[V8] Added line, column and script symbols for SyntaxError

[V8] Added line, column and script symbols for SyntaxError

For exception in promise we generate v8::Message API object from exception object. And in cases of Syntax or Reference Error we don't have enough information in exception object - we can't restore Error location from top stack frame.
In this patch three aditional private fields introduced for exception object. In case of Syntax Error we store line, column and script on Exception object and receive this information when restoring message.

BUG=443140
LOG=Y
R=yurys@chromium.org

Committed: https://crrev.com/4e9e1eedf1ba399d00b53adf6e5722d442ed9797
Cr-Commit-Position: refs/heads/master@{#26393}

[kozy, 2015-01-29 14:08:46]

Yury, please take a look.

[yurys, 2015-01-30 13:56:24]

This needs a test

[kozy, 2015-01-30 17:50:19]

Test added.

[yurys, 2015-02-02 07:07:57]

It might be worth mentioning in description that in case of syntax error call
stack is empty and cannot be used to derive location.

https://codereview.chromium.org/885043002/diff/20001/src/isolate.cc
File src/isolate.cc (right):

https://codereview.chromium.org/885043002/diff/20001/src/isolate.cc#newcode1080
src/isolate.cc:1080: const int start_pos_value =
Handle<Smi>::cast(start_pos)->value();
style: We normally don't mark local variables as const even if they are only
read in the rest of code. The only exception is constants whose value is known
at compile time.

https://codereview.chromium.org/885043002/diff/20001/src/isolate.cc#newcode1093
src/isolate.cc:1093: Handle<Script> casted_script(Script::cast(*script));
typo: casted -> cast

https://codereview.chromium.org/885043002/diff/20001/test/cctest/test-api.cc
File test/cctest/test-api.cc (left):

https://codereview.chromium.org/885043002/diff/20001/test/cctest/test-api.cc#...
test/cctest/test-api.cc:18156: promise_reject_line_number =
stack_trace->GetFrame(0)->GetLineNumber();
Now we don't check line number in the stack trace, is that intentional? Also it
would make sense to add column number checks in other places if you validate it
in the new test.

[kozy, 2015-02-02 09:18:57]

https://codereview.chromium.org/885043002/diff/20001/src/isolate.cc
File src/isolate.cc (right):

https://codereview.chromium.org/885043002/diff/20001/src/isolate.cc#newcode1080
src/isolate.cc:1080: const int start_pos_value =
Handle<Smi>::cast(start_pos)->value();
On 2015/02/02 07:07:56, yurys wrote:
> style: We normally don't mark local variables as const even if they are only
> read in the rest of code. The only exception is constants whose value is known
> at compile time.

Done.

https://codereview.chromium.org/885043002/diff/20001/src/isolate.cc#newcode1093
src/isolate.cc:1093: Handle<Script> casted_script(Script::cast(*script));
On 2015/02/02 07:07:56, yurys wrote:
> typo: casted -> cast

Done.

https://codereview.chromium.org/885043002/diff/20001/test/cctest/test-api.cc
File test/cctest/test-api.cc (left):

https://codereview.chromium.org/885043002/diff/20001/test/cctest/test-api.cc#...
test/cctest/test-api.cc:18156: promise_reject_line_number =
stack_trace->GetFrame(0)->GetLineNumber();
On 2015/02/02 07:07:56, yurys wrote:
> Now we don't check line number in the stack trace, is that intentional? Also
it
> would make sense to add column number checks in other places if you validate
it
> in the new test.

Fixed and added column number checks.

[yurys, 2015-02-02 09:38:31]

lgtm

https://codereview.chromium.org/885043002/diff/40001/test/cctest/test-api.cc
File test/cctest/test-api.cc (right):

https://codereview.chromium.org/885043002/diff/40001/test/cctest/test-api.cc#...
test/cctest/test-api.cc:18166: promise_reject_line_number =
stack_trace->GetFrame(0)->GetLineNumber();
Consider adding promise_reject_line_number here and checking it below.

[kozy, 2015-02-02 10:09:32]

kozyatinskiy@chromium.org changed reviewers:
+ yangguo@chromium.org

[kozy, 2015-02-02 10:09:33]

Yang, please take a look.

https://codereview.chromium.org/885043002/diff/40001/test/cctest/test-api.cc
File test/cctest/test-api.cc (right):

https://codereview.chromium.org/885043002/diff/40001/test/cctest/test-api.cc#...
test/cctest/test-api.cc:18166: promise_reject_line_number =
stack_trace->GetFrame(0)->GetLineNumber();
On 2015/02/02 09:38:31, yurys wrote:
> Consider adding promise_reject_line_number here and checking it below.

Done.

[Yang, 2015-02-03 08:22:39]

On 2015/02/02 10:09:33, kozyatinskiy wrote:
> Yang, please take a look.
> 
> https://codereview.chromium.org/885043002/diff/40001/test/cctest/test-api.cc
> File test/cctest/test-api.cc (right):
> 
>
https://codereview.chromium.org/885043002/diff/40001/test/cctest/test-api.cc#...
> test/cctest/test-api.cc:18166: promise_reject_line_number =
> stack_trace->GetFrame(0)->GetLineNumber();
> On 2015/02/02 09:38:31, yurys wrote:
> > Consider adding promise_reject_line_number here and checking it below.
> 
> Done.

lgtm!

[kozy, 2015-02-03 08:26:24]

The CQ bit was checked by kozyatinskiy@chromium.org

[commit-bot: I haz the power, 2015-02-03 08:27:05]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/885043002/80001

[commit-bot: I haz the power, 2015-02-03 08:57:31]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2015-02-03 08:57:44]

Patchset 5 (id:??) landed as
https://crrev.com/4e9e1eedf1ba399d00b53adf6e5722d442ed9797
Cr-Commit-Position: refs/heads/master@{#26393}