[Bindings] Simplify V8PrivateProperty::Symbol methods

Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods took a context as an argument.
But in case of setting/getting private property, the context to
be used is the current realm and it can be taken from the isolate.
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes the argument |context| from Symbol's methods
and does a clean-up changes around it.


BUG=611864
Review-Url: https://codereview.chromium.org/2785943002
Cr-Commit-Position: refs/heads/master@{#461051}
Committed: https://chromium.googlesource.com/chromium/src/+/55fd47019c03eac50969cd75ebac4119d3474bd4

[peria, 2017-03-30 05:50:53]

Patchset #1 (id:1) has been deleted

[peria, 2017-03-30 06:05:10]

Description was changed from

==========
Simplify V8PrivateProperty methods

BUG=611864
==========

to

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods required context as an argument.
But in most cases, the |context| to be used must be the current realm
and can be taken by isolate->GetCurrentContext().
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes an argument |context| from Symbol's methods
and does few small changes around it.


Note: NOPRESUBMIT is enabled to avoid a presubmit error
[readability/naming/underscores] for V8_PRIVATE_PROPERTY_DEFINE_GETTER. 
Presubmit checker checks NOLINT in only C++ style comment (// NOLINT)
and we can't use it in a macro definition.


BUG=611864
NOPRESUMBIT=true
==========

[peria, 2017-03-30 06:07:10]

The CQ bit was checked by peria@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-03-30 06:07:25]

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

[peria, 2017-03-30 07:08:25]

peria@chromium.org changed reviewers:
+ haraken@chromium.org, yukishiino@chromium.org

[peria, 2017-03-30 07:08:26]

PTL

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:142: static
Symbol V8_PRIVATE_PROPERTY_GETTER_NAME(                              \
A presubmit error warns for "V8_PRIVATE_PROPERTY_GETTER_NAME",
but we can't use a NOLINT comment in C-style comment.

[Yuki, 2017-03-30 07:25:42]

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:104: if
(v8Call(object->GetPrivate(context(), m_privateSymbol), value))
v8Call is deprecated, so it's better to explicitly call Maybe::ToLocalChecked().

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:123: Symbol(const
v8::Eternal<v8::Private>& eternalSymbol, v8::Isolate* isolate)
This looks a little tricky.  Taking v8::Local<v8::Private> seems straightforward
way to me, especially you're going to use ScopedPersistent for window.document.

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:123: Symbol(const
v8::Eternal<v8::Private>& eternalSymbol, v8::Isolate* isolate)
nit: Our convention is to take v8::Context / v8::Isolate as the first argument
following V8's style.

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:142: static
Symbol V8_PRIVATE_PROPERTY_GETTER_NAME(                              \
On 2017/03/30 07:08:26, peria wrote:
> A presubmit error warns for "V8_PRIVATE_PROPERTY_GETTER_NAME",
> but we can't use a NOLINT comment in C-style comment.

/* NOLINT(readability/naming/underscores) */
did work before.  Does this no longer work?

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:146:
v8::Eternal<v8::Private>& eternalProperty =                               \
nit: Hungarian notation is not recommended in general and in Google C++ coding
style, IIRC.  You don't need to include the type info in the name.  The name
should be given based on what it is or for what purpose it is.

[peria, 2017-03-30 07:41:37]

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:104: if
(v8Call(object->GetPrivate(context(), m_privateSymbol), value))
On 2017/03/30 07:25:42, Yuki wrote:
> v8Call is deprecated, so it's better to explicitly call
Maybe::ToLocalChecked().

Done.

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:123: Symbol(const
v8::Eternal<v8::Private>& eternalSymbol, v8::Isolate* isolate)
On 2017/03/30 07:25:41, Yuki wrote:
> This looks a little tricky.  Taking v8::Local<v8::Private> seems
straightforward
> way to me, especially you're going to use ScopedPersistent for
window.document.

Done.

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:142: static
Symbol V8_PRIVATE_PROPERTY_GETTER_NAME(                              \
On 2017/03/30 07:25:41, Yuki wrote:
> On 2017/03/30 07:08:26, peria wrote:
> > A presubmit error warns for "V8_PRIVATE_PROPERTY_GETTER_NAME",
> > but we can't use a NOLINT comment in C-style comment.
> 
> /* NOLINT(readability/naming/underscores) */
> did work before.  Does this no longer work?

Correct.  Presubmit checker works only for "// NOLINT" style.

https://codereview.chromium.org/2785943002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:146:
v8::Eternal<v8::Private>& eternalProperty =                               \
On 2017/03/30 07:25:42, Yuki wrote:
> nit: Hungarian notation is not recommended in general and in Google C++ coding
> style, IIRC.  You don't need to include the type info in the name.  The name
> should be given based on what it is or for what purpose it is.

Done.

