Skip to content

v0.4.0-dev fix: preserve selected-page assets and freeze real table holdouts - #9

Merged
Tom-R-Main merged 3 commits into
mainfrom
codex/real-table-holdouts
Oct 13, 2026
Merged

Tom-R-Main merged 3 commits into
mainfrom
codex/real-table-holdouts

Conversation

@Tom-R-Main

@Tom-R-Main Tom-R-Main commented Oct 13, 2026 •

Copy link
Copy Markdown
Owner

Batch debug assets used page-window ordinals instead of original PDF page indices. This fixes their identity and materialized source filtering, with regressions for noncontiguous source pages, coordinates, filenames, hashes and full-document behavior. Schema 0.19.2 identifies the output correction.

A new frozen real-document pack checks selected merged headers and continuation controls in three hash-verified government originals. Both public JSONL formats retain raw output and failures. Inputs stay in the ignored cache; CI tests the harness without fetching PDFs. Review also fixed descendant cleanup so timeout, normal exit and interruption finalize the owned process group and preserve receipts.

Validation: 62/62 Zig test steps, native/shared and WASM builds, 339 JavaScript tests, 93 Python tests, 16 existing public topology checks and 28 public reading-order checks pass. Retained source/binary hashes were rechecked before this PR. Independent review covered the 19-file implementation scope and reproduced the process-cleanup fix. Review remediation also rejects array-shaped public table bboxes in both JSONL formats while retaining list-shaped frozen truth coordinates.

Qualification: all six real-document case/format runs fail the complete table gate. The preserved baseline exposes region fragmentation, merged stacked regions, extra columns and missing merged-header structure. No thresholds or truth were relaxed; no parser tuning followed holdout assessment. Selected-cell truth and one positive-continuation template do not establish general table accuracy.

The monitored compiler-cache peak was 1.96 GiB against a 6 GiB observed budget. Cached tooling, prior verification evidence and database/container state were preserved.

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 13, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-13T18:31:21.559316Z 8578641 PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 857864123a

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread benchmark/eval/table_topology_real/verify.py Outdated
@Tom-R-Main
Tom-R-Main merged commit ab40fe5 into main Oct 13, 2026
6 checks passed
@Tom-R-Main
Tom-R-Main deleted the codex/real-table-holdouts branch October 13, 2026 18:52
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