What is a mapping?

A mapping is the bridge between a data source and a schema. It tells the system: for each row in the source, produce a record with these schema fields filled in from those source columns.

Why a separate step?

Source columns rarely match schema fields exactly:

• A source column is called Full Name, a schema field is name.
• A source has first and last columns; a schema wants them concatenated.
• A source has a URL in photo_url; a schema wants a clean asset path.
• A source contains extra columns the template does not need.

The mapping captures these decisions once so every dataset derived from the same source uses the same logic.

What a mapping contains

• A data source to read from.
• A schema to produce.
• Per schema field, a source expression: a source column, a constant, a concatenation, a lookup, or a computed value.
• Optional filters — which source rows to include at all.

Multiple mappings for one source

A single source can have many mappings — for example one that filters to "home team" and another that filters to "away team". Each produces a different dataset.

Next

• Read Building a mapping to create your first one.