| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| |/ / / / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
First, update our various project-specific Eclipse compiler settings
configuration files to include the new Eclipse 4 flags.
Second, turn off the "Unchecked conversion from non-annotated type to
@NonNull" warnings; there are hundreds or thousands of these, and
there isn't much we can do about them when they're coming from
platform and library APIs.
Third, make the lint projects warning-clean again by addressing
various warnings Eclipse found (such as some unclosed resources and
some null handling issues; yesterday's null annotation fixes only
addressed errors, not warnings.)
Change-Id: If75f7401a1cbeef1bf58b47ccaa9ad17bede7f91
|
| |\ \ \ \ \
| |/ / / /
|/| | | | |
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This adds a workaround for
https://bugs.eclipse.org/bugs/show_bug.cgi?id=381418
Change-Id: I6a68602cd65a8656e4a6f62bf3b181045956136c
|
| |\ \ \ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
On Linux, the first time a screen size or theme is rendered, the wrong
image is shown (until the scene is re-rendered). The reason for this
is that the first time, the palette previews are generated. And this
has the side-effect of replacing the current render session for the
editor with the most recent render image from the palette preview.
The fix is simple: After rendering previews, trigger another render.
Change-Id: Ic00fa8b2703c4ce8ad4068412d5eea7b0b9f093c
|
| |\ \ \ \ \
| |/ / / /
|/| | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This patch provides the ability to launch an application on
all connected devices with a single launch. This applies only
to Run configurations (and not to Debug/Junit test).
UI changes:
The target tab in the launch configuration dialog provides
two options right now: Manual and Automatic. This CL adds
a third option that allows launching on all connected devices.
A drop down allows the user to further narrow down the list
to just physical devices or just emulators.
Change-Id: I721a4f41e59da24ae722b93e8ef801bc910f5442
|
| |\ \ \ \ \
| |_|_|/ /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | | |
EGL_KHR_fence_sync extension
* commit 'f92f72709089442c68a226ff506464a8fcf792fe':
DO NOT MERGE Implement the EGL_KHR_fence_sync extension
|
| | |\ \ \ \
|/ / / / /
| | | | |
| | | | |
| | | | | |
* commit '212ffd9f6f9685cfa2d6d4556b6ab545b34237a3':
DO NOT MERGE Implement the EGL_KHR_fence_sync extension
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Cherry-pick from aosp-master into jb-dev.
Bug: 6515813
Change-Id: I78f1bd115044aa9dcc534db855eaed24eb92ecee
|
| |\ \ \ \ \ |
|
| |/ / / / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The ManifestInfo class keeps a cache of the manifest metadata, such as
the package name, minSdkVersion, etc.
However, the code which cached the minSdkVersion and targetSdkVersion
fields was incorrectly placed inside an unrelated if-block, which
meant that in some cases (in particular, manifests specifying a
default theme) these fields would not get set, and the class would
return an incorrect value of 1 for minSdkVersion, regardless of the
actual value specified.
Change-Id: I32423cad45c9f6f775b7334ee3fd9bd183497abd
|
| |\ \ \ \ \
| |_|_|_|/
|/| | | | |
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This extends the addons_list XSD to support both
addons sites and OEM's system-image sites.
Then this defines a new schema "sys-img" XSD for
a site that offers only system images.
Change-Id: I77ff56c4c70f57ab8ea44cf4ea3094125653737b
|
| |\ \ \ \ |
|
| | | |/ /
| |/| |
| | | |
| | | | |
Change-Id: I6d9f7bd41e8d9656ef136ba0e2f3c273af049d6d
|
| |\ \ \ \ |
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | | |
Also fix a couple of potential NPEs.
Change-Id: I80d6b625d672ad2e7b96f2fce311aa4347a45a33
|
| |\ \ \ \
| |/ / /
|/| | | |
|
| | | | |
| | | |
| | | |
| | | | |
Change-Id: Ie30c362c71eff0d62e177e06687c362802d416b1
|
| |\ \ \ \
| |/ / /
|/| | | |
|
| | | |/
| |/|
| | |
| | |
| | | |
Bug: 6515813
Change-Id: I738fc2663d81876dc75ad560fd08506b423a21bf
|
| |\ \ \ |
|
| | | | |
| | | |
| | | |
| | | | |
Change-Id: I9023c937b757277962730aa664d8d3a0be85e830
|
| |\ \ \ \
| |/ / /
|/| | | |
|
| | | | |
| | | |
| | | |
| | | | |
Change-Id: I3ba053055c302747082f8a5d6720172cefefb9bc
|
| |\ \ \ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This changeset fixes various code fragments opening
resources without closing them.
Change-Id: I6ed48a32dc5de4c11cab394dd3883ebbb54d2938
|
| |\ \ \ \ \
| |/ / / / |
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Eclipse 4.2 includes analysis support for @Nullable and @NonNull
annotations. However, it requires these annotations to be *repeated*
on every single method implementing or overriding a superclass or
interface method (!).
This changeset basically applies the quickfixes to inline these
annotations. It also changes the retention of our nullness
annotations from source to class, since without this Eclipse believes
that a @NonNull annotation downstream is a redefinition of a @Nullable
annotation.
