Jul 2026

1 commit
JSON_TABLE PLAN clause
9 Jul 2026Alexander KorotkovmasterunreleasedAlso in: SQL/JSON json_table plan clause (landed as posted)Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)

Jul 2024

2 commits
SQL/JSON: Various improvements to SQL/JSON query function docs
9 Jul 2024Amit Langotemaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1REL_17_STABLE17beta3Also in: Doc Rework: Section 9.16.13 SQL/JSON Query Functions (modified before commit)
SQL/JSON: Fix some obsolete comments.
4 Jul 2024Amit Langotemaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1REL_17_STABLE17beta3REL_16_STABLE16.4
LLM analysis, 100% surelanded as postedfrom patchset #2971 of 1 reviewer saw the final version

The parse_expr.c comment changes match #297, including the aggregate-constructor corrections added after the first submission.

  • Alexander Lakhinsaw the final versionLLM analysis, 100% sureReported obsolete comments at #294 and requested the missing aggregate corrections at #296, which were incorporated in #297.

Mar - Apr 2024

7 commits
SQL/JSON: Fix issues with DEFAULT .. ON ERROR / EMPTY
18 Apr 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: sql/json remaining issue (part of a posted patchset)
JSON_TABLE: Add support for NESTED paths and columns
8 Apr 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #2865 of 12 reviewers saw the final version

Commits the NESTED component of #286, with expanded documentation, row-traversal cleanup, and final test corrections reviewed at #287, #288, and #289. This April 2024 matching attachment supersedes the historical discussion-only claims in thread #46113, thread #46380, and thread #38009.

  • Andres Freundsaw an earlier versionLLM analysis, 84% sureReviewed the original combined executor implementation at #72 and #156 before substantial JSON_TABLE restructuring.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in bb766cde63 supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in bb766cde63 supplies the credit; no review appears here.
  • Erik Rijkerssaw an earlier versionLLM analysis, 94% sureRecommended clearer JSON_TABLE common-syntax documentation at #16, before the later implementation changes.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in bb766cde63 supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw an earlier versionLLM analysis, 98% sureTested conversion behavior and nested paths at #204 and #236, before the split and nested-executor rewrite.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in bb766cde63 supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in bb766cde63 supplies the credit; no review appears here.
  • Alvaro Herrerasaw an earlier versionLLM analysis, 99% sureProposed grammar alternatives and missing NESTED tests at #139 and reviewed table-function presentation at #231.
  • Tomas Vondrasaw an earlier versionLLM analysis, 99% sureFound excessive memory use with multiple nested paths at #202 and confirmed the fix at #212.
  • Peter Eisentrautsaw an earlier versionLLM analysis, 99% sureSupplied JSON_TABLE grammar cleanup patches and said the remaining parser changes looked reasonable at #143.
  • Robert Haassaw an earlier versionLLM analysis, 93% sureExamined the NESTED lookahead design and its ambiguity risks at #150 and #154.
Fix JsonExpr deparsing to emit QUOTES and WRAPPER correctly
8 Apr 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 99% surepart of a posted patchsetmoderate changesfrom patchset #2860 of 1 reviewer saw the final version

Takes the deparsing-fix component of #286; the committed ruleutils.c logic additionally emits KEEP QUOTES with wrapper clauses and adjusts the corresponding expected output.

  • jian hesaw an earlier versionLLM analysis, 91% sureDemonstrated changed view semantics when QUOTES and WRAPPER were omitted at #278 and #281, before the final deparser logic was revised.
Fix typo introduced in 6185c9737
4 Apr 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Removes the duplicated word in the PASSING-argument comment identified inline by Jian at #260; there is no standalone attachment for this correction.

  • jian hesaw the final versionLLM analysis, 100% sureIdentified the duplicated word in the committed comment at #260, directly prompting this inline fix.
Add basic JSON_TABLE() functionality
4 Apr 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #2665 of 12 reviewers saw the final version

Commits the basic JSON_TABLE component of #266, with documentation, comment, diagnostic, and redundant pathname-field cleanup while excluding NESTED support. Its April 2024 attachment directly matches the implementation, superseding the historical discussion-only claims in thread #46113, thread #46380, and thread #38009.

