Skip to content

chore: retire testBench, switch to SpreadsheetBench#8

Merged
arnav2 merged 1 commit into
mainfrom
chore/retire-testBench
May 11, 2026
Merged

chore: retire testBench, switch to SpreadsheetBench#8
arnav2 merged 1 commit into
mainfrom
chore/retire-testBench

Conversation

@arnav2
Copy link
Copy Markdown
Collaborator

@arnav2 arnav2 commented May 11, 2026

Summary

Retires the in-tree testBench/ corpus (53 files) and switches all active
docs, scripts, CI, and Makefile targets to point at the public
SpreadsheetBench v0.1 corpus (5,458 real-world xlsx, downloaded on demand
to data/corpora/, gitignored).

What this PR removes

  • testBench/ directory and all 53 bundled workbooks
  • scripts/build_testbench.py, scripts/generate_enterprise_fixtures.py
  • tests/test_testbench_roundtrip.py, tests/test_enterprise_scoring.py,
    tests/test_real_world_datasets.py, tests/test_cross_validation.py
  • static_xlsx pytest fixture
  • testbench CI job + testbench-* Makefile targets
  • testBench-vX.Y.Z.zip release asset in release.yml

What this PR updates

  • README + wiki: point at SpreadsheetBench, make bench-robust/bench-retrieval
  • docs/corpora.md: rewritten around SpreadsheetBench
  • CONTRIBUTING.md, docs/MAINTAINERS.md, docs/RELEASE_PROCESS.md
  • Issue templates + PR template
  • site/index.html landing page (testBench section → SpreadsheetBench)
  • examples/demo.py + examples/generate_examples.py now use
    examples/fixtures/ instead of the removed testBench/real_world/
  • CHANGELOG.md: new Unreleased section flagging the breaking change

Test plan

  • make test → 1041 passed, 11 deselected (~16s)
  • CI green on this PR
  • Manual: verify the SpreadsheetBench-vs-Docling badge / README links still resolve
  • Manual: skim wiki preview to confirm no broken testBench/ links

…heetBench

Removes the 53-file in-tree testBench corpus and switches all active docs,
scripts, CI, and Makefile targets to point at SpreadsheetBench (downloaded
on demand to data/corpora/, gitignored).

Removed: testBench/, scripts/build_testbench.py,
scripts/generate_enterprise_fixtures.py, tests/test_testbench_roundtrip.py,
tests/test_enterprise_scoring.py, tests/test_real_world_datasets.py,
tests/test_cross_validation.py, static_xlsx fixture, testbench CI job,
testbench-* Makefile targets, testBench-vX.Y.Z.zip release asset.

Updated: README, wiki, CONTRIBUTING, MAINTAINERS, RELEASE_PROCESS, corpora.md,
issue + PR templates, site landing page, benchmark harness docs, examples
(redirect output to examples/fixtures/), CHANGELOG.

All 1041 tests pass.
@arnav2 arnav2 force-pushed the chore/retire-testBench branch from 4b65ee9 to 9573625 Compare May 11, 2026 21:18
@arnav2 arnav2 merged commit 80eebe6 into main May 11, 2026
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant