Engine terminology

A few tool descriptions name concepts from the underlying Import.io scraping engine so that behaviour is described exactly. This page defines each term in public behaviour terms; nothing here is required knowledge for the high-level helpers.

A hosted browser session that executes navigations, renders JavaScript, and returns page content. Every importio_* tool is a wrapper over this engine; you never talk to it directly except through importio_call_method.

The options object the engine accepts for navigation and rendering: url, wait, timeout, and output switches (html, png, pdf, force200). importio_goto, importio_render, and importio_get_html pass this object through; importio_web_fetch accepts the common fields (url, wait, wait_for_selector, timeout_ms) as plain arguments instead.

A declarative extraction definition (fields and selectors) the engine applies to a rendered page. Pass one to importio_web_fetch or importio_extract_data when you already know the data shape; discover shapes with importio_get_html or importio_suggest_data first.

Browser interactions (click, input, keyboard, wait-for-selector, pagination, and friends) are engine actions with a constructorName and a fields payload. The importio_action_* tools fill in the constructor name for you; importio_play_action dispatches a raw action envelope when you already know the upstream JSON shape.

When a site serves a block or unexpected status, the high-level URL tools may retry in a fresh browser session with managed proxy selection (country and type selectable, residential only when explicitly enabled). Retries are bounded; repeated retries of an already-blocked page do not help.