The EBNF statemachine description language was to difficult
to remember changed to use a simpler and more obvious psuedo
language.
Added HANDLED and NOT_HANDLED as it makes the psuedo code
more obvious.
Added getCurrentMessage primarily for use by code in enter
so that it can know why the new state is being entered.
Change-Id: I1446e417b77684fbde0020b1da0975eedc57cce4
Merge commit '7361b7410f2427022cc8917b00553dc26107cded' into kraken
* commit '7361b7410f2427022cc8917b00553dc26107cded':
Uncomment a test that was marked as STOPSHIP
Merge commit '4a0823e5486f16b10762076d629f2f7e4d5da476' into froyo-plus-aosp
* commit '4a0823e5486f16b10762076d629f2f7e4d5da476':
Uncomment a test that was marked as STOPSHIP
Initialize the cache path only if the container exists.
This did exist before but was somehow removed.
Fix a known test failure as well.
Change-Id: I63219369b0e7abca4919fbde2a04035a0fea9e35
Merge commit 'b1110149cccc3b99e59ead34ca46e5ac026f6db9' into kraken
* commit 'b1110149cccc3b99e59ead34ca46e5ac026f6db9':
Use integer for loop counter instead of float
New test in FrameworkTest for the VelocityTracker class
Preserve '+' in phone numbers imported from SIM.
Sometimes using PhoneNumberUtils takes unexpected side effects.
For example, PAUSE mark (#) is removed and operation on the phone number does
not work well in some devices.
I also concern removing the PhoneNumberUtils logic since some other
mobile phones cannot understand # or alphabets and starts emitting errors
after the modification.
I'll add a flag instead of just removing the code and let users choose it.
Bug: 2329952
Change-Id: I4ffdd01ebb8ac320b8b67a5848a747af13c6ad1a
This test actually exposes some bugs -- the test passes as written, but
certain parts are commented out with references to the bugs in question.
Bug: 2665914
Bug: 2665954
Change-Id: I61220e0efadc4edbb6ad419d26fa1f0f50bbc24c
when the install location is set to internal only. Similary if install
location is set to external only(via adb), dont fall back to installing on
internal storage if sdcard is filled up.
If nothing is specified ie install location is set to let system decide,
we just check on internal storage.
Fix tests
Change-Id: I7400ccc131782d9c45284bd9ebd0dee43a3b412b
Move install location values to secure settings.
Diable attribute for UI. Set default value to auto.
Add command line interface to set install location via pm.
Change-Id: I80e97b3d24845adad7102f40dcbe238f00efa406
Move all tests for android.* classes from tests/AndroidTests and
tests/CoreTests into framework/base/<core|graphics>/tests.
Consolidate all tests for java.* classes to tests/CoreTests.
Eventually hopefully these will be moved to dalvik/ somewhere.
Remove tests/AndroidTests entirely.
Change-Id: I86584d086ab7bd045bb38a10b699907805298a95
The asserts were removed from that class and replaced by a status String object.
This allows ExpandableListTester to do the asserts instead.
These tests passed on a sapphire and passion devices as well as in the emulator.
The asserts in the main thread are expected to make these tests pass during the
continuous build too.
This is 7fbddb1db1 cherrypicked to froyo.
http://b/issue?id=2525846
The certificates of the package and its shared user certificates
are initialized only when installing the package.
When scanning the packages, the certificates are collected from the apk
if the time stamp changes. If not the one's from
PackageSetting(read from packages.xml) are reused.
Removed mergeSignatures and updateSignatures since we don't support
this any more.
Also fix some bugs related to reinstalling failed upgrades.
Some unit tests.
Change-Id: Ibdeff170bd386d723f774136b18e0ad59d9cdabb
and changed the "get next operation to sync" logic just look
at pending syncs, rather than them and periodic syncs
- made syncoperation dup-detection ignore the initialization
sync extra
- made the sync dispatcher treat initialization syncs as just
a regular sync request and also made it explicitly set or
clear the initialization extra based on whether the sync
adapter was in the syncable or unknown state
- change the getNextSync logic to prioritize syncable "unknown"
syncs above everything else (since they should be fast and
are important)
- make it reschedule completed initialization syncs if the
sync adapter is now marked syncable
- fix some logging in SyncStorageEngine
- change SyncStorageEngine to not reuse authority ids when one
is removed
http://b/issue?id=2531359http://b/issue?id=2429638
Change-Id: I79805b582da74f4f0b6193eafaff24c2371d51e8
- make EntityIterator extend Iterator and thus not throw a
RemoteException, instead converting it into a RuntimeException.
- rename ActiveSyncInfo to SyncInfo
- change getActiveSync to getCurrentSync
- remove the accessors in SyncInfo and instead make the final
fields publicly accessible
- made AbstractThreadedSyncAdapter.cancelSync not take a thread
Change-Id: I99fde5585bc5f1e95f4873ffbba189074a8d6372
http://b/issue?id=2553539http://b/issue?id=2553541http://b/issue?id=2553550
to froyo
- intepret a missing syncavble attribute from donut as "unsynced"
rather than the traditional "true"
- copy the sync settings from the authorities "contacts" and "calendar"
to "com.android.contacts" and "com.android.calendar" if the latter
don't already have settings
- delay the database cleanup until after boot completed, which will give
the GoogleLoginService accounts migration code a chance to run; this
was causing all the settings to get removed upon a donut to froyo upgrade
Change-Id: I8795e97ba0c9b930d1a50784229ca9ab15dff9d2
http://b/issue?id=2531359
version of the accounts.xml file that pre-dated periodic syncs,
e.g. eclair or early froyo. http://b/2515823
- make the AccountManagerService dump() use a getAccounts call that
doesn't check the GET_ACCOUNTS permission to make it useful
in "adb bugreport"
- add some logging to SyncManager to help track down a problem
Change-Id: Icb646909074e2d327d71f6bb39cf06b6fac29e77
with the current locale, not previous one, every time it is possible.
Add a unit test verifying this fix.
BUG: 2514026
Change-Id: I4dd1b047e1ab4bf399e2c3ebc0304c0cfa3a0928
Since there could be an arbitrary number of messages in
the queue but Handler/Looper/MessageQueue do not provide
any mechanism for removing "all" messages the best we can
do is ignore. Throwing an error is probably too heavy.
Change-Id: I13c81ac5786484f5b3218885b010de596d943975
Add support for quiting the looper and stopping a thread.
Allow enter/exit at top or bottom of a STATE EBNF.
Fix missing brace in MSG_LIST.
Change-Id: Iddf5ce908008933bf8f0646e844254183da3d1f3