  • Andres Freundsaw an earlier versionLLM analysis, 84% sureReviewed the original combined executor implementation at #72 and #156 before the JSON_TABLE split and restructuring.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in de3600452b supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in de3600452b supplies the credit; no review appears here.
  • Erik Rijkerssaw an earlier versionLLM analysis, 94% sureReviewed the shared JSON_TABLE syntax documentation at #16, before the implementation was substantially revised.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in de3600452b supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw an earlier versionLLM analysis, 99% sureReported inconsistent numeric and string overflow handling in JSON_TABLE at #204 and #210, prompting coercion changes.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in de3600452b supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in de3600452b supplies the credit; no review appears here.
  • Alvaro Herrerasaw an earlier versionLLM analysis, 99% sureReviewed table-function EXPLAIN behavior at #231 and supported the smaller basic patch at #252, before further plan-node changes.
  • Tomas Vondrasaw an earlier versionLLM analysis, 99% sureReported a large JSON_TABLE memory leak at #202 and confirmed its repair with performance measurements at #212.
  • Shruthi Gowdasaw an earlier versionLLM analysis, 99% sureTested implicit column paths and case folding at #226 and confirmed the quoted-name behavior at #228.
  • Peter Eisentrautsaw an earlier versionLLM analysis, 99% sureReviewed and supplied cleanup patches for the JSON_TABLE grammar at #143, before the later structural changes.
Avoid splitting errmsg string to span multiple lines
22 Mar 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Directly implements Kyotaro Horiguchi's report at #245 by joining an errmsg literal and adding the missing article; no attached patchset supplies this fix.

  • Kyotaro Horiguchisaw the final versionLLM analysis, 100% sureReported the split translatable error-message string at #245, directly prompting the inline correction.
Add SQL/JSON query functions
21 Mar 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: SQL/JSON revisited (not this thread)Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 99% surepart of a posted patchsetmoderate changesfrom patchset #2435 of 12 reviewers saw the final version

Commits the query-functions component of #243, with jump-bookkeeping cleanup, expanded comments, and removal of a scalar-extraction branch in src/backend/executor/execExprInterp.c. This March 2024 submission supersedes the ancestral sources claimed by thread #47160 and thread #38009, and the discussion-only attributions in thread #46113 and thread #46380.

  • Andres Freundsaw an earlier versionLLM analysis, 100% sureHis executor and jump-target reviews at #72 and #156 drove substantial redesign before landing.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in 6185c9737c supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in 6185c9737c supplies the credit; no review appears here.
  • Erik Rijkerssaw an earlier versionLLM analysis, 100% sureReported coercion crashes and later runaway evaluation at #43, #47, and #99, prompting fixes before subsequent redesigns.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in 6185c9737c supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw an earlier versionLLM analysis, 100% sureHis overflow tests at #204 and #210 exposed inconsistent coercion errors and helped motivate the input-function redesign.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in 6185c9737c supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in 6185c9737c supplies the credit; no review appears here.
  • Alvaro Herrerasaw an earlier versionLLM analysis, 99% sureReviewed soft-error handling, measured parser effects, and discussed behavior-clause ordering at #39, #118, #141, and #218.
  • Anton A. Melnikovsaw an earlier versionLLM analysis, 99% sureDiagnosed coercion failures and EMPTY ARRAY behavior at #81 and #83, before later executor rewrites.
  • Nikita Malakhovsaw an earlier versionLLM analysis, 99% sureReported a crashing JSON_QUERY case and forwarded corrective patches at #76, before substantial subsequent changes.
  • Tomas Vondrasaw an earlier versionLLM analysis, 99% sureReported and retested the coercion-cache memory leak at #202 and #212, then examined error-clause ordering at #216.

Jan 2024

4 commits
Add soft error handling to some expression nodes
24 Jan 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: SQL/JSON revisited (not this thread)Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #1869 of 11 reviewers saw the final version

The six-file expression soft-error component of #186, including EEOP_IOCOERCE_SAFE and soft domain checks, matches the committed code exactly. This January 2024 submission supersedes the April 2023 source claimed by thread #47160 and the historical discussion-only claims in thread #46113, thread #46380, and thread #38009.

