[Bug] Unneeded sorting of compressed chunk table with aggregation

[akuzm]

create table pvagg(s int, a int);

select create_hypertable('pvagg', 'a', chunk_time_interval => 1000);

insert into pvagg select 1, generate_series(1, 999);
insert into pvagg select 2, generate_series(1001, 1999);

alter table pvagg set (timescaledb.compress, timescaledb.compress_segmentby = 's');

select count(compress_chunk(x)) from show_chunks('pvagg') x;

analyze pvagg;

set enable_indexscan to off;

explain (costs off) select count(a), s from pvagg group by s;
                            QUERY PLAN                            
──────────────────────────────────────────────────────────────────
 HashAggregate
   Group Key: _hyper_50_339_chunk.s
   ->  Append
         ->  Custom Scan (DecompressChunk) on _hyper_50_339_chunk
               ->  Sort
                     Sort Key: compress_hyper_51_345_chunk.s
                     ->  Seq Scan on compress_hyper_51_345_chunk
         ->  Custom Scan (DecompressChunk) on _hyper_50_340_chunk
               ->  Sort
                     Sort Key: compress_hyper_51_346_chunk.s
                     ->  Seq Scan on compress_hyper_51_346_chunk

We're doing hash aggregation here, so these Sorts are just useless work. The reason is that we don't create an unsorted version of the decompression path if the query pathkeys can be pushed down. Queries with aggregation have query pathkeys set to the grouping columns. See ts_decompress_chunk_generate_paths.