Re: [HACKERS] Function Volatility and Views Unexpected Behavior

`explain analyze verbose select data1 from table1_silly_view where id >=10 and id <= 100;`

does an index scan and only runs the volatile function for rows in the view where id >= 10 and id <=100

Subquery Scan on table1_silly_view  (cost=0.29..33.77 rows=91 width=8) (actual time=2.552..206.563 rows=91 loops=1)

  Output: table1_silly_view.data1

  ->  Index Scan using table1_pkey on public.table1  (cost=0.29..32.86 rows=91 width=20) (actual time=2.550..206.425 rows=91 loops=1)

        Output: NULL::integer, table1.data1, something_silly(table1.id)

        Index Cond: ((table1.id >= 10) AND (table1.id <= 100))

Planning time: 0.526 ms

Execution time: 206.724 ms

whereas

`explain analyze verbose select data1 from table1_silly_view where id in (select id from table1 where id >= 10 and id <=100);`

does a full sequential scan, over the view, producing whatever side effects the volatile function does for every row in the view even though they produce the same output and have what should be equivalent quals. 

Hash Semi Join  (cost=11.24..2793.50 rows=91 width=8) (actual time=23.603..22759.297 rows=91 loops=1)

  Output: table1_1.data1

  Hash Cond: (table1_1.id = table1.id)

  ->  Seq Scan on public.table1 table1_1  (cost=0.00..2655.00 rows=10000 width=20) (actual time=2.468..22720.942 rows=10000 loops=1)

        Output: table1_1.id, table1_1.data1, something_silly(table1_1.id)

  ->  Hash  (cost=10.11..10.11 rows=91 width=4) (actual time=0.484..0.484 rows=91 loops=1)

        Output: table1.id

        Buckets: 1024  Batches: 1  Memory Usage: 12kB

        ->  Index Only Scan using table1_pkey on public.table1  (cost=0.29..10.11 rows=91 width=4) (actual time=0.383..0.430 rows=91 loops=1)

              Output: table1.id

              Index Cond: ((table1.id >= 10) AND (table1.id <= 100))

              Heap Fetches: 91

Planning time: 0.877 ms

Execution time: 22759.448 ms

I recognize that it is an anti-pattern to put a volatile function call in a view, and don't know that there's a better way of dealing with it, as not using indexes in a view that has a volatile function call in it at all seems like a very bad choice, but still think it might be something to document better. 

-David