[dfrankland]

Stacked on #1236. Refs TRUNK-22402.

The validate summary is now built once as data, then rendered, the same way #1236 does it for upload. No output changes, and no JSON flag yet.

What changes.

• New cli/src/report/validate.rs. ValidateReport::new works out everything the validate summary says: the matched files, one FileValidation per file, the CODEOWNERS check, the counts and the exit code. The types derive Serialize, with camelCase keys.
• impl EndOutput for ValidateRunResult moves into report/text.rs as render_validate, which reads only the report. The impl stays on ValidateRunResult (built through the report) because the lock-down tests call it.
• main.rs exits with report.exit_code; ValidateRunResult::exit_code is gone.
• validate and upload's validation section now read the same FileValidation. It gains testSuiteCount and testCaseCount, null for a file that wasn't read as exactly one report. Those files are the ones the text lists under "File parse issues".
• FileValidation.parseError is now {error, causes}, like ErrorSummary, instead of a {:#} string. validate has always printed parse errors the way anyhow's {:?} does, and that can't be rebuilt from the joined string. A unit test checks the rebuilt text against anyhow's own output.
• run_validate collects the --hide-banner deprecation and the BEP parse errors as notices, like run_upload. The deprecation is still only logged.

Kept as-is.

• "📚 Validating N files" counts a file once per glob that matches it, while the closing line counts each file once. counts.files is the deduplicated count.
• upload and validate still collect per-file results differently: upload has no entry for a file with no reports and keeps the last of several, where validate reports both as issues.

#1246 is stacked on this. Its JSON test and doc comments are updated for parseError as an object and the two count fields.

Testing.

• The lock-down tests in cli/tests/summary/ pass unchanged.
• cargo test -p trunk-analytics-cli --lib --test summary --test test --test validate: all pass.
• cargo test -p trunk-analytics-cli --test upload: all pass except the three Swift xunit tests, which fail the same way on the base branch locally.
• cargo clippy -p trunk-analytics-cli --tests: no new warnings. cargo fmt --all --check: clean.

🤖 Generated with Claude Code

[codecov-commenter]

Codecov Report

❌ Patch coverage is 93.44262% with 20 lines in your changes missing coverage. Please review.
✅ Project coverage is 84.85%. Comparing base (f907dd1) to head (bcaa633).

Files with missing lines	Patch %	Lines
cli/src/report/text.rs	90.86%	17 Missing ⚠️
cli/src/validate_command.rs	57.14%	3 Missing ⚠️

Additional details and impacted files

@@                                              Coverage Diff                                               @@
##           dylan/trunk-22401-build-the-upload-summary-as-data-separate-from-rendering    #1254      +/-   ##
==============================================================================================================
+ Coverage                                                                       84.74%   84.85%   +0.10%     
==============================================================================================================
  Files                                                                              78       79       +1     
  Lines                                                                           18058    18144      +86     
==============================================================================================================
+ Hits                                                                            15304    15396      +92     
+ Misses                                                                           2754     2748       -6     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[trunk-staging-io]

   

Failed Test	Failure Summary	Logs
pending_quarantine_test should be quarantined when run with variant	A test marked as pending was expected to fail but unexpectedly passed.	Logs ↗︎
variant_quarantine_test should be quarantined when run with variant	The test failed because it expected 2 + 2 to equal 5, but the result was 4.	Logs ↗︎

View Full Report ↗︎ ⋅ Docs

[trunk-io]

   

View Full Report ↗︎ ⋅ Docs

[TylerJang27]

report/test_report is a bit of an overloaded term in this repo. Can we either:

1. Rename the rspec test_report stuff to gem or integration or something
2. and/or rename this cli/src/report to cli/src/summary_output or something