This is an exact clone of the "public" class, so that applications
that access the raw content provider directly (i.e. Browser) can
continue doing so, at the expense of not being buildable with the
SDK.
The Download Provider itself will also use those constants.
Bug: 2245521
Change-Id: I15f8ca8a3b69719cfbf403fc2607aac0d3c2b79a
this is a clone of https://android-git.corp.google.com/g/#change,35174.
if the cache is full to its capacity and if a new statement is to be cached,
one of the entries in the cache is thrown out to make room for the new one.
but the one that is thrown out doesn't get deallocated by SQLiteProgram
because it doesn't know that it should.
fixed this by having SQLiteProgram finalize its sql statement in
releaseReference*() methods, if the statement is not in cache.
commit cb11364feefc200f10af6a01f776803acba2792a
Author: Andreas Huber <andih@google.com>
Date: Thu Jan 14 14:05:36 2010 -0800
Even if the decoder didn't extract the thumbnail at the specified time, accept it for now.
commit a54c0244b305caf11e67db49b7d3d8dba5f77751
Author: Andreas Huber <andih@google.com>
Date: Thu Jan 14 11:37:15 2010 -0800
Add some checks to make sure we extract the correct thumbnail frame, also revert to the hardware decoders for thumnbail extraction if the software decoders failed.
commit 0014ab17f2e0986044327a5ab22159de50b81e0a
Author: Andreas Huber <andih@google.com>
Date: Thu Jan 14 14:05:09 2010 -0800
Another instance of returning an error instead of asserting makes sense.
commit d2cac8c498a65b449a25ec216601830d23c165e7
Author: Andreas Huber <andih@google.com>
Date: Thu Jan 14 13:34:33 2010 -0800
Instead of asserting on invalid data (incomplete NAL unit), return an error.
to install an app on sdcard, just resources.
Add new install path for /asec in installd.
ignore . when checking for apk path since the sdcard packages id'ed
by package name.
Add new -s option to adb shell pm
Refactor fwd locked from scanMode to ApplicationInfo.
Add new flag for sd install
Add new parse flags for fwd locking and installing on sdcard
New mock api's in PackageManagerService to invoke MountService api's. These
will be refactored again and so have been wrapped internally.
Some error codes in PackageManager
Changes in PackageManagerService to use mPath and mScanPath during installation
and switch to using PackageParser.Package.applicationInfo attributes for
source and public resource directories.
Some known issues that will be addressed later
using system_uid for now. needs some tinkering with uid and packagesetting creation to use the actual app uid
error handling from vold not very robust. ignoring lot of things for now
sending a delayed destroy to delete packages. will revisit later
revisit temp file creation later. just copy for now
Merge commit '0123101af0e15221026f97f19dad23488840700a'
* commit '0123101af0e15221026f97f19dad23488840700a':
Add proper support for fixed position elements
Merge commit 'af2af4e53aedb14c781d0351565fd7bec55a141a' into eclair-mr2-plus-aosp
* commit 'af2af4e53aedb14c781d0351565fd7bec55a141a':
Add proper support for fixed position elements
This change exposes some constants that are needed for
starting the global search activity (QuickSearchBox):
- SearchManager.INTENT_ACTION_GLOBAL_SEARCH
- SearchManager.EXTRA_SELECT_QUERY
Change-Id: I2f1f10070f502a8d5b644e78c03837d5d8b43486
Merge commit 'ae4bd059caa71aa4652c0f07fde7c2687169906e' into eclair-plus-aosp
* commit 'ae4bd059caa71aa4652c0f07fde7c2687169906e':
ADT/Layoutlib: Reimplement parts of BitmapFactory
The original bitmap factory is mostly done in native and deals
with the normal android Bitmap class which has been replaced
in the layoutlib by a bitmap that is merely a wrapper around
an AWT BufferedImage.
This new BitmapFactory creates the layoutlib version of
Bitmap all in Java.
Change-Id: Ice8b5d19141a9a43f83349c159201bf85604b3b0
When the menu comes up over the activity we call removeCallbacks(mFlingRunnable);
The effect of that is that our mTouchMode variable is left dangling in the fling
state. This change calls endFling() after removeCallbacks. In the specific bug
being addressed we depend on reportScrollStateChange() because we avoid doing
requeries while scrolling.
- include basic build & timestamp information in the header
- collect stack traces immediately at bugreport time
- *also* show stack traces from last ANR (as before), if in the last 15 minutes
- alphabetize system properties dump
- remove now-obsolete (and always questionable) "dumpcrash" functionality
- write to a .tmp file first, then rename to the final filename
- friendly usage message
- output sections include the exact file/command run
- source de-weirdification -- it's several hundred lines shorter now
The same data is present in the same order as before, and I've tried
to preserve all the grep/search targets people might use to find their
way through the file.
I added Intent.ACTION_GLOBAL_SEARCH a little while ago, since
I thought we needed it to be different from the old
SearchManager.INTENT_ACTION_GLOBAL_SEARCH. Since GlobalSearch is
no longer built, reusing SearchManager.INTENT_ACTION_GLOBAL_SEARCH
should be fine.
Change-Id: If3a11a77a8e1d9940cf2b20817a60888cdd9d69f
Merge commit '70c4f6cfb4fd48d5153080d038ac0f5b6dbd98ec'
* commit '70c4f6cfb4fd48d5153080d038ac0f5b6dbd98ec':
Add a static method to obtain the mime type from extension.
Merge commit '021b7b443fa4bb24ec73d0c09cf019c14eba67bc' into eclair-mr2-plus-aosp
* commit '021b7b443fa4bb24ec73d0c09cf019c14eba67bc':
Add a static method to obtain the mime type from extension.
