Field schema

The field schema decides what gets pulled out of every document in a collection. Five endpoints read and change it.

Not written yet

This page is a placeholder. The site's build and hosting are in place; the prose is not. What belongs here:

• The difference between reading the schema and reading one document's extracted values (the latter is extracted fields)
• When to use each of the five calls - add, update-or-add, replace, remove, read - because four of them overlap and the wrong one silently discards fields
• What happens to documents already indexed when the schema changes: are they re-extracted, or only new ones?
• That PUT replaces the whole custom set, which is the destructive one