Library projects in the SDK are built using --non-constant-id
to generate a temporary R.java class.
When the library is packaged with the application to generate an
apk, the R class is recreated with the proper IDs due to all the
resources coming from the app and all the libraries.
However for large apps with many libraries (each with their own
R class in their package), this means a lot of unnecessary IDs:
all R classes contains all the IDs including for resources from
by projects they don't have access through the dependency graph.
For really large apps (X,000 resources), with lots of libraries
(10+), this can generate tens of thousands of resources, which
can trigger dalvik's limit of 65K fields and methods per dex
files.
This changes lets aapt generate not only the R class but a simple
text file containing the list of all those IDs so that it is
easier to parse back. The SDK build system will not ask aapt
to generate the R class of the libraries (through the
--extra-packages option), instead it will then read this
file to know what IDs are needed for each library and generate
a much smaller R class for each library (using the same text
file output from compiling all the resources to get the final
integer value).
Change-Id: I4db959fec372cf3ead9950e4b2b82fa1ae7eed2d
The new SDK build system give the ability to insert
versionCode/Name and min/targetSdkVersion in the manifest
but aapt won't replace those if they already exist.
The main problem is that aapt doesn't actually fail when
it doesn't replace them, making the output not what the
developer wanted.
This patch set adds an option to aapt to make it return
an error if the insert failed because the attribute
already existed.
Change-Id: I8938ec1238da407a8562c974e9598db39001ffd9
The code added attributes android:name and class to a
KeyedVector under the same key (fragment) so the 2nd
add (android:name) removed class with was never checked.
This replace the value type in the KeyedVector to be
Vector<NamespaceAttributePair> instead of just
NamespaceAttributePair.
Change-Id: I009b8a8cca878191661c2a63bb14c967d230498d
The android version depends on a custom version of LinkedHashMap
which is not present on desktop VMs. This new implementation is done
in a way that minimizes the difference between the two.
Also some minor fixes.
(cherry picked from commit 01b6c755dbcf24e71192dc44757e2eea2a426091)
Change-Id: Idc7bca820e472e281a3024a5b610fd55606cf428
When using aapt dump xmltree to dump an XML which has a content element with newlines, the output contains a
newline. This makes it very difficult, sometimes impossible, to understand what is part of the content, and
what is the meta-data.
We now pass XML content through the same normalizer used for other tags.
Change-Id: I327321520fac563eb32aecaf796f2473866697fc
Usage: layoutlib_create --list-deps /path/to/layoutlib.jar
Prints:
- all classes found in the Jar and the types they use (the dependencies).
- all the dependencies missing from the Jar and what uses them.
Change-Id: I8b2674df127e1494feed7a653282e88e4d2f5494
This is a revert of 1db36528b12395b9ed9bf8a1005a6d4ace737627,
but with comments added so I don't make this mistake again. :)
Change-Id: I053216279e3721f08f32f561bb989736ef619f82
With SDK 10.6 without _DARWIN_UNLIMITED_STREAMS aapt sometimes fails
right away with the error "Too many open files" when calling fopen().
Change-Id: Ifa7bd8a9530d706aa47f98be8186f1aefe943d76
Add .DS_Store to the list of ignored files that are silently
ignored (other dot-files are ignored but aapt emits a "(skipping <x>)"
message.)
Also, add a "!" prefix to the *~ pattern for Emacs/Vim/Gedit backup
files.
Finally, move the !*.scc pattern up in front of the .* pattern, such
that it doesn't match the earlier .* pattern (which is verbose, unlike
!*.scc).
Change-Id: Id3e96490f1802486aea8c58366d43e9d413971b8
...content provider and updating its oom adj
This introduces the concept of an "unstable" reference on a content
provider. When holding such a reference (and no normal stable ref),
the content provider dying will not cause the client process to be
killed.
This is used in ContentResolver.query(), .openAssetFileDescriptor(),
and .openTypedAssetFileDescriptor() to first access the provider
with an unstable reference, and if at the point of calling into the
provider we find it is dead then acquiring a new stable reference
and doing the operation again. Thus if the provider process dies
at any point until we get the result back, our own process will not
be killed and we can safely retry the operation.
Arguably there is still the potential for a race -- if somehow the
provider is killed way late by the OOM killer after the query or
open has returned -- but this should now be *extremely* unlikely.
We also continue to have the issue with the other calls, but these
are much less critical, and the same model can't be used there (we
wouldn't want to execute two insert operations for example).
The implementation of this required some significant changes to the
underlying plumbing of content providers, now keeping track of the
two different reference counts, and managing them appropriately. To
facilitate this, the activity manager now has a formal connection
object for a client reference on a content provider, which hands to
the application when opening the provider.
These changes have allowed a lot of the code to be cleaned up and
subtle issues closed. For example, when a process is crashing, we
now have a much better idea of the state of content provider clients
(olding a stable ref, unstable ref, or waiting for it to launch), so
that we can correctly handle each of these.
The client side code is also a fair amount cleaner, though in the
future there is more than should be done. In particular, the two
ProviderClientRecord and ProviderRefCount classes should be combined
into one, part of which is exposed to the ContentResolver internal
API as a reference on a content provider with methods for updating
reference counts and such. Some day we'll do that.
Change-Id: I87b10d1b67573ab899e09ca428f1b556fd669c8c
After skipping * with "token++", the length should decrease by 1 as
well.
(merged from 996b073e813ba1a22a13282ccdebb664f14ba898)
Change-Id: Ie6232ef603bb31e25e03b926e6c1bb92ac34902d
AAPT has a fixed built-in list of files and directories
to ignore when parsing resource files. Over the years we
always had developers requiring specific patterns.
Added a command-line option for it:
aapt di --ignore-assets "foo*:*.blah"
If the env var ANDROID_AAPT_IGNORE is set, it is parsed
to find which file/directory patterns to ignore.
Otherwise a default is used that matches the current behavior.
SDK Bug: 5343 24067
(cherry-pick from AOSP 90897ed87bce639bf6bb2ccf15fbabb59b131bab)
Change-Id: Ia4caa2a8188c8c1df143f884e459b8182645995f
Also fix Activity menu inflater when using the dark on light
theme wrapper to still be able to find onClick listeners.
Change-Id: Ie206db26d1df96041bc477804e476b02ad99dc9d
