[Lldb-commits] [lldb] [lldb] Don't rely on ScriptInterpreterPythonImpl::Initialize in the unit tests (PR #82096)

[Alex Langford via lldb-commits]

bulbazord wrote:

> The downside of doing the initialization manually is that we do lose a bit of test coverage. For example, issue #70453 also manifested itself in the unit tests.

I think this is an acceptable tradeoff. The unit tests are for testing LLDB's python internals, not for testing LLDB's python interface and integration.

https://github.com/llvm/llvm-project/pull/82096