Add a blurb to the `Napi::Array` documentation that refers readers to
`Napi::TypedArray` and `Napi::ArrayBuffer` for using arrays with large
amounts of data.

PR-URL: #465
Reviewed-By: Michael Dawson <[email protected]>
 Reviewed-By: Nicola Del Gobbo <[email protected]>

PR-URL: #481
Reviewed-By: Michael Dawson <[email protected]>
Reviewed-By: NickNaso <[email protected]>

PR-URL: #478
Reviewed-By: Anna Henningsen <[email protected]>
Reviewed-By: Gabriel Schulhof <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

PR-URL: #470
Reviewed-By: Michael Dawson <[email protected]>
Reviewed-By: Hitesh Kanwathirtha <[email protected]>

PR-URL: #486
Reviewed-By: Michael Dawson <[email protected]>
Reviewed-By: NickNaso <[email protected]>

`AsyncWorker` contained the assumption that instances of its subclasses
were allocated using `new`, because it unconditionally destroyed
instances using `delete`.

This change replaces the call to `delete` with a call to a protected
instance method `Destroy()`, which can be overridden by subclasses.
This ensures that users can employ their own allocators when
creating `AsyncWorker` subclass instances because they can override
the `Destroy()` method to use their deallocator of choice.

Re: #231 (comment)
PR-URL: #488
Reviewed-By: Michael Dawson <[email protected]>

Refs: #485

The test case was relyingon the ordering of
"for in" which is not guarranteed as per
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/for...in

Update the testcase to check in an way that does
not depend on ordering.

PR-URL: #495
Refs: #485
Reviewed-By: Gabriel Schulhof <[email protected]>
Reviewed-By: NickNaso <[email protected]>

PR-URL: #490
Reviewed-By: Michael Dawson <[email protected]>

`AsyncWorker` assumed that after work is complete,  a JavaScript
 callback would need to execute.

This change removes the restriction of specifying a `Function` callback,
and instead replaces it with an `Env` parameter. Since the purpose of
`receiver` was for the `this` context for the callback, it has also been
removed from the constructors.

Re: #231 (comment)

PR-URL: #489
Reviewed-By: Michael Dawson <[email protected]>

PR-URL: #493
Reviewed-By: Michael Dawson <[email protected]>

This PR is implementing ThreadSafeFunction class wraps
napi_threadsafe_function features.

FYI, the test files that included in this PR have come from Node.js
repo[1]. They've been rewritten based on C   and node-addon-api.

[1] https://github.com/nodejs/node/tree/e800f9d/test/node-api/test_threadsafe_function

PR-URL: #442
Fixes: #312
Reviewed-By: Michael Dawson <[email protected]>
Reviewed-By: Gabriel Schulhof <[email protected]>

Errors in `NAPI_THROW()` and Co. were being thrown as `Error:New(...)`
but they should be thrown as `Napi::Error::New(...)` because the former
assumes that the user has opted to declare `using namespace Napi;`.

PR-URL: #506
Reviewed-By: Nicola Del Gobbo <[email protected]>
Reviewed-By: Gus Caplan <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

PR-URL: #507
Reviewed-By: Nicola Del Gobbo <[email protected]>
Reviewed-By: Gabriel Schulhof <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

* doc: document ThreadSafeFunction

PR-URL: #494
Reviewed-By: Gabriel Schulhof <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

PR-URL: #510
Reviewed-By: Gabriel Schulhof <[email protected]>
Reviewed-By: NickNaso <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

PR-URL: #513
Reviewed-By: Gabriel Schulhof <[email protected]>
Reviewed-By: NickNaso <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

Adds an overridable `GetResult()` method, providing arguments
to the callback invoked in `OnOK()`.

Refs: #231 (comment)
PR-URL: #512
Reviewed-By: Michael Dawson <[email protected]>
Reviewed-By: NickNaso <[email protected]>
Reviewed-By: Gabriel Schulhof <[email protected]>