Mar - Apr 2022

4 commits
psql: Show all query results by default
4 Apr 2022Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #1241 of 5 reviewers saw the final version

Lands the remaining feature from #124 after the COPY refactoring was extracted separately, retaining its documentation, tests, and asynchronous result handling. Peter preserves existing function names and simplifies incidental edits, as explained in #127.

  • Peter Eisentrautsaw the final versiontrailerReviewed timing, compatibility, savepoints, replication errors, and crash output throughout the thread, then edited and committed the final version in #127.
  • Iwata, Ayasaw an earlier versiontrailerTested combined queries and reported misplaced warnings and a regression-output whitespace error in #2, before substantial result-handling changes.
  • Daniel Veritesaw an earlier versiontrailerFound the errverbose regression in #4 and reviewed removal of the option and COPY-test changes in #13, well before the later fixes.
  • Kyotaro Horiguchisaw an earlier versiontrailerReviewed documentation, FETCH_COUNT interaction, COPY failure propagation, and timing in #15 and #20, before subsequent functional revisions.
  • vignesh Csaw an earlier versiontrailerRequested function comments and removal of an accidental duplicate in #23 and #25, before watch support and the post-revert corrections.
psql: Refactor ProcessResult()
1 Apr 2022Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 98% surepart of a posted patchsetlight changesfrom patchset #1242 of 2 reviewers saw the final version

Extracts HandleCopyResult from src/bin/psql/common.c in #124, retaining ProcessResult as its caller so the refactoring can land before the feature. The helper closely follows Fabien's implementation with comment and formatting adjustments.

  • Kyotaro Horiguchisaw the final versionLLM analysis, 88% sureReviewed HandleCopyResult's return contract, redundant success checks, and comments in #20, with the accepted cleanup retained in the extracted helper.
  • Peter Eisentrautsaw the final versioncommittercommitted the patch
psql: Refactor SendQuery()
31 Mar 2022Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 97% surepart of a posted patchsetmoderate changesfrom patchset #1231 of 2 reviewers saw the final version

Extracts the query-execution separation and savepoint-status propagation underlying #123 into ExecQueryAndProcessResult. It deliberately retains PQexec and single-result processing until the subsequent feature commit.

  • Tom Lanesaw an earlier versionLLM analysis, 90% sureIdentified the lost result-dependent savepoint checks in #60, which led to the explicit savepoint-status propagation used by this extraction.
  • Peter Eisentrautsaw the final versioncommittercommitted the patch
psql: Add test for psql behavior on server crash
29 Mar 2022Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #1221 of 1 reviewer saw the final version

Extracts the voluntary-disconnection test from src/bin/psql/t/001_basic.pl in #122, polishing test labels and expecting the pre-feature connection-closure text. The remaining output adjustment persists in #123.

  • Peter Eisentrautsaw the final versioncommittercommitted the patch

Feb - Mar 2022

2 commits
psql: Additional tests
1 Mar 2022Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #110

Matches the second patch of #110: TAP coverage for timing, ENCODING updates, and LISTEN/NOTIFY in src/bin/psql/t/001_basic.pl. Only this member of the two-patch series is committed here.

Improve some psql test code
28 Feb 2022Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #110

Matches the first patch of #110, separating psql_like and psql_fails_like and standardizing argument order. The additional tests are committed separately.

Sep - Oct 2021

4 commits
psql: Fix some scan-build warnings
12 Oct 2021Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 86% surepart of a posted patchsetlight changesfrom patchset #94

Corresponds to the independent scan-build cleanup attachment described in #94, caching timing settings in common.c and removing dead stores in copy.c and describe.c. Attachment extraction failed, so exact textual fidelity cannot be checked.

psql: Add test for handling of replication commands
12 Oct 2021Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 95% surepart of a posted patchsetlight changesfrom patchset #94

Separately lands the START_REPLICATION error-handling test explicitly described in #94, adding replication configuration and an expected PQresultStatus failure to 001_basic.pl. The basic tests from the same attachment landed in another commit.

psql: More tests
12 Oct 2021Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 95% surepart of a posted patchsetlight changesfrom patchset #94

Creates src/bin/psql/t/001_basic.pl with command-line and help tests, matching the basic-test portion described in #94. The replication test and scan-build fixes were split into separate commits.

psql: Add various tests
29 Sep 2021Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #921 of 1 reviewer saw the final version

The AUTOCOMMIT, ON_ERROR_ROLLBACK, and ECHO tests match Peter's #91 exactly and are also included unchanged in the later #92. Only those tests are extracted from that combined submission.

