Re: SQL:2023 JSON simplified accessor support

> On Aug 26, 2025, at 11:52, Alexandra Wang <alexandra(dot)wang(dot)oss(at)gmail(dot)com> wrote:

> <v14-0002-Allow-Generic-Type-Subscripting-to-Accept-Dot-No.patch><v14-0003-Export-jsonPathFromParseResult.patch><v14-0001-Allow-transformation-of-only-a-sublist-of-subscr.patch><v14-0005-Implement-read-only-dot-notation-for-jsonb.patch><v14-0007-Implement-jsonb-wildcard-member-accessor.patch><v14-0006-Implement-Jsonb-subscripting-with-slicing.patch><v14-0004-Extract-coerce_jsonpath_subscript.patch>

A few other comments:

-static Node *
-transformIndirection(ParseState *pstate, A_Indirection *ind)
+Node *
+transformIndirection(ParseState *pstate, A_Indirection *ind,
+ bool *trailing_star_expansion)
{
Node *last_srf = pstate->p_last_srf;
Node *result = transformExprRecurse(pstate, ind->arg);
@@ -454,12 +454,7 @@ transformIndirection(ParseState *pstate, A_Indirection *ind)
if (IsA(n, A_Indices))
subscripts = lappend(subscripts, n);
else if (IsA(n, A_Star))
- {
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("row expansion via \"*\" is not supported here"),
- parser_errposition(pstate, location)));
- }
+ subscripts = lappend(subscripts, n);
else
{
Assert(IsA(n, String));

Now, the 3 clauses are quite similar, the if-elseif-else can be simplified as:

If (!IsA(n, A_Indices) && !Is_A(n, A_Start))
Assert(IsA(n, String));
Subscripts = lappend(subscripts, n)

+static bool
+jsonb_check_jsonpath_needed(List *indirection)
+{
+ ListCell *lc;
+
+ foreach(lc, indirection)
+ {
+ Node *accessor = lfirst(lc);
+
+ if (IsA(accessor, String))
+ return true;

Like I mentioned in my previous comment, if we convert String (dot notation) to Indices in rewritten phase, then jsonpath might be no longer needed, and the code logic is much simplified.

--- a/src/backend/parser/parse_node.c
+++ b/src/backend/parser/parse_node.c
@@ -244,7 +244,7 @@ transformContainerSubscripts(ParseState *pstate,
Node *containerBase,
Oid containerType,
int32 containerTypMod,
- List *indirection,
+ List **indirection,
bool isAssignment)
{
SubscriptingRef *sbsref;
@@ -280,7 +280,7 @@ transformContainerSubscripts(ParseState *pstate,
* element. If any of the items are slice specifiers (lower:upper), then
* the subscript expression means a container slice operation.
*/
- foreach(idx, indirection)
+ foreach(idx, *indirection)
{
A_Indices *ai = lfirst_node(A_Indices, idx);

Should this foreach be pull up to out of transformContainerSubscripts()? Originally transformContainerSubscripts() runs only once, but now it’s placed in a while loop, and every time this foreach needs to go through the entire indirection list, which are duplicated computation.

Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/