  • Andres Freundsaw the final versionLLM analysis, 99% sureRequested a separate opcode to avoid ordinary-cast overhead at #72 and reviewed error-context ownership at #156, after which this prerequisite changed only in comments and rebasing.
  • Alvaro Herrerasaw an earlier versionLLM analysis, 99% sureReviewed error-context management, approved the earlier implementation at #63, and examined LLVM fixes at #70, before the separate SAFE-opcode redesign.
  • Erik Rijkerssaw an earlier versionLLM analysis, 96% sureReported successful check, check-world, and custom tests for the series containing the expression prerequisite at #37, before its SAFE-opcode redesign.
  • jian hesaw the final versionLLM analysis, 95% sureTested domain and typmod coercion and examined ExecEvalCoerceViaIOSafe at #180, with only comment changes afterward in the committed prerequisite.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in aaaf9449ec supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in aaaf9449ec supplies the credit; no review appears here.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in aaaf9449ec supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw the final versionLLM analysis, 100% sureThe prose reviewer list in aaaf9449ec supplies historical credit; his messages here begin after the commit.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in aaaf9449ec supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in aaaf9449ec supplies the credit; no review appears here.
  • Peter Eisentrautsaw the final versionLLM analysis, 99% sureThe prose reviewer list in aaaf9449ec supplies the credit, while his visible reviews concern grammar rather than this expression prerequisite.
Adjust populate_record_field() to handle errors softly
24 Jan 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: SQL/JSON revisited (not this thread)Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #1867 of 11 reviewers saw the final version

Commits the populate_record_field() soft-error component of #186, including jsonb_populate_record_valid(), with documentation and comment edits only. This January 2024 source supersedes the April 2023 attachment claimed by thread #47160 and the historical discussion-only claims in thread #46113, thread #46380, and thread #38009.

  • Andres Freundsaw an earlier versionLLM analysis, 99% sureReviewed error propagation, readability, and performance at #156, before domain handling and the testing API were added.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in 1edb3b491b supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in 1edb3b491b supplies the credit; no review appears here.
  • Erik Rijkerssaw an earlier versionLLM analysis, 90% sureTested problematic array and wrapper coercions at #43 and #47, before later changes to the population and domain-error paths.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in 1edb3b491b supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw the final versionLLM analysis, 100% sureThe commit supplies historical reviewer credit; his messages here postdate 1edb3b491b.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in 1edb3b491b supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in 1edb3b491b supplies the credit; no review appears here.
  • Alvaro Herrerasaw an earlier versionLLM analysis, 100% sureRequested a separate prerequisite and reliable error propagation at #39 and #141, before additional domain logic and test support.
  • jian hesaw an earlier versionLLM analysis, 99% sureReported domain and coercion behavior at #177 and #180, leading to further soft-error handling changes.
  • Peter Eisentrautsaw the final versionLLM analysis, 98% sureThe commit supplies reviewer credit, while his visible reviews concern grammar rather than record population 1edb3b491b.
Refactor code used by jsonpath executor to fetch variables
24 Jan 2024Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: SQL/JSON revisited (not this thread)Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #18611 of 11 reviewers saw the final version

The jsonpath_exec.c variable-fetch component of #186 matches the committed changes exactly. This January 2024 submission supersedes the older sources claimed by thread #47160 and thread #38009, while the discussion-only claims in thread #46113 and thread #46380 reflect historical context.

  • Andres Freundsaw the final versionLLM analysis, 95% sureThe prose reviewer list supplies credit, while his visible reviews do not isolate this callback refactoring faa2b953ba.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in faa2b953ba supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in faa2b953ba supplies the credit; no review appears here.
  • Erik Rijkerssaw the final versionLLM analysis, 98% sureThe commit credits him in its prose reviewer list, but no callback-specific review is identifiable here faa2b953ba.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in faa2b953ba supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw the final versionLLM analysis, 100% sureThe commit supplies historical reviewer credit; his messages here postdate faa2b953ba.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in faa2b953ba supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in faa2b953ba supplies the credit; no review appears here.
  • Alvaro Herrerasaw the final versionLLM analysis, 97% sureThe commit supplies reviewer credit, but his visible reviews address other components of the series faa2b953ba.
  • jian hesaw the final versionLLM analysis, 97% sureRequested updated callback documentation at #157, which Amit incorporated as confirmed at #161, with no later functional callback change.
  • Peter Eisentrautsaw the final versionLLM analysis, 98% sureThe commit supplies reviewer credit, while his visible reviews concern grammar rather than variable fetching faa2b953ba.
Test EXPLAIN (FORMAT JSON) ... XMLTABLE
22 Jan 2024Alvaro Herreramaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 87% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Álvaro replaced the proposed functional EXPLAIN change in #175 with FORMAT JSON coverage, explaining the correction at #184; the commit retains its alias-test adjustment but introduces the main regression tests without a matching attachment.

