Re: WIP: a way forward on bootstrap data

John Naylor <jcnaylor(at)gmail(dot)com> writes:
> On 3/3/18, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> * I'm a little disturbed by the fact that 0002 has to "re-doublequote
>> values that are macros expanded by initdb.c". I see that there are only
>> a small number of affected places, so maybe it's not really worth working
>> harder, but I worry that something might get missed. Is there any way to
>> include this consideration in the automated conversion, or at least to
>> verify that we found all the places to quote? Or, seeing that 0004 seems
>> to be introducing some quoting-related hacks to genbki.pl anyway, maybe
>> we could take care of the issue there?

> The quoting hacks are really to keep the postgres.bki diff as small as
> possible (attached). The simplest and most air-tight way to address
> your concern would be to double-quote everything when writing the bki
> file. That could be done last as a follow-up.

Oh, if you're cross-checking by diff'ing the produced .bki file, then
that's sufficient to address my concern here. No need to do more.

regards, tom lane