Retrieve bounded pages through same-origin next links

Usage

paginated_links(
  request,
  items,
  next_link = function(response) httr2::resp_link_url(response, "next"),
  max_pages = 100,
  max_items = 10000,
  warn_limits = FALSE,
  error_policy = "stop",
  format = NULL,
  policy = list()
)

Arguments

request: An httr2 GET request for the first page. Configure credentials, query parameters, timeout, and retries on this request beforehand.
items: Function extracting a collection from an httr2 response.
next_link: Function extracting the next URL from an httr2 response. Defaults to the HTTP Link header's next relation. NULL or an empty string means completion. Relative links resolve against the current response URL.
max_pages, max_items: Finite positive retrieval limits, as in paginated().
warn_limits, error_policy, format, policy: As in paginated(). Policy is passed to iteration; link security checks still apply before every request.

Value

The same structure as paginated(), with no_next_link for completion without another link. Request, extraction, repeated-link, redirect, and cross-origin errors abort by default. Selected error recovery applies only to request failures; redirects and unsafe links always abort.

Details

Requires httr2 1.3.0 or later. Only GET requests without a body are supported. Every link must use the initial request's scheme, hostname, and effective port. URL credentials and fragments are rejected. Redirects are disabled, including same-origin redirects. Headers and authentication remain on the original request; query arguments come from each next URL. Single- request generated functions and their decoded response types are unchanged.