SQL/MED will have some kinds of planner hooks to support FDW-depending
plan execution. Then, we will need to support user-defined executor nodes.
The proposed SQL/MED has own "executor node hooks" in ForeignTableScan,
http://wiki.postgresql.org/wiki/SQL/MED#Executor
but I think it will be cleaner to support it in executor level.
The attached patch is an experimental code to do it; Plan struct has
"vtable" field as a set of functions to generate and execute PlanState
nodes. It changes large switch-case blocks in the current executor
into function-pointer calls as like as virtual functions in C++.
Is it worth doing? If we will go to the direction, I'll continue to
research it, like extensibility of Path nodes and EXPLAIN support.
-------- Essence of the patch --------
typedef struct Plan
{
NodeTag type;
PlanVTable *vtable; /* executor procs */
...
struct PlanVTable
{
ExecInitNode_type InitNode;
ExecProcNode_type ProcNode;
MultiProcNode_type MultiProcNode;
ExecEndNode_type EndNode;
...
make_seqscan()
{
node = makeNode(SeqScan);
node->vtable = &SeqScanVTable;
...
ExecReScan(node)
{
node->plan->vtable->ReScan(node);
...
--------
--
Itagaki Takahiro