Finally, the null analysis revealed a dozen or so places where the
nullness annotation was either wrong, or some null checking on
parameters or return values needed to be done.
Change-Id: I43b4e56e2d025a8a4c92a8873f55c13cdbc4c1cb
|
| |\ \ \ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fragments should provide a default constructor, and only a default
constructor, such that they can be recreated by the system on
configuration changes.
This changeset adds a lint check to catch cases where this is not the
case -- such as fragments that are non-static innerclasses, or where
the the class or constructor is not public. It also warns if the
fragment contains any *other* constructors, since the fragment
documentation strongly advises against it.
Change-Id: I8cdd00fd7c74259f84977804e36ace7c43864026
|
| |\ \ \ \ \
| |/ / / / |
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This changeset contains various unrelated fixes to the lint
infrastructure:
(1) Tweak the way the classpaths are computed in the default lint
client method such that rather than reading and parsing the
.classpath file 3 times, once for each of source-path, output-path
and library-path, it's now processing it once and storing the
results for all 3.
(2) Override the lookup-classpath method in Eclipse to directly query
the Eclipse APIs for obtaining the classpath info.
(3) Add in user libraries found in libs/, since these don't
necessarily show up in the .classpath file.
(4) Fix a couple of bugs related to checking .class files: First, when
locating the project for a .class file, lint would search upwards
for the surrounding project, which meant looking for the nearest
parent containing an AndroidManifest.xml file. However, in the
case of .class files, it will first encounter the bin/ directory,
which can contain a manifest file, so it would compute a project
for the bin/ folder rather than its parent, which meant the source
paths would be wrong.
Second, the list of class entries to be processed by lint must be
sorted prior to processing; the code dealing with innerclasses
depends on that.
(5) Some minor code cleanup: Move some generic utility code and some
string literals out of specific detectors and into the generic
utility and constant classes.
(6) Cache results of the lint-project to eclipse-project lookup method
since that method is called repeatedly with the same (current)
project.
Change-Id: I33603eed8381ca54314202620cb1bb033e70f775
|
| |\ \ \ \ |
|
| | | |/ /
| |/| |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This CL fixes two issues:
- Fully expanding the path to the libraries (i.e commong -> out/host/../common.jar)
allows the use of "mm" from sdk/monitor/ folder.
- The target of the build (the monitor script) will now have a timestamp that
corresponds to when it was last built, as opposed to when the source script was
last modified. As a result, timestamps of the dependency jar files are correctly
checked with the last build time of the target.
Change-Id: I77469561d6750d1942cfeda1d6672d6d2a5dad51
|
| |\ \ \ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | | |
Change-Id: Ia47944b3e5420859913201c8e430b6b53841ee1c
|
| |\ \ \ \ \ |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change-Id: I92386ed8f6e8c2c17c3dc2803f4ac937cf3616c6
|
| |\ \ \ \ \ \
| |_|_|_|/ /
|/| | | | | |
|
| | |/ / / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
If you have both res/menu/main.xml and res/layout/main.xml, and the
menu file is open, then double clicking on the layout file would *not*
open up the layout. There was a bug in LayoutEditorMatchingStrategy
which meant that it only considered the project and base names of the
files, not whether the existing editor represented a layout file.
Change-Id: Idaf7c1b145b60510f4913896d7d0b2162744a399
|
| |\ \ \ \ \
| |/ / / /
|/| | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This allows us to only keep classes that are really used either
through code or through XML. Also tweak the default rules for
better control of animated properties.
Added a test of a custom property animation and fixed some other
misc test files.
Change-Id: I7cc5839a764881d8d3c7bfce0a3f12ea7cba660e
|
| |\ \ \ \ \ |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The layout editor stores both an SWT image (of the most recently
rendered scene), as well as an AWT image which is kept around such
that it can be used over again and over again as the scratch buffer
for layout lib. However, if you're opening a lot of editors, there's a
lot of these AWT buffers sitting around and most of the time not
getting used. In a low memory situation they should be able to be
freed up.
To allow for this, the AWT image buffers are now held only by a soft
(not weak) reference. If the soft reference is cleared by the garbage
collector, then a new AWT image buffer will be allocated when needed
for layoutlib.
There is one wrinkle in all this: On Linux, we prescale images
(because scaling while painting is very slow -- see issue 19447). This
means we need to keep the AWT image around. Thus, there's a second
field which is strong reference to the AWT image, which we set if
prescaling is enabled.
Change-Id: Ifb64e476c29194f0e99fec023612f9dc880dd6f4
|
| |\ \ \ \ \ \ |
|
| |/ / / / / /
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Also fix a couple of NPEs.
Change-Id: Iab4544d3538d370dd76d96e3a97d49555b7c6359
|
| |\ \ \ \ \ \ |
|
| | | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
When ADT encounters a custom view, it creates a view descriptor for
that custom view, and in ADT 20 it also finds the associated styles
and records these as attributes, which means code completion and the
property sheet will show these properties.
However, once the descriptor has been lazily created, it never
changes. If you go and edit the styles.xml file specifying the
attributes for your custom view, you need to restart the IDE before
code completion will pick up these changes.
This changeset makes this behavior more dynamic: It will now pick up
edits to the files specifying styles for a custom view.
Change-Id: I365119e18c74378a0a039c1e7a22641c94acf002
|
