http://gwt-code-reviews.appspot.com/1236801/diff/1/3
File user/src/com/google/gwt/resources/ext/ResourceContext.java (right):
http://gwt-code-reviews.appspot.com/1236801/diff/1/3#newcode147
user/src/com/google/gwt/resources/ext/ResourceContext.java:147: * a
given method. This data will be populated and available prior to any
On 2010/12/21 17:03:03, bobv wrote:
method, based on the presence of an {...@link Source} annotation or the
method or
the {...@link DefaultExtensions} annotation of the ResourceGenerator.
Hopefully a bit clearer now
The whole populated and available statement seems kind of a weird
thing to
guarantee here.
Is there any time where it's invalid to call this method?
http://gwt-code-reviews.appspot.com/1236801/diff/1/3#newcode170
user/src/com/google/gwt/resources/ext/ResourceContext.java:170:
On 2010/12/21 17:03:03, bobv wrote:
Whitespace.
Done.
http://gwt-code-reviews.appspot.com/1236801/diff/1/6
File
user/src/com/google/gwt/resources/rebind/context/AbstractClientBundleGenerator.java
(right):
http://gwt-code-reviews.appspot.com/1236801/diff/1/6#newcode231
user/src/com/google/gwt/resources/rebind/context/AbstractClientBundleGenerator.java:231:
On 2010/12/21 17:03:03, bobv wrote:
Whitespace, here and elsewhere.
Done.
http://gwt-code-reviews.appspot.com/1236801/diff/1/6#newcode362
user/src/com/google/gwt/resources/rebind/context/AbstractClientBundleGenerator.java:362:
return new RebindResult(RebindStatus.USE_ALL_NEW, createdClassName,
cpi);
On 2010/12/21 17:03:03, bobv wrote:
Where do the previously-emitted resource files live (e.g.
*.cache.png)? Are
they deleted between refreshes so that staleness will become glaring
obvious?
This is just reverting to the same behavior prior to the this patch,
that is, previously emitted resource files have already been overwritten
(or renamed) by the time we get here, and thus we return the status to
USE_ALL_NEW resources/artifacts. If we had bailed early above prior to
re-running the generation (and returning a status of USE_ALL_CACHED),
we'd be using the previously generated resource files in place (and not
overwritten).
http://gwt-code-reviews.appspot.com/1236801/diff/1/6#newcode438
user/src/com/google/gwt/resources/rebind/context/AbstractClientBundleGenerator.java:438:
long lastModified =
generatorContext.getSourceLastModifiedTime(sourceType);
On 2010/12/21 17:03:03, bobv wrote:
How does this handle the case where a CssResource subtype is not
necessarily
defined in the same compilation unit as the ClientBundle subtype?
Good point, I've added the method's return type (e.g. a resource type),
and it's parents, to the list of source types that need to be checked
(see below).
http://gwt-code-reviews.appspot.com/1236801/diff/1/6#newcode655
user/src/com/google/gwt/resources/rebind/context/AbstractClientBundleGenerator.java:655:
cacheable = false;
On 2010/12/21 17:03:03, bobv wrote:
Why not just return false immediately?
I've added a comment to clarify above the decl for 'cacheable', but
essentially, we need to iterate over all the methods to completion
regardless of whether cacheable becomes false, since if it is false,
then we still need to find and add all the resouces to the context, for
subsequent processing.
http://gwt-code-reviews.appspot.com/1236801/diff/1/7
File
user/src/com/google/gwt/resources/rebind/context/AbstractResourceContext.java
(right):
http://gwt-code-reviews.appspot.com/1236801/diff/1/7#newcode142
user/src/com/google/gwt/resources/rebind/context/AbstractResourceContext.java:142:
On 2010/12/21 17:03:03, bobv wrote:
Whitespace, here and elsewhere.
Done.
http://gwt-code-reviews.appspot.com/1236801/show
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
