Data Convert

Take a clean-enough source and re-express it in the structure or contract a downstream system requires. The interoperability counterpart to data-tidy: tidy = quality (messy → clean, same-ish shape); convert = interoperability (clean → a different structure/contract). If the source is messy, convert delegates the cleaning to data-tidy first, then converts the clean result.

Two jobs, often combined:

1. Contract mapping — map a source onto another system's import contract (columns, order, types, required fields). Unmapped sources are reported; required-but-missing rows follow rules.on_missing_required (flag / exclude / error) — never invent values.
2. Structural reshape — unpivot (wide→long), pivot (long→wide), flatten/nest (JSON↔table), split (one file→many by a key), union (many→one with column alignment).

Workflow