Make it possible to run a test in sky_shell

Make it possible to run a test in sky_shell

This CL makes it possible to run sky/tests/raw/bogus_import.dart in sky_shell
on Linux. We'd eventually like to scale this up to running all the tests.

R=eseidel@chromium.org

Committed: https://chromium.googlesource.com/external/mojo/+/c7a7d2c08ea8ae1f5107e320b11493020a0d6c27

[eseidel, 2015-06-29 16:53:59]

lgtm

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
File sky/shell/testing/test_runner.cc (right):

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
sky/shell/testing/test_runner.cc:26: void WaitForURL(UrlData& data) {
This code is banana pants.  file://path/to/test?hash=HASH would made slightly
more sense (although admittedly not a huge amount).

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
sky/shell/testing/test_runner.cc:59:
viewport_observer_->OnViewportMetricsChanged(320, 640, 1.0);
How did we standardize on sub-VGA?

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
sky/shell/testing/test_runner.cc:73: std::cout << "#READY\n";
You're gonna want a helper for all these flush-every-line commands.  Or maybe
you can set cout into a flush-on-newline mode?

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
sky/shell/testing/test_runner.cc:82: std::cout << "#EOF\n";
Comments here as to what goes between the various #EOFs would be nice.

[abarth-chromium, 2015-06-29 17:39:48]

Committed patchset #3 (id:40001) manually as
c7a7d2c08ea8ae1f5107e320b11493020a0d6c27 (presubmit successful).

[abarth-chromium, 2015-06-30 21:38:31]

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
File sky/shell/testing/test_runner.cc (right):

https://codereview.chromium.org/1218633003/diff/20001/sky/shell/testing/test_...
sky/shell/testing/test_runner.cc:59:
viewport_observer_->OnViewportMetricsChanged(320, 640, 1.0);
On 2015/06/29 at 16:53:59, eseidel wrote:
> How did we standardize on sub-VGA?

We can change this to be whatever you like.  This value matches the value in the
existing test runner.