webdriver
webdriver copied to clipboard
w3c
Remote control interface that enables introspection and control of user agents.
[GitHub will no longer support the Ubuntu 20.04 runner image starting from April 1, 2025](https://github.com/actions/runner-images/issues/11101). That PR updates the actions to use ubuntu-latest instead
In the "process capabilities" algorithm, steps 2.2 and 5.1 provide a "flags" parameter to the "validate capabilities" algorithm: ``` 2.2. Let required capabilities be the result of trying to validate...
We've run into a couple issues implementing https://github.com/w3c/webdriver/commit/7d8beb949159e0f7d86a34410aedd3a1af3d4b0c in the Ladybird browser. 1. Release Actions now enqueues a token to the input state's actions queue in step 5 ("Wait for...
While crawling [WebDriver](https://w3c.github.io/webdriver/), the following algorithms fire an event, or resolve or reject a Promise, within a step that runs [in parallel](https://html.spec.whatwg.org/multipage/infrastructure.html#in-parallel) without first queuing a task: * [ ]...
https://w3c.github.io/webdriver/#dfn-url-prefix defines "URL prefix" to be (ECMAScript's) undefined or a path-absolute URL (an Infra string). Then in [match a request](https://w3c.github.io/webdriver/#dfn-match-a-request) we look at the "concatenation of the URL prefix and...
As discussed in https://github.com/w3c/webdriver-bidi/issues/795, the actions should be dispatched from the top-level browsing context. *** Preview | Diff
We have tests in WPT that assert a non-zero pressure when pointerdown gets fired for touch/pen input, but nothing actually defines how much pressure they should have; it seems sensible...
The WebDriver specification expresses the following: > A [session](https://w3c.github.io/webdriver/#dfn-sessions) has an associated browsing context input state map, which is a [weak map](https://w3c.github.io/webdriver/#dfn-weak-map) with [top-level browsing contexts](https://html.spec.whatwg.org/multipage/document-sequences.html#top-level-browsing-context) as keys, and [input...
As noted in [this GitHub comment](https://github.com/w3c/webdriver/pull/1802/files#r1816671281), we introduced a regression in our `wait for navigation to complete` algorithm, which causes timeouts to be mishandled when a user prompt is opened....
Given a user defines a frame via: ```ts ``` Interactions with elements within that frame fail, e.g. don't happen. I assume it is because the driver fails to properly calculate...
