Fix two bugs in the toast

Fix two bugs in the toast

1- The US locale is sometimes "en" and sometimes "en-US"
2- Users that hit "don't bug me" where shown the same toast again when we ran the toast twice in a row. Now we compare the prefix of the experiment witht the 'client' prefix if equal it means we already toasted this user with this same experiment.

BUG=130169
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=147972

[cpu_(ooo_6.6-7.5), 2012-07-22 20:58:07]

please review at your earliest chance.

[Mark Larson, 2012-07-22 21:18:45]

On 2012/07/22 20:58:07, cpu wrote:
> please review at your earliest chance.

LGTM

[cpu_(ooo_6.6-7.5), 2012-07-23 22:36:08]

I checked in as-is but I still need finnur's feedback. I will create another CL
that builds on top of this so in that CL you can request changes to this as
well.

As you can see I am rushing to get this in m22

[Finnur, 2012-07-25 10:51:10]

http://codereview.chromium.org/10802063/diff/1/chrome/installer/util/google_c...
File chrome/installer/util/google_chrome_distribution.cc (right):

http://codereview.chromium.org/10802063/diff/1/chrome/installer/util/google_c...
chrome/installer/util/google_chrome_distribution.cc:761: if (base_group ==
client.substr(0, 2)) {
nit: Probably more readable to combine these into one if:

if (client.size() > 2 && base_group == client.substr(0, 2)) {
...
}

http://codereview.chromium.org/10802063/diff/1/chrome/installer/util/google_c...
chrome/installer/util/google_chrome_distribution.cc:763: return;
Are you intentionally setting no results for Omaha to pick up and report back?
In other words, shouldn't you be using SetClient with a specific results code,
like we do on lines 782, 787, 793 (to name a few)?