CivicSpleen / ambry

A comprehensive data package manager
BSD 2-Clause "Simplified" License
4 stars 5 forks source link

Stronger separation of ingestion phase #155

Closed ericbusboom closed 8 years ago

ericbusboom commented 8 years ago

Create a stronger separation of the ingestion phase.

Don't auto create source tables. Create them only when the wranger requests it. The source tables are always loaded in from config.

So, the wrangler runs an ingest, then exports the source tables, and edits them. Then they are synced in.

When a source is run, if a source table exists, a new pipe segment is created to map the source headers and column types.

This change would also make it easier to load sources directly into build pipelines.