PDF Manual Mapping Workflow

Use Build Your Own Map when automatic mapping is not reliable (common with complex XFA structures).

Manual mapping is the third mapping mode in the PDF pipeline:

  1. Deterministic mode
  2. AI mode
  3. Manual mapping flow (this guide)

For complex XFA forms, this flow is the final best-effort step to complete mappings that could not be resolved automatically.

Where to open it in the add-in PDF, Word, PPT → select a PDF file → Edit Fields Mapping → Build Your Own Map

Stage Overview

  1. Stage 1 - Select Parent and Child Tables
    • Select exactly one parent table.
    • Keep related tables checked as child tables.
    • Uncheck unrelated tables to ignore them.
  2. Stage 2 - Parent Column Mapping
    • Map parent Excel headers to PDF fields.
  3. Stage 3 - Child Table Mapping
    • Map child table headers.
  4. Stage 4 - Review and Apply
    • Review the full mapping and apply it.

Stage 1 Validation Rules

Manual mapping enforces these rules before you can continue:

  • Exactly one parent table is required.
  • At least one used table is required.
  • Only used tables must contain an ID column.
  • Child tables without ID are auto-ignored and cannot be selected.
  • Used tables must have unique header names (duplicate headers block progression).

Mapping Rules

  • PDF field selections are globally unique across parent and child mapping grids.
  • If you select a PDF field in one row, it becomes unavailable in other rows.
  • Server-side validation also blocks duplicate PDF target selections at apply time.

After Apply

  • Mapping is applied in manual source mode.
  • The edit tree refreshes from your manual mapping payload.
  • Hover tooltips show diagnostics for both Acro and XFA nodes (with XFA fallback metadata when needed).

Tips

  • Start with the cleanest parent table (unique headers + ID).
  • Ignore unrelated tables early to reduce noise in Stage 2/3.
  • Map required PDF fields first, then optional fields.