pg_dump: Separate table and sequence data object types
Instead of handling both sequence data and table data internally as
"table data", handle sequences separately under a "sequence set" type.
We already handled materialized view data differently, so it makes the
code somewhat cleaner to handle each relation kind separately at the top
level.
This does not change the output format, since there already was a
separate "SEQUENCE SET" archive entry type. A noticeable difference is
that SEQUENCE SET entries now always appear after TABLE DATA entries.
And in parallel mode there is less sorting to do, because the sequence
data entries are no longer considered table data.
Reviewed-by: Anastasia Lubennikova <a.lubennikova@postgrespro.ru>
Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
Branch
------
master
Details
-------
http://git.postgresql.org/pg/commitdiff/27d2c1232879aab5636da56b698764d553b9c5c6
Modified Files
--------------
src/bin/pg_dump/pg_dump.c | 11 +++++++----
src/bin/pg_dump/pg_dump.h | 1 +
src/bin/pg_dump/pg_dump_sort.c | 28 +++++++++++++++++-----------
3 files changed, 25 insertions(+), 15 deletions(-)