  • Amit Langotesaw the final versionLLM analysis, 80% sureApproved the proposed patch, including its retained alias-test change, at #176; the final implementation adds test coverage as explained in #184.

Dec 2023

1 commit
Simplify productions for FORMAT JSON [ ENCODING name ]
11 Dec 2023Alvaro Herreramaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #1531 of 1 reviewer saw the final version

Matches the FORMAT JSON grammar component of #153 modulo whitespace; Amit carried Álvaro's independent patch from #151 in the larger series.

  • Amit Langotesaw the final versiontrailerReviewed and incorporated Álvaro's grammar simplification in #153, whose component matches the committed version.

Oct 2023

1 commit
Add soft error handling to some expression nodes
2 Oct 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 99% surepart of a posted patchsetmoderate changesfrom patchset #640 of 1 reviewer saw the final version

The original soft-error prerequisite comes from #64, with a local change to track InputFunctionCallSafe() failure directly rather than reading state->escontext; later SAFE-opcode attachments postdate this commit.

  • Álvaro Herrerasaw an earlier versiontrailerReviewed context ownership and placement and approved #62 at #63, before the final local error-state logic change.

Jul 2023

9 commits
doc: add missing <returnvalue> and whitespace
28 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surelanded as postedfrom patchset #291 of 1 reviewer saw the final version

The func.sgml returnvalue markup and whitespace changes match Shinoda's attachment in #29.

  • Amit Langotesaw the final versioncommittercommitted the patch
Add more SQL/JSON constructor functions
26 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1Also in: PG 15 (and to a smaller degree 14) regression due to ExprEvalStep size (not this thread)Also in: SQL/JSON features for v15 (not this thread)Also in: SQL/JSON: functions (not this thread)
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #269 of 10 reviewers saw the final version

Commits the constructor component of #26 with light documentation, catalog-version, and enum cleanup. This July 2023 submission supersedes the March 2022 source claimed by thread #38009; the discussion-only claims in thread #46113 and thread #46380 describe historical context rather than the latest matching patchset.

  • Andres Freundsaw the final versionLLM analysis, 100% sureThe commit supplies historical reviewer credit; his messages here postdate 03734a7fed.
  • Pavel Stehulesaw the final versionLLM analysis, 100% sureThe prose reviewer list in 03734a7fed supplies the credit; no review appears here.
  • Andrew Alsupsaw the final versionLLM analysis, 100% sureThe prose reviewer list in 03734a7fed supplies the credit; no review appears here.
  • Erik Rijkerssaw the final versionLLM analysis, 98% sureThe commit supplies reviewer credit, while his visible early review concerns query and JSON_TABLE documentation 03734a7fed.
  • Zihong Yusaw the final versionLLM analysis, 100% sureThe prose reviewer list in 03734a7fed supplies this spelling and credit; no review appears here.
  • Himanshu Upadhyayasaw the final versionLLM analysis, 100% sureThe commit supplies historical reviewer credit; his messages here postdate 03734a7fed.
  • Daniel Gustafssonsaw the final versionLLM analysis, 100% sureThe prose reviewer list in 03734a7fed supplies the credit; no review appears here.
  • Justin Pryzbysaw the final versionLLM analysis, 100% sureThe prose reviewer list in 03734a7fed supplies the credit; no review appears here.
  • Alvaro Herrerasaw the final versionLLM analysis, 96% sureReviewed constructor grammar, conversion interfaces, and tests through #25, followed by extraction of prerequisites and light final edits.
  • Peter Eisentrautsaw an earlier versionLLM analysis, 99% sureIdentified inappropriate SQL:2016 keyword additions at #3, before subsequent constructor and conversion-interface revisions.
Rename a nonterminal used in SQL/JSON grammar
26 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #261 of 1 reviewer saw the final version

Exactly matches the extracted grammar-renaming component of #26, changing json_output_clause_opt to json_returning_clause_opt.

