The explicit wait after an action that navigates – a clicked link, a submitted form, a JS redirect. Paparazzi never detects navigations on its own, so a wait marks exactly where one is expected. Call it immediately after the action. It waits for the new document to finish loading and then hold still for a moment. Three cases pass:
Usage
pz_wait_for_navigation(
ctx,
...,
wait = c("auto", "load", "shiny", "none"),
timeout = NULL
)Arguments
- ctx
A paparazzi context.
- ...
Checked empty; reserved for future use.
- wait
What to wait for:
"load"settles the navigation;"shiny"also waits for Shiny idle after load."auto"uses"shiny"only when the page was opened on an app handle or app path and lands on that app's origin; otherwise it uses"load"."none"skips settling and only resets the scope.- timeout
Seconds before giving up;
NULLuses the session default.
Details
the navigation finished before the wait started: the current document differs from the document where the preceding action began (including a page restored from the back/forward cache);
the navigation is in flight when the wait starts, and is waited out;
the navigation begins while the wait is running. One scheduled beyond the timeout can't be caught – block on its trigger with
pz_wait_for_js()first.
A page where nothing navigates times out with a classed error rather than passing, and so does a second wait after the same action: a successful wait uses up the action's navigation.
On success the wait resets the scope to the root and releases every pinned scope object: contexts scoped before the navigation error on their next use instead of acting on a stale set.
Examples
page <- pz_open(pz_example("tasks"))
page |> pz_act_click(pz_loc(".task-done", within = pz_loc(".task", has_text = "bank")))
pz_get_count(page, target = ".task.done")
#> [1] 2
# "Start over" is a link to a new copy of the page
page |>
pz_act_click("#start-over") |>
pz_wait_for_navigation()
pz_get_count(page, target = ".task.done")
#> [1] 1
pz_close(page)