Beadloom

Appearance

✅ fresh

last synced 2026-09-29T21:12:19.681221+00:00 · coverage 100% (markdown-tables)

Validation by Beadloom doc_sync — same source as sync-check.

Markdown Tables (component) ​

Internal building block of the doc-sync domain.

Source: src/beadloom/doc_sync/tables.py

Overview ​

What a markdown table row is, and where one table ends and the next begins. One module, because three readers of that answer needed it and the first two disagreed twice in one slice.

BDL-UX #213 and BDL-UX #244 are one sentence found in two places: a section holding two tables was read as one, so the second table's rows were judged against the FIRST table's column index and the second table's header row came back as data. In doc-quality that reported a measurement table's header as a decision row with a missing reason cell — four findings on BDL-067's ACTIVE.md, all false. In axes-section it produced an approved node literally named Node. Measured on this repository's own RFC laid out in the shape its own rule describes — 74 rows in five per-slice tables — the axes reader returned 78 rows, four of them header rows, all four approved.

Both second tables exist because this project's own document rules ask for them. The RFC says each slice appends its axis rows under its own Derived by line, and the /coordinator playbook asks for a verification table beside the decision table it verifies. Neither reader was wrong about one table; both were wrong that there is one.

BDL-UX #259 is the THIRD reader, work-item-routing, converted by beadloom-0mdo.77. It matched the first ROW that spelled the routing header and read every table row below it as a route, and it fired no instance because it discarded a row whose second cell named neither simplified nor full. That is the shape this component exists to remove: a vocabulary guard standing in for a boundary, in the reader whose sibling had already measured that vocabulary cannot decide this. Measured over this repository, the guard admits 27 of 5 122 table rows in 456 markdown files, and 17 of the 27 belong to other tables in other documents.

Public surface ​

  • cells_of(line) — the cells of a table line, alignment row included, or None when the line is not a table row. The raw reading, for a caller that asks whether a line is a row at all.
  • is_separator(cells) — whether those cells are an alignment row rather than content.
  • table_cells(line) — the cells of a table row that says something, or None. Re-exported from doc_sync.doc_shape, where it used to live, so no caller moved.
  • table_blocks(lines) — the contiguous tables in numbered lines, each leading with its own header row. A separator row is dropped and does not end a table; everything that is not a table row does.
  • Table — one table, as (line number, cells) rows.

Collaborators ​

doc-quality spends table_blocks for decision-reason, risk-mitigation and pending-in-approved, and cells_of for unfilled-placeholder. axes-section spends it to read the ## Axes section's per-slice tables. work-item-routing spends it to find the /task-init routing tables — every block whose own header row leads with Type and Flow — and reads their union, each table's rows judged against its own header. wave-plan spends cells_of.

doc-shape still re-exports table_cells, and since beadloom-0mdo.77 moved work-item-routing onto table_blocks that re-export has no caller in this repository. It is kept as the documented alias its comment describes rather than removed here, because removing a public name is a surface change and not this bead's subject.

The boundary rule reads no header vocabulary at all. The residual class that a boundary cannot decide — a measurement table carrying a reason column of its own — is doc-quality's, and is answered there by asking what the DOCUMENT declares.

Component doc (BDL-068 S6, beadloom-0mdo.46; third reader added by beadloom-0mdo.77). Public surface verified against tables.py.