Repository navigation
feat: add failed_summary option to list failed tests with logs - #1622
Conversation
There was a problem hiding this comment.
Copilot review overview
π‘ Changes recommended
Failed-test grouping, cross-platform paths, and unbounded log payloads can produce incorrect output or publishing failures.
Review effort: Balanced
Findings: 1
Open (4)
What changed in this PR
Adds configurable failed-test details to summaries, comments, and outputs.
Changes:
- Adds
failed_summaryandfailed_summary_logsinputs. - Renders escaped failed-test details and logs.
- Adds table tests and rebuilds the distribution bundle.
| File | Description |
|---|---|
src/βutils.ts |
Adds HTML escaping. |
src/βtable.ts |
Builds failed-test sections. |
src/βmain.ts |
Reads inputs and publishes sections. |
src/βannotator.ts |
Adds sections to summaries/comments. |
README.md |
Documents new inputs. |
action.yml |
Declares new action inputs. |
__tests__/βtable.test.ts |
Tests failed-summary rendering. |
dist/βindex.js |
Updates bundled action code. |
π‘ Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
There was a problem hiding this comment.
Copilot review overview
π‘ Changes recommended
Detail cleanup can discard valid rows, and comment fitting can still exceed GitHubβs size limit.
Review effort: Balanced
Findings: 5
Open (6)
Truncation can exceed maxLength and select partial HTML Β· New Strip configured path prefixes from test titles Preserve no-annotations diagnostic row Collect nested annotations for inline failure reporting Prevent oversized truncation notice from exceeding comment length limit Correctly name default failed-summary log test
|
Oh. the closing of this PR was not intentional. |
ddea25c to
d9fa2e4
Compare
|
Thanks for managing so quickly! really appreciate! |
There was a problem hiding this comment.
Copilot review overview
π‘ Changes recommended
Windows paths are rendered incorrectly, and log rendering performs unbounded intermediate allocations.
Review effort: Balanced
Findings: 1
Open (5)
Apply output cap before materializing large failure messages Β· New Normalize Windows path prefixes before stripping test titles Downgrade log blocks before truncating entries Suppress missing-annotation warning after rendering nested failures Update detailed_summary documentation for details blocks
Yeah copilot loves to do endless loops of feedback :D Thank you very much for the PR and the proposed new feature |
There was a problem hiding this comment.
Copilot review overview
π‘ Changes recommended
Windows-style titles remain duplicated, public prerequisites are incomplete, and the committed source map is stale.
Review effort: Balanced
Findings: 1
Open (7)
Apply output cap before materializing large failure messages Normalize Windows path prefixes before stripping test titles Downgrade log blocks before truncating entries Suppress missing-annotation warning after rendering nested failures Document detailed and failed summary prerequisites Β· New Document both prerequisites for failed summary logs Β· New Update detailed_summary documentation for details blocks
Preserve existing summary defaults and check context. Group testcase failures and fit titles before logs within each destination limit. Co-authored-by: Ludovic DEHON <tchiot.ludo@gmail.com>
10ad22b to
767bd31
Compare
|
You are fast !! β© |
Yes, can't say when exactly, but maybe even today. |
|
thanks for the really fast release, btw ! |
|
You are welcome! thanks again for the PR |



Adds optional failed-test logs to the detailed summary. Existing configurations keep failed tests in the table by default.
failed_summary, defaulttrue, controls whether failed tests appear in the detailed summary. Requiresdetailed_summary.failed_summary_logs, defaultfalse, moves failed tests from table rows into collapsible log blocks. Requiresdetailed_summaryandfailed_summary. Durations followinclude_time_in_summary.Applies to the job summary, PR comment, and
detailed_summaryHTML output. Multiple failures of one testcase share a block, and each block retains its check name. The failure section is capped at 50,000 characters and respects the remaining PR comment space. Titles take priority over logs; logs are rendered within the remaining budget before HTML escaping. Nested failures and Windows-qualified titles are covered.Validation: 102 tests passed, build, lint, changed-file formatting, packaging, and
git diff --check. Compared 8,192 default configurations against main and verified byte-identical default comment and job-summary output. Regression coverage includes oversized logs, narrow comment budgets, 10,000 hidden-entry counts, Windows paths, nested-only failures, and identical failures from different checks.