[alternate fix] Ensure Content-Type and encoding consistency in HttpResponse

Slightly modified approach to:
http://codereview.appspot.com/1881043/show

Ensures that charset in Content-Type header is in sync w/ any explicitly-specified encoding to MutableContent/HttpResponseBuilder.

[zhoresh, 2010-07-27 19:48:05]

lgtm, see minor comment within.
A test for the specific issue that you try to resolve would be good too.

http://codereview.appspot.com/1903045/diff/1/2
File
java/gadgets/src/main/java/org/apache/shindig/gadgets/rewrite/MutableContent.java
(right):

http://codereview.appspot.com/1903045/diff/1/2#newcode115
java/gadgets/src/main/java/org/apache/shindig/gadgets/rewrite/MutableContent.java:115:
// Never happens.
now it can happen?

http://codereview.appspot.com/1903045/diff/1/2#newcode191
java/gadgets/src/main/java/org/apache/shindig/gadgets/rewrite/MutableContent.java:191:
protected void setContentBytesState(byte[] newBytes, Charset newEncoding) {
Add explanation - can be override by derived class that need to know when
encoding of content is changed.

[johnfargo, 2010-07-27 20:15:09]

Removing unnecessary helper method, adding comments (simplifyng String creation

[johnfargo, 2010-07-27 20:33:41]

Adding a test for the affected case: always a good idea. Includes key bugfix in
HttpResponse ctor.

[johnfargo, 2010-07-27 20:45:57]

A few more comments for exposition.

[johnfargo, 2010-07-27 20:52:32]

Oops, missed the added test.

[zhoresh, 2010-07-27 20:56:54]

LGTM, thanks for the test and all the inline docs!

[johnfargo, 2010-07-27 21:00:07]

Committed.

Many thanks to Gagan for finding this issue!

On 2010/07/27 20:56:54, zhoresh wrote:
> LGTM, thanks for the test and all the inline docs!