ContextAccess

Works out which parts of a request a route actually uses, so the rest can be skipped.

Parsing a body, a query string and path parameters costs work, and most handlers only use one of them. Before serving a route, App asks this module what the handler chain reads and then parses only that.

The answer comes from reading the handler's own source: fn.toString() gives back the function's text, and every mention of the context parameter is checked. c.body is a read of the body. ({ params }) => … is a read of the params. Anything less obvious — passing c to another function, returning it, indexing it with a variable — means the handler could reach anything, so everything gets parsed.

That fallback is the rule the whole module follows: when in doubt, parse it all. Guessing "unused" wrongly would hand a handler an empty body it was counting on. Guessing "used" wrongly just does some work nobody needed.

getContextAccess

function

function getContextAccess(
	handlers: ReadonlyArray<Function>,
	config: Optional<RouteConfig>,
): ContextAccess;

Combines every handler in a route's chain into one answer.

A property is parsed if any handler reads it, so a Middleware that needs the body still gets one even when the route handler ignores it. A key with a schema in the route's RouteConfig is always parsed — skipping it would let an invalid payload through unvalidated.

Parameters

• handlers — Every function in the chain, middleware and route handler alike.
• config — The route's config, read for its schemas.

Returns — Which properties App should populate on the Context.