Horton

Re: remaining sql/json patches

jian he <jian.universality@gmail.com>

View thread
From: jian he <jian.universality@gmail.com>
To: Amit Langote <amitlangote09@gmail.com>
Cc: Alvaro Herrera <alvherre@alvh.no-ip.org>, Tomas Vondra <tomas.vondra@enterprisedb.com>, Himanshu Upadhyaya <upadhyaya.himanshu@gmail.com>, Erik Rijkers <er@xs4all.nl>, Andres Freund <andres@anarazel.de>, Andrew Dunstan <andrew@dunslane.net>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2024-04-01T08:56:00Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. SQL/JSON: Various improvements to SQL/JSON query function docs

  2. SQL/JSON: Fix some obsolete comments.

  3. SQL/JSON: Fix issues with DEFAULT .. ON ERROR / EMPTY

  4. JSON_TABLE: Add support for NESTED paths and columns

  5. Fix JsonExpr deparsing to emit QUOTES and WRAPPER correctly

  6. Fix typo introduced in 6185c9737

  7. Add basic JSON_TABLE() functionality

  8. Avoid splitting errmsg string to span multiple lines

  9. Add SQL/JSON query functions

  10. Implement various jsonpath methods

  11. Add soft error handling to some expression nodes

  12. Adjust populate_record_field() to handle errors softly

  13. Refactor code used by jsonpath executor to fetch variables

  14. Test EXPLAIN (FORMAT JSON) ... XMLTABLE

  15. Simplify productions for FORMAT JSON [ ENCODING name ]

  16. Add trailing commas to enum definitions

  17. doc: add missing <returnvalue> and whitespace

  18. Add more SQL/JSON constructor functions

  19. Rename a nonterminal used in SQL/JSON grammar

  20. Some refactoring to export json(b) conversion functions

  21. Don't include CaseTestExpr in JsonValueExpr.formatted_expr

  22. Code review for commit b6e1157e7d

  23. Pass constructName to transformJsonValueExpr()

  24. Unify JSON categorize type API and export for external use

  25. Make some indentation in gram.y consistent

  26. Allow most keywords to be used as column labels without requiring AS.

  27. Reduce size of backend scanner's tables.

  28. Use perfect hashing, instead of binary search, for keyword lookup.

Attachments

On Mon, Apr 1, 2024 at 8:00 AM jian he <jian.universality@gmail.com> wrote:
>
> +-- Should fail (JSON arguments are not passed to column paths)
> +SELECT *
> +FROM JSON_TABLE(
> + jsonb '[1,2,3]',
> + '$[*] ? (@ < $x)'
> + PASSING 10 AS x
> + COLUMNS (y text FORMAT JSON PATH '$ ? (@ < $x)')
> + ) jt;
> +ERROR:  could not find jsonpath variable "x"
>
> the error message does not correspond to the comments intention.
> also "y text FORMAT JSON" should be fine?

sorry for the noise, i've figured out why.

> only the second last example really using the PASSING clause.
> should the following query work just fine in this context?
>
> create table s(js jsonb);
> insert into s select '{"a":{"za":[{"z1": [11,2222]},{"z21": [22,
> 234,2345]}]},"c": 3}';
> SELECT sub.* FROM s,JSON_TABLE(js, '$' passing 11 AS "b c", 1 + 2 as y
> COLUMNS (xx int path '$.c ? (@ == $y)')) sub;
>
>
> I thought the json and text data type were quite similar.
> should these following two queries return the same result?
>
> SELECT sub.* FROM s, JSON_TABLE(js, '$' COLUMNS(
> xx int path '$.c',
> nested PATH '$.a.za[1]' columns (NESTED PATH '$.z21[*]' COLUMNS (a12
> jsonb path '$'))
> ))sub;
>
> SELECT sub.* FROM s,JSON_TABLE(js, '$' COLUMNS (
> c int path '$.c',
> NESTED PATH '$.a.za[1]' columns (z json path '$')
> )) sub;
sorry for the noise, i've figured out why.

there are 12 appearances of "NESTED PATH" in  sqljson_jsontable.sql.
but we don't have a real example of  NESTED PATH nested with NESTED PATH.
so I added some real tests on it.
i also added some tests about the PASSING clause.
please check the attachment.


/*
 * JsonTableInitPlan
 * Initialize information for evaluating a jsonpath given in
 * JsonTablePlan
 */
static void
JsonTableInitPathScan(JsonTableExecContext *cxt,
  JsonTablePlanState *planstate,
  List *args, MemoryContext mcxt)
{
JsonTablePlan *plan = (JsonTablePlan *) planstate->plan;
int i;

planstate->path = DatumGetJsonPathP(plan->path->value->constvalue);
planstate->args = args;
planstate->mcxt = AllocSetContextCreate(mcxt, "JsonTableExecContext",
ALLOCSET_DEFAULT_SIZES);

/* No row pattern evaluated yet. */
planstate->currentRow = PointerGetDatum(NULL);
planstate->currentRowIsNull = true;

for (i = plan->colMin; i <= plan->colMax; i++)
cxt->colexprplans[i] = planstate;
}

JsonTableInitPathScan's work is to init/assign struct
JsonTablePlanState's elements.
maybe we should just put JsonTableInitPathScan's work into JsonTableInitPlan
and also rename JsonTableInitPlan to "JsonTableInitPlanState" or
"InitJsonTablePlanState".



JsonTableSiblingJoin *join = (JsonTableSiblingJoin *) plan;
just rename the variable name, seems unnecessary?