[Yuki, 2017-03-30 07:53:03]

LGTM.

[peria, 2017-03-30 07:55:08]

Description was changed from

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods required context as an argument.
But in most cases, the |context| to be used must be the current realm
and can be taken by isolate->GetCurrentContext().
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes an argument |context| from Symbol's methods
and does few small changes around it.


Note: NOPRESUBMIT is enabled to avoid a presubmit error
[readability/naming/underscores] for V8_PRIVATE_PROPERTY_DEFINE_GETTER. 
Presubmit checker checks NOLINT in only C++ style comment (// NOLINT)
and we can't use it in a macro definition.


BUG=611864
NOPRESUMBIT=true
==========

to

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods required context as an argument.
But in most cases, the |context| to be used must be the current realm
and can be taken by isolate->GetCurrentContext().
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes an argument |context| from Symbol's methods
and does few small changes around it.


BUG=611864
==========

[peria, 2017-03-30 07:55:54]

The CQ bit was checked by peria@chromium.org

[peria, 2017-03-30 07:55:54]

The patchset sent to the CQ was uploaded after l-g-t-m from
yukishiino@chromium.org
Link to the patchset: https://codereview.chromium.org/2785943002/#ps60001
(title: "NOLINT comment hack")

[commit-bot: I haz the power, 2017-03-30 07:56:14]

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, 2017-03-30 08:02:57]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-30 08:02:58]

Try jobs failed on following builders:
  chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[peria, 2017-03-30 10:33:56]

On 2017/03/30 08:02:58, commit-bot: I haz the power wrote:
> Try jobs failed on following builders:
>   chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
>
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

Ah. It needs haraken's LGTM for core/events/MessageEvent.cpp

[haraken, 2017-03-30 13:53:34]

LGTM

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:95:
v8::Local<v8::Value> getOrUndefined(v8::Local<v8::Object> object) const {

Not related to this CL, I might want to remove this method and unify all call
sites to getOrEmpty(). If the call site wants to get undefined when the private
property is not set, the call site can explicitly use hasValue() or
getOrEmpty().

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:122:
v8::Local<v8::Context> context() const {

I'd prefer removing this method to prevent someone from using context() without
understanding what context she is in.

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/custom/V8ErrorEventCustom.cpp
(right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/custom/V8ErrorEventCustom.cpp:51:
ScriptState* scriptState = ScriptState::from(context);

We should use ScriptState::current and remove line 41.

[Yuki, 2017-03-30 14:18:14]

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:95:
v8::Local<v8::Value> getOrUndefined(v8::Local<v8::Object> object) const {
On 2017/03/30 13:53:34, haraken wrote:
> 
> Not related to this CL, I might want to remove this method and unify all call
> sites to getOrEmpty(). If the call site wants to get undefined when the
private
> property is not set, the call site can explicitly use hasValue() or
> getOrEmpty().

I'd like to do the opposite.  As V8 never returns the empty handle except for
cases that an exception is thrown (and the empty Maybe is returned), as haraken@
made ToV8() never return the empty handle, I don't want to return the empty
handle.

So, if possible, I'd like to remove getOrEmpty(), not getOrUndefined().  Also
note that getOrEmpty() is slower than getOrUndefined().

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:122:
v8::Local<v8::Context> context() const {
On 2017/03/30 13:53:34, haraken wrote:
> 
> I'd prefer removing this method to prevent someone from using context()
without
> understanding what context she is in.

This method is introduced in order to prevent someone from using the wrong
context.  Note that this method is private:, so client code cannot use this
method.

v8::Private::{Get,Set}Private never create a new object and the context
passed-in is only used for simple get / set operations otherwise for throwing an
exception.  In either case, the context we should use is the current context. 
We don't want call sites to pass in any other context.

This CL guarantees that the current context (i.e. the right context) is always
used.

As we think that (some of) V8 APIs do not need to take a v8::Context as the
first argument, this class does not need, too.  As we think that
v8::Private::{Get,Set}Private do not need to take a v8::Context, should it be
okay to always pass in the current context.

[haraken, 2017-03-30 14:21:37]

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:95:
v8::Local<v8::Value> getOrUndefined(v8::Local<v8::Object> object) const {
On 2017/03/30 14:18:14, Yuki wrote:
> On 2017/03/30 13:53:34, haraken wrote:
> > 
> > Not related to this CL, I might want to remove this method and unify all
call
> > sites to getOrEmpty(). If the call site wants to get undefined when the
> private
> > property is not set, the call site can explicitly use hasValue() or
> > getOrEmpty().
> 
> I'd like to do the opposite.  As V8 never returns the empty handle except for
> cases that an exception is thrown (and the empty Maybe is returned), as
haraken@
> made ToV8() never return the empty handle, I don't want to return the empty
> handle.
> 
> So, if possible, I'd like to remove getOrEmpty(), not getOrUndefined().  Also
> note that getOrEmpty() is slower than getOrUndefined().

