s QUERY" to be a useful thing for translators.
Sure. Maybe I should have used an enum have a explicit switch in echoQuery, but I do not like writing this kind of code.
Attached a v5 without hinting at the origin of the query beyond internal or not.
I had just one question - with this patch, the format of output of modes ECHO ALL and ECHO QUERIES will be different, and that can be a little bit messy. On second hand, the prefix --QUERY can be disturbing in echo queries mode. It is not a problem in echo all mode, because queries and results are mixed together. So in the end, I think the current design can work.
All tests passed, this is trivial patch without impacts on users