Fix fuzzer's bools to be 0 or 1 only

Fix fuzzer's bools to be 0 or 1 only

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2447823002

Committed: https://skia.googlesource.com/skia/+/85d301745a9031b13f1b716f07e1041f2fdd1ce3

[kjlubick, 2016-10-24 17:36:42]

Description was changed from

==========
Fix fuzzer's bools to be 0 or 1 only

BUG=skia:
==========

to

==========
Fix fuzzer's bools to be 0 or 1 only

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2447823002
==========

[kjlubick, 2016-10-24 17:38:23]

kjlubick@google.com changed reviewers:
+ mtklein@google.com

[mtklein_C, 2016-10-24 17:45:10]

mtklein@chromium.org changed reviewers:
+ mtklein@chromium.org

[mtklein_C, 2016-10-24 17:45:10]

https://codereview.chromium.org/2447823002/diff/1/fuzz/FuzzGradients.cpp
File fuzz/FuzzGradients.cpp (right):

https://codereview.chromium.org/2447823002/diff/1/fuzz/FuzzGradients.cpp#newc...
fuzz/FuzzGradients.cpp:83: !fuzz->next(&useLocalMatrix) ||
This code reads inconsistently, before this CL and more now.  If we're going to
line up ||, let's line them all up, and while we're at it, line up all the
!fuzz->next too.

Again, it'd be nice to either always use the type or never use the type when
calling next().  I'd recommend never using the type... it can be inferred in
every single one of these cases.

If you want to use the type, instead of overloading next(T*) with next(bool*),
we can make an explicit template function specialization.  It'd work very
similarly except that next<bool>(...) would pick up the specialized version. 
The downside is that there are some more restrictions on where you can define
it, so let's not get into that unless you really like the <Type> all over the
place.

[kjlubick, 2016-10-24 18:27:01]

https://codereview.chromium.org/2447823002/diff/1/fuzz/FuzzGradients.cpp
File fuzz/FuzzGradients.cpp (right):

https://codereview.chromium.org/2447823002/diff/1/fuzz/FuzzGradients.cpp#newc...
fuzz/FuzzGradients.cpp:83: !fuzz->next(&useLocalMatrix) ||
On 2016/10/24 at 17:45:10, mtklein_C wrote:
> This code reads inconsistently, before this CL and more now.  If we're going
to line up ||, let's line them all up, and while we're at it, line up all the
!fuzz->next too.
> 
> Again, it'd be nice to either always use the type or never use the type when
calling next().  I'd recommend never using the type... it can be inferred in
every single one of these cases.
> 
> If you want to use the type, instead of overloading next(T*) with next(bool*),
we can make an explicit template function specialization.  It'd work very
similarly except that next<bool>(...) would pick up the specialized version. 
The downside is that there are some more restrictions on where you can define
it, so let's not get into that unless you really like the <Type> all over the
place.

All the types removed and the || are lined up as well as !fuzz->next

[mtklein_C, 2016-10-24 18:31:52]

lgtm

[kjlubick, 2016-10-24 18:32:12]

The CQ bit was checked by kjlubick@google.com

[commit-bot: I haz the power, 2016-10-24 18:32:20]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-24 18:53:37]

Description was changed from

==========
Fix fuzzer's bools to be 0 or 1 only

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2447823002
==========

to

==========
Fix fuzzer's bools to be 0 or 1 only

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2447823002

Committed:
https://skia.googlesource.com/skia/+/85d301745a9031b13f1b716f07e1041f2fdd1ce3
==========

[commit-bot: I haz the power, 2016-10-24 18:53:38]

Committed patchset #2 (id:20001) as
https://skia.googlesource.com/skia/+/85d301745a9031b13f1b716f07e1041f2fdd1ce3