3. Terminology
Permalink to 3. TerminologyThis section is normative.
Use case. A description of a single, goal-oriented workflow, identified by
an id, with a title, a start_location, an expected_result, and an
ordered sequence of steps.
Step. A single instruction within a use case. Every step has a keyword (see §5.1) and zero or more arguments and modifiers.
Target. The element a step acts upon. A target is identified by a role token (see §5.4) and an accessible name, and optionally by a scope (see §5.5).
Accessible name. As defined in [WAI-ARIA] and the [HTML AAM]: the string
exposed to assistive technology as the name of the element. UCDL targets match
against the accessible name as computed by the user agent and exposed through
Playwright's getByRole({ name }), getByLabel(), or getByText() queries.
Effective data map. The merged map produced by combining: external data
files (see §12.2) ⊂ parent data (for extension cases) ⊂ own
data ⊂ data overrides supplied at runtime (e.g., --set on the CLI). Later
sources override earlier sources.
Resolved use case. A use case whose extends chain has been fully followed,
whose data block has been merged, whose template variables have been
interpolated, and whose steps_override (if any) has been applied. All
references to "use case" in §§5–11 refer to
the resolved form unless stated otherwise.
Processor. A program that consumes a use case document and produces a parse tree, a Playwright artifact, an execution result, or a report.
User agent. The browser instance under test, controlled via Playwright.
Engine. A separate scanner — specifically @afixt/afixt-engine — invoked by
the audit keyword.