Replace TCPThread with RAII ConnectionManager + Incoming/Outgoing threads

[ProfessorTom]

Before submitting a pull request:

• Did you fork from the development branch? yes
• Are you submitting the pull request to the development branch? (not master) yes

This will likely be a Cthulhu PR that will need to be broken up into a few smaller PRs once the work is done. Unfortunately, the way through is a long running branch in the short term.

Some benefits and thoughts as work continues on this branch:

TcpThread class will be replaced with BaseTcpThread, OutgoingTcpThread and IncomingTcpThread.

The idea here is to combine common code in the base class but separate the different directions so that they have room to breath and you can follow the main flows more easily. All three of those classes will automagically clean up both the socket and the threads via RAII and Qt's object lifecycle and parenting system.

There will be a root Connection class, a BaseTcpConnection class, an IncomingTCPConnection and an OutgoingTcpConnection class.

Here's the basic idea:

Connection (abstract)
│
├── BaseTcpConnection
│ ├── OutgoingTcpConnection
│ └── IncomingTcpConnection
│
├── BaseDtlsConnection
│ ├── OutgoingDtlsConnection
│ └── IncomingDtlsConnection
│
├── BaseUdpConnection
│ └── UdpConnection (usually no separate in/out)
│
└── ... (future protocols: RS-232 [aka Serial] QUIC, WebSocket, SCTP, etc.)

---

Unit tests have been added for code I have modified or added and will continue to be added for code I modify or add.

---

Connection state machine

BaseTcpConnection keeps a small explicit state (Created → Active / Inactive / Error / Closing, etc.) updated from the status strings the worker threads emit (connectionStatus). Not every status string we emit is mapped yet—CLion only shows a handful of distinct messages in practice, and several human-readable variants still fall through to a default. That’s intentional for now: the machine is mostly documentation and a single place to reason about lifecycle, not a full protocol state chart. The one place it earns its keep is when we decide whether a connection is still usable for send-queue traffic (e.g. reuse / drain on a live persistent session versus tearing down and creating a new worker). Refining the remaining status → state mappings is straightforward follow-up if we want stricter checks later.

---

qMake is going away.

I'm not sure how Dan builds for each platform.

I was under the impression that most if not all platforms were using CMake now. I know snapcraft and macOS does. Regardless, the unit test only work with CMake, in part, because I am developing on a Mac. There may be some way to get the unit tests to build and run with qMake, but as I understand it, Dan wanted to move away from qMake anyway, so now's the chance.

This means I should be able to delete the .pro files in this PR. At the time of this writing I am not doing so because I don't know what the consequences are for taking that action. e.g. there may be one build that is still using qMake. I talked with @dannagle and he said that Windows stills needs to be converted to CMake. I know that the snapcraft build uses CMake.

Not sure about the Debian build in the pipeline. The "Debian" build that's in the pipeline is not a true Debian build: it uses an Ubuntu container to build the code and then create the .deb package.

I was able to convert the "Debian" build on GitHub to use CMake. This also means running the unit tests, which now pass on both macOS and Ubuntu. I don't have access to a Windows system, so I'm not sure what it would take to get Windows using CMake and if there will need to be any modifications to the unit tests.

Given we are dealing with some low-level functionality (e.g. IPv4 vs IPv6, os default network stacks, etc.), it's possible that some unit tests may fail on Windows. (Presumably snapcraft will build and run, though we may need to modify the script to get snapcraft running the unit tests.)

Fixes #151

[ProfessorTom]

@dannagle I believe this PR is finally ready for review.

Thank you for giving me the opportunity to do this work.