Make apks be built in a consistent way

Make apks be built in a consistent way

Simplify apk targets by doing most of the work in build/java_apk.gypi
(like jars are done with build/java.gypi). The apk targets still use
their own ant build files, but everything else is more consistent
between them.

BUG=


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=156233

[cjhopman, 2012-09-07 20:06:41]

tedchoc: content/shell/android, build/
brettw: content/

[Ted C, 2012-09-07 20:21:10]

lgtm

Aside from the first couple nits.  Looks like currently we can't reduce the
number of paths specified, but I do think we should investigate if we are doing
something wrong currently that prevents that.

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi
File build/java_apk.gypi (right):

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi#newcode6
build/java_apk.gypi:6: # to build Java in a consistent manner.
s/Java/an Android APK

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi#newcode10
build/java_apk.gypi:10: #   'target_name': 'my-package_apk',
is the underscore important before the apk?  Just wondering why it is mixing
hyphens w/ underscores

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi#newcode47
build/java_apk.gypi:47: '<!@(find <(resource_dir) -name "*")',
is resource_dir ever different from <(java_in_dir)/res ?

http://codereview.chromium.org/10914130/diff/2001/content/content_shell.gypi
File content/content_shell.gypi (right):

http://codereview.chromium.org/10914130/diff/2001/content/content_shell.gypi#...
content/content_shell.gypi:608: 'resource_dir': 'shell/android/res',
hmm...I see resource_dir is I guess different.  Sadly I think that is a bug.  I
believe the res directory should live at the same level at the java source
dir...i.e.

java_in_dir: shell/android/java/src
resource_dir: shell/android/java/res

If we enforce this paradigm, we could just specify shell/android/java and look
for src/ and res/ under it.

http://codereview.chromium.org/10914130/diff/2001/content/content_tests.gypi
File content/content_tests.gypi (right):

http://codereview.chromium.org/10914130/diff/2001/content/content_tests.gypi#...
content/content_tests.gypi:826: 'resource_dir': 'shell/android/res',
blast...proven wrong again.  Does the test actually need to point to the same
resources?  It would be nice if there was a way for test apks to use the
underlying R.java in the activity is is testing, but I don't know if that is
feasible.

[cjhopman, 2012-09-07 20:52:36]

For the resource_dir stuff: Yes, I consider this a bug. There are some other
issues with how content_shell_test_apk is being built (like it directly refers
to the .class files created by content_shell's javac task). This is on my list
of things to fix, and I expect that in the future we can enforce a proper
directory structure on these targets.

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi
File build/java_apk.gypi (right):

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi#newcode6
build/java_apk.gypi:6: # to build Java in a consistent manner.
On 2012/09/07 20:21:10, Ted C wrote:
> s/Java/an Android APK

Done.

http://codereview.chromium.org/10914130/diff/2001/build/java_apk.gypi#newcode10
build/java_apk.gypi:10: #   'target_name': 'my-package_apk',
On 2012/09/07 20:21:10, Ted C wrote:
> is the underscore important before the apk?  Just wondering why it is mixing
> hyphens w/ underscores

The target name isn't actually important. The _apk suffix is just a convention
that seems to have been adopted by all (most?) of the current apk targets.

We don't ever really use hyphens in target names, though, so I've changed that
to an underscore.

[cjhopman, 2012-09-11 18:24:59]

Brettw: please take a look at the gypi files in content/

[brettw, 2012-09-11 18:31:28]

lgtm rubberstamp

[commit-bot: I haz the power, 2012-09-12 02:09:29]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/cjhopman@chromium.org/10914130/5003

[commit-bot: I haz the power, 2012-09-12 04:31:42]

Change committed as 156233