Exposing NewAllocator and AllocatorOption was unnecessary, and it made
the API more complex to use and understand.
Instead, have users call NewExecAllocator directly. This removes some
code, and simplifies the examples and tests.
This way, the simple examples and tests don't need to do that
separately. Practically all users will want this cleanup work to be
synchronous, and practically all Go APIs are synchronous by default, so
this makes chromedp easier to use.
This can simplify some common use cases, like running a few actions
directly, or running no actions at all. It's also an almost entirely
backwards compatible change, as all Run call sites should continue to
compile and work.
Leave Tasks, as it can still be useful for functions to return complex
sequences of actions as a single Action.
That way, we avoid the racy map access via Browser.executorForTarget. If
a context is attached to a target, the Target field must be non-nil.
The Browser.pages map is still racy, since multiple tabs can be created
concurrently; we'll fix this other data race in another commit.
Using a smaller viewport speeds up both tests, and lets us know what
dimensions to expect in TestCaptureScreenshot.
For TestScreenshot, we can know what dimensions to expect in advance, as
we have the images in testdata.
'go test -run Screenshot' goes from ~0.9s to ~0.5s on my machine.
Finally, don't run ExampleTitle as part of 'go test', as it's slow.
The Default folder is created asynchronously to Chrome starting to
listen on the debugging protocol port. So we can't expect it to exist.
Instead, base the example on DevToolsActivePort, which we can rely on.
All it did was wait on the entire allocator, which is confusing. From
the user's perspective, this wait method should instead wait for the
resources for its own browser, and not any other browsers sharing the
same allocator.
We haven't decided how to integrate that into our API, so simply replace
it with Allocator.Wait.
