Record a missing method field without borrowing from a nearby row

A blank does not always mean repeat the previous method

Keep the gap visible so the next user knows which comparison still needs supporting evidence.

Our public material-register example shows a synthetic report beside downloadable output. Use it to inspect how values and source references travel together. The self-service ParseMill trial is for invoice PDFs; supplier test reports and engineering review need a separate fit discussion.

A correct result can become misleading if paired with an assumed method. Keep the distinction between explicitly reported in cell, confirmed shared context and method not stated. This lets a later reviewer assess comparability without mistaking your inference for source text.

Use a blank-context fixture

The example is invented and includes a deliberately unsupported blank. S3 must not inherit Method A merely because it follows S2. The ditto interpretation in S4 is valid only because the fixture supplies a defining legend; real documents may use marks differently.

Microsoft: Document layout analysis includes row and column spans in table output. Span information can help identify merged-cell context, but verify it against the rendered source before assigning a method.

Review the missing field

1. Inspect the cell, heading and surrounding table boundaries. Determine whether the source is blank, merged, marked or unreadable.
2. Search for an explicit shared-method note or legend and record its location and scope. Keep ambiguous scope unresolved.
3. Assign a method only when evidence supports the affected observation. Store the assignment basis separately from the raw cell text.
4. Check records at merged-cell and section boundaries. Preserve an unresolved method state where evidence remains insufficient and identify the clarification needed.

Microsoft: M language values distinguishes absent or indeterminate information from an ordinary value. Use a missing-state field rather than silently manufacturing a method string to avoid a blank.

Avoid inference from similar tests

Two adjacent rows with the same test name can use different methods or revisions. A familiar laboratory practice does not establish what this source used. Do not fill the blank from another supplier report or from a previous month’s export without explicit supporting evidence.

If a clarification later confirms the method, record who supplied it, which source revision it concerns and which observations it covers. Keep the earlier unstated state in the change history.

Check the resulting context

The fixture passes when S1 is explicit, S2 is context-confirmed, S3 remains unstated and S4 carries its legend-supported assignment. The recipient can inspect the basis for each populated method field. That distinction is more useful than a column with no visible blanks.

Include merged and genuinely blank method cells in a supplier-report scope discussion. Agree how unsupported assignments would be kept visible and how the source would be reviewed. Do not infer this specialist context handling from the self-service invoice offer.