If SDCH is disabled, don't add an SDCH filter

If SDCH is disabled, don't add an SDCH filter

This avoids a crash when SDCH is disabled, but the "Content-Encoding" response header contains "sdch".

BUG=167218
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=250395

[cbentzel, 2014-02-07 03:34:19]

Is this for https://code.google.com/p/chromium/issues/detail?id=167218

[andrea, 2014-02-07 04:02:05]

On 2014/02/07 03:34:19, cbentzel wrote:
> Is this for https://code.google.com/p/chromium/issues/detail?id=167218

Yes.

[mef, 2014-02-07 15:59:03]

On 2014/02/07 04:02:05, huey1 wrote:
> On 2014/02/07 03:34:19, cbentzel wrote:
> > Is this for https://code.google.com/p/chromium/issues/detail?id=167218
> 
> Yes.

Looks reasonable, but what would happen to the content if it is indeed sdch
encoded?
Will it just proceed as garbage or somehow fail gracefully?

[huey, 2014-02-07 22:30:45]

On 2014/02/07 15:59:03, mef wrote:
> On 2014/02/07 04:02:05, huey1 wrote:
> > On 2014/02/07 03:34:19, cbentzel wrote:
> > > Is this for https://code.google.com/p/chromium/issues/detail?id=167218
> > 
> > Yes.
> 
> Looks reasonable, but what would happen to the content if it is indeed sdch
> encoded?
> Will it just proceed as garbage or somehow fail gracefully?

It will just proceed as garbage. This seems preferable to crashing, though.

In my use case, the hosting app handles SDCH on its own. We explicitly disable
SDCH in CrNet. Without this change, we get a crash on SDCH-encoded responses.

[mef, 2014-02-07 22:34:18]

lgtm

[huey, 2014-02-11 04:06:07]

The CQ bit was checked by huey@google.com

[commit-bot: I haz the power, 2014-02-11 04:06:49]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/huey@chromium.org/138533008/320001

[commit-bot: I haz the power, 2014-02-11 14:17:17]

Change committed as 250395

[cbentzel, 2014-02-13 18:46:37]

Can this get a unit test? Seems like an easy thing to regress.

[mef, 2014-02-14 19:31:46]

On 2014/02/13 18:46:37, cbentzel wrote:
> Can this get a unit test? Seems like an easy thing to regress.

I've added a unit test: https://codereview.chromium.org/166273004