I added a new option to the screenshot method to customize the color of
the box when we want to mask some elements for the screenshot.
The default color is pink `#FF00FF`, but with this new option you can
specify the color you like the most, like a nice green `#00FF00`:
```js
await page.screenshot({
mask: [page.locator('div').nth(5)],
maskColor: "#00FF00",
})
```
---------
Signed-off-by: Jasiel Guillén <darkensses@gmail.com>
The method accepts a `ratio` option to assert the ratio
of the element in viewport. `ratio` defaults to `Number.MIN_VALUE`.
NOTE: this reverts commit d950f5b6ee3fee4b825831983d5af5b197bda769 and
adds `ratio` option support + does the rename.
Fixes#8740
This patch has 2 fixes:
- screenshot code was accidentally using main page context to fetch
page layout metrics instead of a utility context
- Avoid usage of `self.eval` inside utility context since it escapes
Firefox sandbox. This turns out to be an upstream bug:
https://bugzilla.mozilla.org/show_bug.cgi?id=1814898Fixes#20434
This is a new web-first assertion that should be used like this:
```ts
test('should work', async ({ page }) => {
const locator = page.locator('body');
// New web-first assertion.
await expect(locator).toIntersectViewport();
// The same functionality.
await expect.poll(() => locator.viewportRatio()).toBeGreaterThan(0);
});
```
Fixes#8740
- Rename internal selectors `has`, `control` and `attr` to
`internal:has`, `internal:control` and `internal:attr`.
- Fix `getByLabel()` to respect strictness, by introducing
`internal:label` selector.
- Move tests essential for ports to `selectors-by.spec`.
Instead of requiring all frames in the subtree to receive a particular
event, we rely on the browser's definition of load and DOMContentLoaded.
This changes logic in a few edge cases:
- Some browsers do not emit load event upon window.stop() at all.
- DOMContentLoaded does not wait for subframes, so they might not be
ready when passing `{ waitUntil: 'domcontentloaded' }`.
`networkidle` preserves the old logic.
We lack `documentId` when doing a reload over browser protocols, so
`reload()` waits for the next navigation to finish. Sometimes, the page
might issue a same-document navigation while reload is in progress,
which confuses the reload command.
To fix the issue, just ignore same-document navigations for reload,
because it is always a new document.
Previously, when some iframe started/finished a new navigation, we
could have removed and then re-addded load/domcontentloaded on the
main frame.
Drive-by: unflake wheel test in Firefox.
