Thread

Commits

  1. Fix incorrect comment for get_agg_clause_costs

  1. Outdated comment in get_agg_clause_costs

    David Rowley <dgrowleyml@gmail.com> — 2021-07-22T14:29:50Z

    I noticed that get_agg_clause_costs still claims that it recursively
    finds Aggrefs in the expression tree, but I don't think that's been
    true since 0a2bc5d61.
    
    I've attached a patch that adjusts the comment so it's more aligned to
    what it now does.
    
    David
    
  2. Re: Outdated comment in get_agg_clause_costs

    David Rowley <dgrowleyml@gmail.com> — 2021-07-26T02:58:39Z

    On Fri, 23 Jul 2021 at 02:29, David Rowley <dgrowleyml@gmail.com> wrote:
    > I've attached a patch that adjusts the comment so it's more aligned to
    > what it now does.
    
    This was a bit more outdated than I first thought.  I also removed the
    mention of the function setting the aggtranstype and what it mentions
    about also gathering up "counts". I assume that was related to
    numOrderedAggs which is now done in preprocess_aggref().
    
    I ended up pushing to master and PG14.  The code was new to PG14 so I
    thought it made sense to keep master and 14 the same since 14 is not
    yet out the door.
    
    David