Implement Date constructor.

[LaszloLango]

JerryScript-DCO-1.0-Signed-off-by: László Langó [email protected]

[LaszloLango]

Depends on #305

[galpeter]

Space before =

[galpeter]

Should this be ecma_get_number_from_value ? As the parse_res_value is not a completion value.

[LaszloLango]

why is parse_res_value not a completion value?

[galpeter]

See: https://github.com/Samsung/jerryscript/blob/master/jerry-core/ecma/operations/ecma-try-catch-macro.h#L43

[LaszloLango]

I missed that. I'll fix it

[ruben-ayrapetyan]

JERRY_ASSERT?

[LaszloLango]

@ruben-ayrapetyan, sorry, I don't get it. What do you mean exactly? What would you like to test with the assert?

[ruben-ayrapetyan]

@LaszloLango, I mean that ret_value can be either empty or contain throw value just before the if block. Here we handle empty case in if and throw in else if. As there should be no other cases, we could just assert this, replacing else if with else and adding JERRY_ASSERT with contents of current else if condition to the else block.

Like the following:

if (ecma_is_completion_value_empty (ret_value))
{
 ...
}
else
{
  JERRY_ASSERT (ecma_is_completion_value_throw (ret_value));

  ...
}

[LaszloLango]

@ruben-ayrapetyan, I see. Thanks for explain. I'll do it.

[ruben-ayrapetyan]

@LaszloLango, thank you.

[galpeter]

We usually test it this way:

assert (e instanceof Error);
assert (e.message === "foo");

but this is also ok imho.

[LaszloLango]

@galpeter, @ruben-ayrapetyan, I've updated the PR.

[ruben-ayrapetyan]

We can use ECMA_SET_NON_NULL_POINTER in the case, as prim_value_num_p is not NULL.

[ruben-ayrapetyan]

Looks good to me

[egavrin]

make push

[galpeter]

lgtm