Seems to me that your tool could set the env var that you want.
If you don't export it, I think it shouldn't "leak" but your child process
should get it as part of their environment.
On Fri, Sep 22, 2023 at 12:43 PM Dominique Devienne <ddevie...@gmail.com>
wrote:
> On Fri, Sep 22, 2023 at 5:19 PM Luca Ferrari <fluca1...@gmail.com> wrote:
>
>> On Fri, Sep 22, 2023 at 3:24 PM Dominique Devienne <ddevie...@gmail.com>
>> wrote:
>> You added information I was not aware before: you are dumping
>> PostgreSQL to restore it into SQLite, while I was thinking you wanted
>> to do some stuff with a PostgreSQL-to-PostgreSQL backup and restore.
>>
>
> I must be really bad at explaining things today, sorry...
>
> Because of course I'm doing PostgreSQL-to-PostgreSQL backup/restore.
> The fact I use a custom SQLite DB file as the backup's "format" instead of
> a "pile-of-files" (or a TAR of that pile), is the only difference.
>
> That, and the fact it's a backup that mixes one full schema with part of
> another.
> And that my SQLite backup format is much more convenient and "expressive"
> IMHO,
> since fully introspectable ("semi-opaque", since rows are still COPY
> BINARY encoded).
>
> Writing the backup as an SQLite DB incurs a little overhead, of course,
> compared to just appending to per-table files, but not that much when
> in non-transactional mode. Acceptable tradeoff compared to the fact one
> can open the backup easily and see the tables and rows almost as-is,
> which also opens the door to partial restores (there are separates "roots",
> independent entity trees, in these schemas).
>