  • Fabien COELHOsaw the final versiontrailerIntegrated Peter's tests and repaired his feature to pass them in #92, confirming the COMMIT AND CHAIN problem.

Aug 2021

1 commit
psql: Add test for query canceling
20 Aug 2021Peter Eisentrautmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #841 of 1 reviewer saw the final version

Extracts Peter's cancellation TAP test from #84, where Fabien had incorporated #78. The committed test updates node construction, adds -X, shortens pg_sleep, and revises comments and labels.

  • Fabien COELHOsaw the final versiontrailerIntegrated, annotated, and ran Peter's cancellation test with the revised feature in #84, before light committer adjustments.

Apr 2021

2 commits
Revert "psql: Show all query results by default"
15 Apr 2021Peter Eisentrautmaster14beta1REL_19_STABLE14beta1REL_18_STABLE14beta1REL_17_STABLE14beta1REL_16_STABLE14beta1REL_15_STABLE14beta1REL_14_STABLE14beta1reverts 1 commit here
LLM analysis, 100% surewritten from the discussion, no matching patch0 of 3 reviewers saw the final version

Reverses 3a51306722 after the regression reports and requests to revert in #60, #64, and #67. Peter announces the revert in #73; no revert patch was attached.

  • Tom Lanesaw an earlier versionLLM analysis, 99% sureReported broken savepoint handling and requested a revert in #60, with later notice-buffer and crash-output findings preceding the eventual corrected feature.
  • Michael Paquiersaw an earlier versionLLM analysis, 99% sureReviewed cancellation fixes in #53 and #57, then supported reverting the defective feature in #64 and #70.
  • Nathan Bossartsaw an earlier versionLLM analysis, 98% sureReproduced the START_REPLICATION infinite error loop, supplied an inline fix, and explicitly supported reverting in #67.
psql: Show all query results by default
6 Apr 2021Peter Eisentrautmaster14beta1REL_19_STABLE14beta1REL_18_STABLE14beta1REL_17_STABLE14beta1REL_16_STABLE14beta1REL_15_STABLE14beta1REL_14_STABLE14beta1
LLM analysis, 100% suremodified before commitmoderate changesfrom patchset #431 of 5 reviewers saw the final version

Lands Peter's revised submission of Fabien's patch from #43, with ShowErrorMessage inlined, result-pointer cleanup changed, and normal notice processing restored around COPY. Peter explains the COPY notice-order correction in #45.

  • Iwata, Ayasaw an earlier versiontrailerReported warning ordering and a failing expected-output line in #2, before notice buffering, watch support, and other functional changes.
  • Daniel Veritesaw an earlier versiontrailerFound broken errverbose in #4 and reviewed option removal and test changes in #13, before later watch and error-handling revisions.
  • Peter Eisentrautsaw the final versiontrailerReviewed timing in #9, revised the feature in #43, and fixed COPY notice ordering while committing it in #45.
  • Kyotaro Horiguchisaw an earlier versiontrailerReviewed documentation, FETCH_COUNT, COPY processing, and timing in #15 and #20, before substantive subsequent changes.
  • vignesh Csaw an earlier versiontrailerRequested missing function headers and identified a duplicate describe.c comment in #23 and #25, before watch support and compatibility changes.

Feb 2021

1 commit
Extend a test case a little
26 Feb 2021Peter Eisentrautmaster14beta1REL_19_STABLE14beta1REL_18_STABLE14beta1REL_17_STABLE14beta1REL_16_STABLE14beta1REL_15_STABLE14beta1REL_14_STABLE14beta1
LLM analysis, 99% surepart of a posted patchsetlight changesfrom patchset #392 of 2 reviewers saw the final version

Extracts the copydml.sql and copydml.out changes adding tg_when to distinguish BEFORE and AFTER trigger notices, present through #39. These hunks disappear from Peter's subsequent #43.

  • Daniel Veritesaw the final versionLLM analysis, 97% sureSpecifically examined and questioned the tg_when test changes in #13, prompting Fabien's notice-order justification in #14, with those changes retained.
  • Peter Eisentrautsaw the final versioncommittercommitted the patch

Credits across 14 commits

Author
Peter EisentrautLLM7
Fabien COELHO6
Reviewer
Peter Eisentrautcommitter6
Daniel Verite2
Fabien COELHO2
Iwata, Aya2
Kyotaro Horiguchi2
vignesh C2
Committer
Peter Eisentraut14

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