Re: [HACKERS] How to look at the Expression Trees

2011-03-22 Thread Heikki Linnakangas
On 22.03.2011 05:39, Vaibhav Kaushal wrote: You said: ExecMakeFunctionResultNoSets is used to evaluate function calls. What are the 'functions' there? Are they the user supplied pl/PGSQL style user functions, the functions handled by fmgr or are they just another C function which make the Expre

Re: [HACKERS] How to look at the Expression Trees

2011-03-21 Thread Vaibhav Kaushal
Hi, You said: ExecMakeFunctionResultNoSets is used to evaluate function calls. What are the 'functions' there? Are they the user supplied pl/PGSQL style user functions, the functions handled by fmgr or are they just another C function which make the Expression Evaluator? Regards, Vaibhav

Re: [HACKERS] How to look at the Expression Trees

2011-03-21 Thread Heikki Linnakangas
On 21.03.2011 14:37, Vaibhav Kaushal wrote: I am already using the postgresql server with -d 4 option and it shows a lot of things. But I am not able to see the Expression State trees. To be precise, debug_print_plan=on prints the expression tree that comes from planner, not the execution tree

Re: [HACKERS] How to look at the Expression Trees

2011-03-21 Thread Vaibhav Kaushal
On Mon, Mar 21, 2011 at 5:47 PM, Heikki Linnakangas < heikki.linnakan...@enterprisedb.com> wrote: > > Yes. There's actually two "trees" involved. The planner produces a tree of > Expr nodes, and ExecInitExpr prepares a tree of ExprState nodes that mirrors > the first tree. The ExprStates contain r

Re: [HACKERS] How to look at the Expression Trees

2011-03-21 Thread Heikki Linnakangas
On 21.03.2011 13:44, Vaibhav Kaushal wrote: Hi all, I have been trying to work on the expression evaluator (trying to alter it just for the seqscan case). I have understood a few things. I wish someone could tell me if I am wrong at some point. As far as I have gone through the code, I think: 1

[HACKERS] How to look at the Expression Trees

2011-03-21 Thread Vaibhav Kaushal
Hi all, I have been trying to work on the expression evaluator (trying to alter it just for the seqscan case). I have understood a few things. I wish someone could tell me if I am wrong at some point. As far as I have gone through the code, I think: 1. Quals are formulated by planner 2. Quals are