Re: Avoid full GIN index scan when possible

Yeah, I can confirm those results, although on my system the timings are
a bit different (I haven't tested v8):

                                        |        Query time, ms
             WHERE condition            | master |    v9
---------------------------------------+--------+---------
  a @> '{}'                             |    610 |    589
  a @> '{}' and b @> '{}'               |    185 |    665
  a @> '{}' and b @> '{}' and c @> '{}' |    185 |    741

So that's something we probably need to address, perhaps by using the
GIN_CAT_EMPTY_QUERY entries as proposed.

I've also tested this on a database storing mailing lists archives with
a trigram index, and in that case the performance with short values gets
much better. The "messages" table has two text fields with a GIN trigram
index - subject and body, and querying them with short/long values works
like this:

                    WHERE                    |  master  |    v9
  --------------------------------------------------------------
  subject LIKE '%aa%' AND body LIKE '%xx%'   |    4943  |  4052
  subject LIKE '%aaa%' AND body LIKE '%xx%'  |      10  |    10
  subject LIKE '%aa%' AND body LIKE '%xxx%'  |     380  |    13
  subject LIKE '%aaa%' AND BODY LIKE '%xxx%' |       2  |     2

which seems fairly nice. I've done tests with individual columns, and
that seems to be working fine too.