This is an automated email from the ASF dual-hosted git repository.

github-merge-queue[bot] pushed a commit to branch 
gh-readonly-queue/main/pr-7022-a61702fd10ccde130f0a24a1298adff24c4a38a5
in repository https://gitbox.apache.org/repos/asf/texera.git

commit 729e3d8bb53e86c37e13ba249059250015b71f8b
Author: Yicong Huang <[email protected]>
AuthorDate: Wed Jul 29 07:35:27 2026 -0400

    fix(ci): rename backport auto-label output so its report step runs (#7022)
    
    ### What changes were proposed in this PR?
    
    The `Report backport decisions` step in `backport-auto-label.yml` is
    skipped on every run, so the report comment and release-manager review
    request added in #6962 never fire.
    
    Root cause: the step's guard reads `steps.label.outputs.result`.
    `actions/github-script` always writes the script's return value to an
    output named `result` after the body runs. The `Label fix PRs` script
    has no `return`, so that post-run write blanks out the explicit
    `core.setOutput("result", …)` — the guard is therefore always false.
    
    Fix: rename the output to `decisions` (github-script only reserves
    `result`), and update the guard and the `RESULT` env reference to match.
    No logic in either step changes.
    
    ### Any related issues, documentation, discussions?
    
    Closes #7021.
    
    ### How was this PR tested?
    
    - Confirmed the failure on run `30423275226` (PR #7013): the `Report
    backport decisions` step shows `skipped`.
    - Verified 0 `<!-- backport-auto-label-report -->` comments across
    #7013, #6983, #6958, #6908 (including PRs that were auto-labeled),
    confirming the step is systemically skipped.
    - Change is name-only; the report/review logic is untouched, so it runs
    unchanged once the guard sees the output.
    
    ### Was this PR authored or co-authored using generative AI tooling?
    
    Generated-by: Claude Code (Claude Opus 4.8)
    
    Co-authored-by: Claude Opus 4.8 (1M context) <[email protected]>
---
 .github/workflows/backport-auto-label.yml | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/backport-auto-label.yml 
b/.github/workflows/backport-auto-label.yml
index e99a93abee..c2ee450100 100644
--- a/.github/workflows/backport-auto-label.yml
+++ b/.github/workflows/backport-auto-label.yml
@@ -284,9 +284,13 @@ jobs:
 
             // Only hand work to the report step when there is something to 
say;
             // a "none" outcome leaves the output unset so that step is 
skipped.
+            // The output is named `decisions`, not `result`: 
actions/github-script
+            // always writes the script's return value to an output called 
`result`
+            // after the body runs, which would clobber an explicit setOutput 
of the
+            // same name (this script returns nothing, so that would blank it 
out).
             const result = await decide();
             if (result.mode !== "none") {
-              core.setOutput("result", JSON.stringify(result));
+              core.setOutput("decisions", JSON.stringify(result));
             }
 
       # Report step: runs under the default GITHUB_TOKEN so the review request
@@ -294,10 +298,10 @@ jobs:
       # retrigger comment-/review-driven workflows. It only touches the PR when
       # the label step handed it something to say.
       - name: Report backport decisions
-        if: ${{ steps.label.outputs.result }}
+        if: ${{ steps.label.outputs.decisions }}
         uses: actions/github-script@v9
         env:
-          RESULT: ${{ steps.label.outputs.result }}
+          RESULT: ${{ steps.label.outputs.decisions }}
         with:
           github-token: ${{ secrets.GITHUB_TOKEN }}
           script: |

Reply via email to