Support linearized loading

Support linearized loading

BUG=N/A

[jun_fang, 2015-09-23 12:41:52]

jun_fang@foxitsoftware.com changed reviewers:
+ thestig@chromium.org, tsepez@chromium.org

[jun_fang, 2015-09-23 12:41:53]

Hi Tom and Lei,

I submitted a new patch for you to review. All review comments have been
addressed.

[Tom Sepez, 2015-09-23 19:01:13]

On 2015/09/23 12:41:53, jun_fang wrote:
> Hi Tom and Lei,
> 
> I submitted a new patch for you to review. All review comments have been
> addressed.

Why did you choose to go to a new patch?  If you created a new branch, then
you can use 
$ git cl issue 1353093003
to cause it to upload to the old isse number, so that the comments aren't lost.

That aside, I'm concerned that were parsing a new binary format, and there is
only the one positive example as a test case.  Where are the tests that show
what happens when pathologically mangled input is presented to your decoder?

In particular, it should be possible to create a unit test for ReadHintTables
and
friends, where we feed desired bitstreams from the test harness into the
decoder.

[jun_fang, 2015-09-24 12:45:02]

On 2015/09/23 19:01:13, Tom Sepez wrote:
> On 2015/09/23 12:41:53, jun_fang wrote:
> > Hi Tom and Lei,
> > 
> > I submitted a new patch for you to review. All review comments have been
> > addressed.
> 
> Why did you choose to go to a new patch?  If you created a new branch, then
> you can use 
> $ git cl issue 1353093003
> to cause it to upload to the old isse number, so that the comments aren't
lost.
> 
> That aside, I'm concerned that were parsing a new binary format, and there is
> only the one positive example as a test case.  Where are the tests that show
> what happens when pathologically mangled input is presented to your decoder?
> 
> In particular, it should be possible to create a unit test for ReadHintTables
> and
> friends, where we feed desired bitstreams from the test harness into the
> decoder.

OK. Let's still use https://codereview.chromium.org/1353093003/.