fix compilation for Visual Studio 2022

[vmoroz]

Fix PR fixes compilation errors and warnings while compiling code in VS 2022.

• async_worker.cc used deprecated std::allocator APIs. Their use is is replaced with APIs that is still valid.
• There was a strange error in ObjectWrap<T>::WrappedMethod. It was saying that & operation requires an l-value. The issue is fixed by copying method pointer to a local variable before invocation.
• VS identified that some fields were not initialized. They are changed to initialize with {}.
• VS identified that some functions must be noexcept. They are augmented with NAPI_NOEXCEPT.

BEGIN_COMMIT_OVERRIDE
fix: fix compilation for Visual Studio 2022 (#1492)
END_COMMIT_OVERRIDE

[codecov-commenter]

Codecov Report

❌ Patch coverage is 0% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 64.71%. Comparing base (4a9d74b) to head (2b52856).
⚠️ Report is 114 commits behind head on main.

Files with missing lines	Patch %	Lines
napi-inl.h	0.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #1492 +/- ##
=======================================
Coverage 64.71% 64.71% =======================================
Files 3 3 Lines 1981 1981 Branches 687 687 =======================================
Hits 1282 1282 Misses 138 138 Partials 561 561 

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[mhdawson]

@vmoroz thanks for the investigation and the PR. One question about the noexcept is that because they are called by other methods that specify noexcept?

I'd have to look closer but adding noexcept I think means it would be breaking to remove those. I understand that if we did make them throw exceptions we'd have to change some other methods to catch those, but the way adding them now would change that to a breaking versus non-breaking change so we should make sure we believe that for those with the additions we believe it would never make sense for them to through an exception. Otherwise we should possibly add a try catch around the calls versus adding the noexcept.

[vmoroz]

@vmoroz thanks for the investigation and the PR. One question about the noexcept is that because they are called by other methods that specify noexcept?

I'd have to look closer but adding noexcept I think means it would be breaking to remove those. I understand that if we did make them throw exceptions we'd have to change some other methods to catch those, but the way adding them now would change that to a breaking versus non-breaking change so we should make sure we believe that for those with the additions we believe it would never make sense for them to through an exception. Otherwise we should possibly add a try catch around the calls versus adding the noexcept.

@mhdawson , it were the recommendations from Visual Studio. The move constructors and assignment operators are typically must be noexcept. In other cases VS saw that methods never through and suggested to make them noexcept.
These changes were just some "good to have". I can remove them from this PR to keep the bare minimum that fixes only compilation errors.

[mhdawson]

@vmoroz I'd say lets keep this PR to just what is needed to fix the compile failures, and then we can look more closely at the good to haves :)

[vmoroz]

@mhdawson , I have removed all newly added NAPI_NOEXCEPT, but I kept the initialization of previously uninitialized variables. They do not cause compilation errors, but I believe they are quite important to avoid undefined behavior.

[mhdawson]

LGTM

[mhdawson]

@vmoroz many thanks. I'm going to go ahead an land this so that we can get the cross platform CI testing green again.