pgsql: Parse catalog .dat files as a whole when compiling the backend

Parse catalog .dat files as a whole when compiling the backend

Previously Catalog.pm eval'd each individual hash reference
so that comments and whitespace can be preserved when running
reformat-dat-files. This is unnecessary when building, and we can save
~15% off the run time of genbki.pl by simply slurping and eval'-ing
the whole file at once. This saves a bit of time, especially in highly
parallel builds, since most build targets depend on this script's outputs.

Report and review by Andres Freund
Discussion: https://www.postgresql.org/message-id/CAFBsxsGW%3DWRbnxXrc8UqqR479XuxtukSFWV-hnmtgsbuNAUO6w%40mail.gmail.com

Branch
------
master

Details
-------
https://git.postgresql.org/pg/commitdiff/69eb643b2582f180c91b2d14c896402e54040231

Modified Files
--------------
src/backend/catalog/Catalog.pm | 22 ++++++++++++++++++++--
1 file changed, 20 insertions(+), 2 deletions(-)