Re: pg_dump multi VALUES INSERT - Mailing list pgsql-hackers

From Surafel Temesgen
Subject Re: pg_dump multi VALUES INSERT
Date
Msg-id CALAY4q-B1uFAe9k12Lcm=ADr+t51M2eYrV77C_3igm2TS-1+Xw@mail.gmail.com
Whole thread Raw
In response to Re: pg_dump multi VALUES INSERT  (Fabien COELHO <coelho@cri.ensmp.fr>)
Responses Re: pg_dump multi VALUES INSERT
Re: pg_dump multi VALUES INSERT
List pgsql-hackers


On Tue, Dec 25, 2018 at 2:47 PM Fabien COELHO <coelho@cri.ensmp.fr> wrote:

Thank you for looking into it

Hello Surafel,

> Thank you for informing, Here is an updated patch against current master

Patch applies cleanly, compiles, "make check" is okay, but given that the
feature is not tested...

Feature should be tested somewhere.

ISTM that command-line switches with optional arguments should be avoided:
This feature is seldom used (hmmm... 2 existing instances), because it
interferes with argument processing if such switches are used as the last
one. It is only okay with commands which do not expect arguments. For
backward compatibility, this suggests to add another switch, eg
--insert-multi=100 or whatever, which would possibly default to 100. The
alternative is to break compatibility with adding a mandatory argument,
but I guess it would not be admissible to committers.

Function "atoi" parses "1zzz" as 1, which is debatable, so I'd suggest to
avoid it and use some stricter option and error out on malformed integers.

The --help output does not document the --inserts argument, nor the
documentation.


done

There is an indendation issue within the while loop.

Given that the implementation is largely a copy-paste of the preceding
function, I'd suggest to simply extend it so that it takes into account
the "multi insert" setting and default to the previous behavior if not
set. 
 

At first i also try to do it like that but it seems the function will became long and more complex to me

Regards

Surafel

Attachment

pgsql-hackers by date:

Previous
From: Fabien COELHO
Date:
Subject: Re: Offline enabling/disabling of data checksums
Next
From: Tom Lane
Date:
Subject: Re: Poor buildfarm coverage of strong-random alternatives