Fix auto-res namespace handling

[lexs]

@holmari notified me that 8d0e029 might cause a regression in regards to library handling. This is because resources from a library project will have http://schemas.android.com/apk/res-auto replaced with the library package name instead of the application package name.

This commit changes this so we never actually replace package names but instead look up by http://schemas.android.com/apk/res-auto if looking up by package name fail. I don't have a specific test case but regression tests pass.

[lexs]

The relevant changes are https://github.com/robolectric/robolectric/pull/1847/files#diff-1480b00057d655fbfd6b8fa7a7b128d9R267, https://github.com/robolectric/robolectric/pull/1847/files#diff-1480b00057d655fbfd6b8fa7a7b128d9R298 and https://github.com/robolectric/robolectric/pull/1847/files#diff-1480b00057d655fbfd6b8fa7a7b128d9R327

[holmari]

I tested this locally. It fixes the regression - we had 18 tests cases that failed earlier, but they pass now with this patch.

[jongerrish]

I think this PR might have caused some breakages in our build:-

On one project:-
Caused by: java.lang.RuntimeException: no value found for true
at org.robolectric.shadows.Converter$EnumOrFlagConverter.findValueFor(Converter.java:431)
at org.robolectric.shadows.Converter$EnumConverter.fillTypedValue(Converter.java:394)
at org.robolectric.shadows.Converter$EnumConverter.fillTypedValue(Converter.java:386)
at org.robolectric.shadows.Converter.convertAndFill(Converter.java:157)
at org.robolectric.shadows.Converter.convertAndFill(Converter.java:50)
at org.robolectric.shadows.ShadowResources.createTypedArray(ShadowResources.java:232)
at org.robolectric.shadows.ShadowResources.attrsToTypedArray(ShadowResources.java:209)
at org.robolectric.shadows.ShadowResources.access$000(ShadowResources.java:55)
at org.robolectric.shadows.ShadowResources$ShadowTheme.obtainStyledAttributes(ShadowResources.java:485)
at sun.reflect.GeneratedMethodAccessor31.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at org.robolectric.internal.bytecode.ShadowWrangler$ShadowMethodPlan.run(ShadowWrangler.java:437)
at android.content.res.Resources$Theme.obtainStyledAttributes(Resources.java)
at android.widget.TextView.$$robo$$constructor(TextView.java:752)
at android.widget.TextView.(TextView.java)
at android.widget.TextView.(TextView.java)

On another:-

Caused by: java.lang.NumberFormatException: For input string: "/path/to/resources/color/some_resource.xml"
at java.lang.NumberFormatException.forInputString(NumberFormatException.java:65)
at java.lang.Integer.parseInt(Integer.java:481)
at com.android.internal.util.XmlUtils.convertValueToInt(XmlUtils.java:122)
at android.content.res.TypedArray.getInt(TypedArray.java:255)
at android.widget.ImageView.constructor(ImageView.java:155)
at android.widget.ImageView.(ImageView.java)
at android.widget.ImageButton.(ImageButton.java)

[lexs]

@jongerrish Ugh, I'll see if I can figure something out during the day. The resource code is icky and hard to test as you and @holmari have conflicting cases here. It would be nice to have a few samples that we test against that can be more complex.

By the way it looks like you have stack trace filtering off, is this for performance reasons?

[jongerrish]

It's for confidentiality reasons I can't leak stack traces unfortunately...

...let me know if you have any questions I'll see what info I can provide.
On May 29, 2015 1:49 AM, "Alexander Blom" notifications@github.com wrote:

@jongerrish https://github.com/jongerrish Ugh, I'll see if I can figure
something out during the day. The resource code is icky and hard to test as
you and @holmari https://github.com/holmari have conflicting cases
here. It would be nice to have a few samples that we test against that can
be more complex.

By the way it looks like you have stack trace filtering off, is this for
performance reasons?

—
Reply to this email directly or view it on GitHub
#1847 (comment)
.

[lexs]

@jongerrish Would be great if you could give #1851 a try.

[jongerrish]

I just ran this through our submit queue and it works. Previous failures are fixed and no new failures. I'll go ahead and merge this. Thanks for the quick response.