I'm fine as long as the two methods are unified into one :)

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:122:
v8::Local<v8::Context> context() const {
On 2017/03/30 14:18:14, Yuki wrote:
> On 2017/03/30 13:53:34, haraken wrote:
> > 
> > I'd prefer removing this method to prevent someone from using context()
> without
> > understanding what context she is in.
> 
> This method is introduced in order to prevent someone from using the wrong
> context.  Note that this method is private:, so client code cannot use this
> method.
> 
> v8::Private::{Get,Set}Private never create a new object and the context
> passed-in is only used for simple get / set operations otherwise for throwing
an
> exception.  In either case, the context we should use is the current context. 
> We don't want call sites to pass in any other context.
> 
> This CL guarantees that the current context (i.e. the right context) is always
> used.
> 
> As we think that (some of) V8 APIs do not need to take a v8::Context as the
> first argument, this class does not need, too.  As we think that
> v8::Private::{Get,Set}Private do not need to take a v8::Context, should it be
> okay to always pass in the current context.

OK, then I'd like to rename it to currentContext() to clarify what context it
returns though.

[Yuki, 2017-03-30 15:02:43]

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:122:
v8::Local<v8::Context> context() const {
On 2017/03/30 14:21:37, haraken wrote:
> On 2017/03/30 14:18:14, Yuki wrote:
> > On 2017/03/30 13:53:34, haraken wrote:
> > > 
> > > I'd prefer removing this method to prevent someone from using context()
> > without
> > > understanding what context she is in.
> > 
> > This method is introduced in order to prevent someone from using the wrong
> > context.  Note that this method is private:, so client code cannot use this
> > method.
> > 
> > v8::Private::{Get,Set}Private never create a new object and the context
> > passed-in is only used for simple get / set operations otherwise for
throwing
> an
> > exception.  In either case, the context we should use is the current
context. 
> > We don't want call sites to pass in any other context.
> > 
> > This CL guarantees that the current context (i.e. the right context) is
always
> > used.
> > 
> > As we think that (some of) V8 APIs do not need to take a v8::Context as the
> > first argument, this class does not need, too.  As we think that
> > v8::Private::{Get,Set}Private do not need to take a v8::Context, should it
be
> > okay to always pass in the current context.
> 
> OK, then I'd like to rename it to currentContext() to clarify what context it
> returns though.

context() is a short and good name in this case, I think.  I'd rather propose to
add a comment why we always use the current context.

The important point is that context() returns the *right* context, it's not so
important that context() returns the current context.  If the call sites of this
helper function need to be aware of that it's *current* context, this function
helps nothing at all.  It's important that the call sites can use this helper
function without caring much about what the exact context is.

But, the explanation in a comment about why it's always the current context is
good for code readers.

[haraken, 2017-03-30 15:03:32]

On 2017/03/30 15:02:43, Yuki wrote:
>
https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
> File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):
> 
>
https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
> third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:122:
> v8::Local<v8::Context> context() const {
> On 2017/03/30 14:21:37, haraken wrote:
> > On 2017/03/30 14:18:14, Yuki wrote:
> > > On 2017/03/30 13:53:34, haraken wrote:
> > > > 
> > > > I'd prefer removing this method to prevent someone from using context()
> > > without
> > > > understanding what context she is in.
> > > 
> > > This method is introduced in order to prevent someone from using the wrong
> > > context.  Note that this method is private:, so client code cannot use
this
> > > method.
> > > 
> > > v8::Private::{Get,Set}Private never create a new object and the context
> > > passed-in is only used for simple get / set operations otherwise for
> throwing
> > an
> > > exception.  In either case, the context we should use is the current
> context. 
> > > We don't want call sites to pass in any other context.
> > > 
> > > This CL guarantees that the current context (i.e. the right context) is
> always
> > > used.
> > > 
> > > As we think that (some of) V8 APIs do not need to take a v8::Context as
the
> > > first argument, this class does not need, too.  As we think that
> > > v8::Private::{Get,Set}Private do not need to take a v8::Context, should it
> be
> > > okay to always pass in the current context.
> > 
> > OK, then I'd like to rename it to currentContext() to clarify what context
it
> > returns though.
> 
> context() is a short and good name in this case, I think.  I'd rather propose
to
> add a comment why we always use the current context.
> 
> The important point is that context() returns the *right* context, it's not so
> important that context() returns the current context.  If the call sites of
this
> helper function need to be aware of that it's *current* context, this function
> helps nothing at all.  It's important that the call sites can use this helper
> function without caring much about what the exact context is.
> 
> But, the explanation in a comment about why it's always the current context is
> good for code readers.

Makes sense.

