On Wed, Oct 26, 2022 at 2:13 AM jack...@gmail.com <jack...@gmail.com> wrote:
> typedef struct A_Expr
>
>
>
> {
>
>
>
> pg_node_attr(custom_read_write)
>
>
>
> NodeTag type;
>
>
>
> A_Expr_Kind kind; /* see above */
>
>
>
> List *name; /* possibly-qualified name of operator */
>
>
>
> Node *lexpr; /* left argument, or NULL if none */
>
>
>
> Node *rexpr; /* right argument, or NULL if none */
>
>
>
> int location; /* token location, or -1 if unknown */
>
>
>
> } A_Expr;
>
>
>
> I run a sql like select a,b from t3 where a > 1; and I get the parseTree
> for selectStmt:
>
>
>
> why the name is '-' but not '>'?
>
>
Given the general lack of interest in this so far I'd suggest you put
together a minimal test case that includes a simple print-to-log command
patched into HEAD showing the problematic parse tree in its internal form.
Posting an image of some custom visualization that seems evidently produced
by custom code that itself may be buggy against an unspecified server with
no indication how any of this all works doesn't seem like enough detail and
there is little reason to think that such an obvious bug could exist. I
do agree that your expectation seems quite sound. Though I do not have the
faintest idea how to actually go about reproducing your result even in the
minimal way described above (though I know enough to know it is possible,
just not where to patch).
David J.