  • Alvaro Herrerasaw the final versionLLM analysis, 99% sureProposed the RETURNING-based nonterminal name at #6 and reviewed the later series at #25 before the rename was extracted.
Some refactoring to export json(b) conversion functions
26 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #261 of 1 reviewer saw the final version

Exactly matches the conversion-export prerequisite in #26, exporting datum_to_json(), datum_to_jsonb(), and jsonb_from_text().

  • Alvaro Herrerasaw the final versionLLM analysis, 99% sureReviewed conversion-function naming and exports at #25, which Amit incorporated in the extracted prerequisite at #26.
Code review for commit b6e1157e7d
21 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #231 of 1 reviewer saw the final version

Exactly matches the revised formatted_expr cleanup carried in #23, rather than the smaller initial standalone patch in #19.

  • Alvaro Herrerasaw the final versionLLM analysis, 100% sureRequested clearer formatted_expr guarantees at #18 and reviewed the revised cleanup at #25.
Unify JSON categorize type API and export for external use
20 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #171 of 1 reviewer saw the final version

Takes the unified categorization prerequisite from #17, differing only by an explicit extern on the json_categorize_type() declaration.

  • Álvaro Herrerasaw the final versiontrailerProposed unification, supplied API corrections at #13, and approved the resulting prerequisite at #18.
Don't include CaseTestExpr in JsonValueExpr.formatted_expr
13 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1REL_16_STABLE16beta3
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #111 of 1 reviewer saw the final version

Takes the CaseTestExpr-removal prerequisite from #11; its code matches, with only a small parse_expr.c comment wording difference.

  • Álvaro Herrerasaw the final versiontrailerReviewed the prerequisite at #9, followed by explanation and readability changes in #10 and #11.
Pass constructName to transformJsonValueExpr()
13 Jul 2023Amit Langotemaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1REL_16_STABLE16beta3
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #111 of 1 reviewer saw the final version

Exactly matches the constructName prerequisite in #11, passing the actual SQL/JSON construct name through transformJsonValueExpr().

  • Álvaro Herrerasaw the final versiontrailerReviewed the argument order and endorsed the prerequisite at #9, with only light changes before commit.
Make some indentation in gram.y consistent
8 Jul 2023Peter Eisentrautmaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1REL_16_STABLE16beta3
LLM analysis, 100% surelanded as postedfrom patchset #31 of 1 reviewer saw the final version

The gram.y indentation changes match Peter's standalone attachment in #3.

  • Amit Langotesaw the final versiontrailerApproved Peter's indentation correction at #4.

Credits across 25 commits

Author
Amit Langote9 LLM17
Alexander Korotkov7
Andrew Dunstan7
Nikita Glukhov7
Oleg Bartunov7
Teodor Sigaev7
jian he4
Peter Eisentraut1 LLM2
Alvaro HerreraLLM1
Nikita Malakhov1
Shinoda, Noriyoshi (PN Japan FSIP)1
Reviewer
Alvaro Herrera5
Amit Langotecommitter3
Erik Rijkers2
jian he2
Andres Freund1
Andrew Alsup1
Daniel Gustafsson1
Himanshu Upadhyaya1
Justin Pryzby1
Pavel Stehule1
Vladlen Popolitov1
Zhihong Yu1
Reported by
jian he3
Alexander Lakhin1
Kyotaro Horiguchi1
Co author
Alexander Korotkov1
Alvaro Herrera1
Amit Langote1
Andrew Dunstan1
Anton A. Melnikov1
Nikita Glukhov1
Oleg Bartunov1
Teodor Sigaev1
Committer
Amit Langote21
Alvaro Herrera2
Alexander Korotkov1
Peter Eisentraut1

N LLM means N of the credits come from LLM analysis, not from commit trailers.