[sync] Add java InvalidationController for Android.

[sync] Add java InvalidationController for Android.

* Adds InvalidationController to facilitate the move to the new
  invalidation client, but still supporting the old way.
* Sets up the chromium test shell to be able to run the new tests for
  sync.
* Upstreaming a few base test utilities for MockContext and
  SharedPreferences.

InvalidationController and InvalidationControllerTest provided by
dsmyers@chromium.org.

BUG=159221
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=176533

[nyquist, 2013-01-10 18:39:39]

yfriedman: All parts
dsmyers: sync/notifier parts

[Yaron, 2013-01-10 19:29:40]

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
File
base/test/android/javatests/src/org/chromium/base/test/util/AdvancedMockContext.java
(right):

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
base/test/android/javatests/src/org/chromium/base/test/util/AdvancedMockContext.java:18:
* Simple ContextWrapper that adds functionality for SharedPreferences and a
simple
Nit: extra "simple" is unneeded :)

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
File
base/test/android/javatests/src/org/chromium/base/test/util/InMemorySharedPreferences.java
(right):

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
base/test/android/javatests/src/org/chromium/base/test/util/InMemorySharedPreferences.java:14:
* An implementation of SharedPreferences that can be used in tests.
Why are these two utils now in base? We've never used them outside of sync.

https://codereview.chromium.org/11778084/diff/1010/chrome/chrome_tests.gypi
File chrome/chrome_tests.gypi (right):

https://codereview.chromium.org/11778084/diff/1010/chrome/chrome_tests.gypi#n...
chrome/chrome_tests.gypi:2815: '../sync/sync.gyp:sync_javatests',
It seems odd to have sync_javatests roll up into the test apk but not have the
binary depend on the actual code. Also, nobody currently depends on sync_java
aside from this test. I suspect chrome_java will eventually and I guess we can
add it then. wdyt?

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
File sync/android/java/src/org/chromium/sync/ModelType.java (right):

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/ModelType.java:1: // Copyright (c) 2011
The Chromium Authors. All rights reserved.
I suppose this should be 2013.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/ModelType.java:15: // A bookmark folder
or a bookmark URL object.
Nit: use javadoc for IDEs?

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/ModelType.java:19: // An object
representing a browser session.
.. browser session or tab.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
File
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java
(right):

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:39:
/**
Nit: missing ws above

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:91:
private static final String IMPL_CLASS_MANIFEST_PROPERTY =
nit: IMPLEMENTING_CLASS_NAME? Seems like the brevity is hurting.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:160:
@Nullable ApplicationInfo appInfo;
We don't really make use of @Nullable elsewhere so it seems to be stylistically
at odds. Also, isn't it more useful as a formal argument? Here you're not
instantiating it so it's obvious.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:165:
@Nullable String className = null;
same

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:176:
Log.e(TAG, "Cannot read own application info", exception);
THis is also a wtf moment?

[nyquist, 2013-01-10 22:01:42]

addressed all comments.
yfriedman: PTAL

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
File
base/test/android/javatests/src/org/chromium/base/test/util/AdvancedMockContext.java
(right):

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
base/test/android/javatests/src/org/chromium/base/test/util/AdvancedMockContext.java:18:
* Simple ContextWrapper that adds functionality for SharedPreferences and a
simple
On 2013/01/10 19:29:40, Yaron wrote:
> Nit: extra "simple" is unneeded :)

Done.

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
File
base/test/android/javatests/src/org/chromium/base/test/util/InMemorySharedPreferences.java
(right):

https://codereview.chromium.org/11778084/diff/1010/base/test/android/javatest...
base/test/android/javatests/src/org/chromium/base/test/util/InMemorySharedPreferences.java:14:
* An implementation of SharedPreferences that can be used in tests.
On 2013/01/10 19:29:40, Yaron wrote:
> Why are these two utils now in base? We've never used them outside of sync.

It is used by omaha, crash uploading, identify (unique IDs), Chrome to Mobile,
preferences (connectivity tests), and the exponential backoff scheduler in
addition to sync, so I felt that sync/ would be a bad place to put it in?

https://codereview.chromium.org/11778084/diff/1010/chrome/chrome_tests.gypi
File chrome/chrome_tests.gypi (right):

https://codereview.chromium.org/11778084/diff/1010/chrome/chrome_tests.gypi#n...
chrome/chrome_tests.gypi:2815: '../sync/sync.gyp:sync_javatests',
On 2013/01/10 19:29:40, Yaron wrote:
> It seems odd to have sync_javatests roll up into the test apk but not have the
> binary depend on the actual code. Also, nobody currently depends on sync_java
> aside from this test. I suspect chrome_java will eventually and I guess we can
> add it then. wdyt?

Added it to chrome_java now.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
File sync/android/java/src/org/chromium/sync/ModelType.java (right):

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/ModelType.java:1: // Copyright (c) 2011
The Chromium Authors. All rights reserved.
On 2013/01/10 19:29:40, Yaron wrote:
> I suppose this should be 2013.

Done here and in all the new files.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/ModelType.java:15: // A bookmark folder
or a bookmark URL object.
On 2013/01/10 19:29:40, Yaron wrote:
> Nit: use javadoc for IDEs?

Done.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/ModelType.java:19: // An object
representing a browser session.
On 2013/01/10 19:29:40, Yaron wrote:
> .. browser session or tab.

Done.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
File
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java
(right):

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:39:
/**
On 2013/01/10 19:29:40, Yaron wrote:
> Nit: missing ws above

Done.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:91:
private static final String IMPL_CLASS_MANIFEST_PROPERTY =
On 2013/01/10 19:29:40, Yaron wrote:
> nit: IMPLEMENTING_CLASS_NAME? Seems like the brevity is hurting.

Done.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:160:
@Nullable ApplicationInfo appInfo;
On 2013/01/10 19:29:40, Yaron wrote:
> We don't really make use of @Nullable elsewhere so it seems to be
stylistically
> at odds. Also, isn't it more useful as a formal argument? Here you're not
> instantiating it so it's obvious.

Done.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:165:
@Nullable String className = null;
On 2013/01/10 19:29:40, Yaron wrote:
> same

Done.

https://codereview.chromium.org/11778084/diff/1010/sync/android/java/src/org/...
sync/android/java/src/org/chromium/sync/notifier/InvalidationController.java:176:
Log.e(TAG, "Cannot read own application info", exception);
On 2013/01/10 19:29:40, Yaron wrote:
> THis is also a wtf moment?

Done.

[Yaron, 2013-01-11 03:12:46]

lgtm

[nyquist, 2013-01-11 20:54:20]

thakis: chrome/ non-android folders (ie. gyp)
tim: sync/

[tim (not reviewing), 2013-01-11 20:58:34]

On 2013/01/11 20:54:20, nyquist wrote:
> thakis: chrome/ non-android folders (ie. gyp)
> tim: sync/

Assuming we're going to add more to sync/android as we go, is there a directory
structure in mind (I see ModelType at the logical "root" with a 'notifier'
directory sibling)?  Would it make sense to mirror the sync/ structure?  So
ModelType would live in something like public/base?

[Nico, 2013-01-11 21:16:46]

chrome lgtm

[nyquist, 2013-01-12 01:36:52]

tim: Moved ModelType. PTAL.

[tim (not reviewing), 2013-01-12 01:43:48]

LGTM

[commit-bot: I haz the power, 2013-01-12 01:54:02]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nyquist@chromium.org/11778084/26001

[commit-bot: I haz the power, 2013-01-12 16:07:21]

Change committed as 176533