[peria, 2017-03-30 15:30:54]

Description was changed from

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods required context as an argument.
But in most cases, the |context| to be used must be the current realm
and can be taken by isolate->GetCurrentContext().
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes an argument |context| from Symbol's methods
and does few small changes around it.


BUG=611864
==========

to

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods took a context as an argument.
But in case of setting/getting private property, the context to
be used is the current realm and it can be taken from the isolate.
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes the argument |context| from Symbol's methods
and does a clean-up changes around it.


BUG=611864
==========

[peria, 2017-03-30 15:53:24]

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h (right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/V8PrivateProperty.h:122:
v8::Local<v8::Context> context() const {
On 2017/03/30 15:02:43, Yuki wrote:
> On 2017/03/30 14:21:37, haraken wrote:
> > On 2017/03/30 14:18:14, Yuki wrote:
> > > On 2017/03/30 13:53:34, haraken wrote:
> > > > 
> > > > I'd prefer removing this method to prevent someone from using context()
> > > without
> > > > understanding what context she is in.
> > > 
> > > This method is introduced in order to prevent someone from using the wrong
> > > context.  Note that this method is private:, so client code cannot use
this
> > > method.
> > > 
> > > v8::Private::{Get,Set}Private never create a new object and the context
> > > passed-in is only used for simple get / set operations otherwise for
> throwing
> > an
> > > exception.  In either case, the context we should use is the current
> context. 
> > > We don't want call sites to pass in any other context.
> > > 
> > > This CL guarantees that the current context (i.e. the right context) is
> always
> > > used.
> > > 
> > > As we think that (some of) V8 APIs do not need to take a v8::Context as
the
> > > first argument, this class does not need, too.  As we think that
> > > v8::Private::{Get,Set}Private do not need to take a v8::Context, should it
> be
> > > okay to always pass in the current context.
> > 
> > OK, then I'd like to rename it to currentContext() to clarify what context
it
> > returns though.
> 
> context() is a short and good name in this case, I think.  I'd rather propose
to
> add a comment why we always use the current context.
> 
> The important point is that context() returns the *right* context, it's not so
> important that context() returns the current context.  If the call sites of
this
> helper function need to be aware of that it's *current* context, this function
> helps nothing at all.  It's important that the call sites can use this helper
> function without caring much about what the exact context is.
> 
> But, the explanation in a comment about why it's always the current context is
> good for code readers.

Thank you Shiino-san for the explanation.
I'll put a comment for this method.

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
File third_party/WebKit/Source/bindings/core/v8/custom/V8ErrorEventCustom.cpp
(right):

https://codereview.chromium.org/2785943002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/bindings/core/v8/custom/V8ErrorEventCustom.cpp:51:
ScriptState* scriptState = ScriptState::from(context);
On 2017/03/30 13:53:34, haraken wrote:
> 
> We should use ScriptState::current and remove line 41.

No, ScriptState::current(isolate) is deprecated.
I'll move line 41 to this argument.

[peria, 2017-03-31 03:39:54]

The CQ bit was checked by peria@chromium.org

[peria, 2017-03-31 03:39:55]

The patchset sent to the CQ was uploaded after l-g-t-m from
haraken@chromium.org, yukishiino@chromium.org
Link to the patchset: https://codereview.chromium.org/2785943002/#ps80001
(title: "work for comments")

[commit-bot: I haz the power, 2017-03-31 03:40:28]

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, 2017-03-31 05:44:46]

CQ is committing da patch.

Bot data: {"patchset_id": 80001, "attempt_start_ts": 1490931594857840,
"parent_rev": "e12d70fb8a80561804cf6b033cb355ca6fc74aed", "commit_rev":
"55fd47019c03eac50969cd75ebac4119d3474bd4"}

[commit-bot: I haz the power, 2017-03-31 05:45:34]

Description was changed from

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods took a context as an argument.
But in case of setting/getting private property, the context to
be used is the current realm and it can be taken from the isolate.
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes the argument |context| from Symbol's methods
and does a clean-up changes around it.


BUG=611864
==========

to

==========
Simplify V8PrivateProperty::Symbol methods

Before this CL, those methods took a context as an argument.
But in case of setting/getting private property, the context to
be used is the current realm and it can be taken from the isolate.
Plus, a context in the argument can be wrong if a developer
misunderstands what should be passed.

Hence this CL removes the argument |context| from Symbol's methods
and does a clean-up changes around it.


BUG=611864

Review-Url: https://codereview.chromium.org/2785943002
Cr-Commit-Position: refs/heads/master@{#461051}
Committed:
https://chromium.googlesource.com/chromium/src/+/55fd47019c03eac50969cd75ebac...
==========

[commit-bot: I haz the power, 2017-03-31 05:45:35]

Committed patchset #4 (id:80001) as
https://chromium.googlesource.com/chromium/src/+/55fd47019c03eac50969cd75ebac...