We no longer instantiate the transport just for the duration of handling a
backup or restore operation. Instead, we hold the object forever (replacing it
if instructed to do so). This makes it easier for transports to watch system
state and help set backup timing policy.
Also fixes up the IBackupTransport documentation a bit.
This change adds new APIs to control the gesture's path by deciding whether the
path should be drawn and by getting the ability to get the Path itself and
draw it differently.
The 'list sets' and 'restore token#' commands from bmgr now do what they are
supposed to. At this point we see the restore target's data being cleared
properly and its agent being launched and invoked for restore.
It now reconstructs the backup data stream as it was originally structured, no
longer assuming that its backed-up records can simply be concatenated into the
destination stream.
Fixes http://b/issue?id=1905863
This is needed to address two security issues with global search:
http://b/issue?id=1871088 (Apps can read content providers through GlobalSearch)
http://b/issue?id=1819627 (Apps can use GlobalSearch to launch arbirtrary intents)
This also fixes http://b/issue?id=1693153 (SearchManager.OnDismissListener
never gets called)
To fix the security issues, GlobalSearch also needs to require
a non-app permission to access its content provider and launch intents.
As of this change, LocalTransport is successfully propagating data changes from
the backup data format into a repository stored in /cache/backup/[packagename].
Each backup key gets a separate file there for ease of manipulation and testing.
The general semantics of BackupDataReader have been tweaked, too; it now just
returns simple "we're done with the data" when it hits the end, even if no
footer has been found, because on the writing side the footer isn't being
written. Also, reading an entity now merely requires a "big enough" buffer, not
an exactly-sized one.
This is all a work in progress, but this is at least working now for purposes of
this local transport.
Still to do: proper change vs deletion detection, as well as expanding the data
format itself to include necessary metadata etc.
ViewGroup was trying to access the AttachInfo without checking whether the
info object is null. The object can be null at varioous times in the lifecycle
of a list.
It is only used for numeric dates -- spelled-out dates have such a complex
variety of formats that they can only be meaningfully formatted from
locale strings.
In addition, the preference is left null when initializing, on the assumption
that the locale will still specify a more useful numeric format than we can
guess as part of a build-wide configuration.
But if the user has specified a format, the date will be formatted in the
order they asked for, with locale-appropriate punctuation substituted in.
With the new implementation of using a SuggestionItemView, we no longer need to draw a solid background for all items in the suggestion listview. Hence removing the code to draw a default white background, and make the item transparent so the listview-drawn selection hilite shows through.
* changes:
Make the file backup helper not crash if a file you requested can't be stated. This means you don't need to know if the files you are backing up exist or not -- we'll figure it out for you.
Fix SharedPrefsBackupHelper so it doesn't hard code the paths to the files.
which allows us to set the language and country (based on MCC otherwise
locale country). If we are unable to retrieve the ToS, or verify the
content, we load the default ToS (en_US) from the device.
BUG=1856455
Automated import of CL 149797
- harmonized private member variable names
- added success / failure codes (used for now in initialization)
- synchronized access to speech completed listener.
The format strings are newly generated from CLDR. The code is once again
the same as in cupcake: do the natural thing for the locale if the user
has never specified, but follow the checkbox if the user has ever set it.
* changes:
Renaming the Tts class to TextToSpeech for clarity. Switching to HashMap for parameter lists to pass key-value parameter pairs. Using TODO instead of FIXME in comments. Removing methods that don't belong anymore to the TTS interface now the feature becomes part of the platform.
