Oliver Stöneberg

This is in preparation of using the shared code in `ThreadExecutor` and `ProcessExecutor`

Use the same logic as `ThreadExecutor` - processes will now be run in a separate thread. This might sound like overhead but increases the responsiveness when a process finishes and...

ProcessExecutor: avoid explicit usage of suppressions

1

The `hasToLog()` call performs this check.

fixed #11372 - try more encodings for Python `open()`/`subprocess.Popen()` calls and do not silently fail

1

This should be a proper solution which should supersede #4603.

fixed #11372 - potential `UnicodeDecodeError` with `subprocess.Popen()`

2

Passing `universal_newlines=True` to `subprocess.Popen()` will implicitly perform `decode()`. But this might encounter errors which need to be ignored. Unfortunately the required options `encoding` and `errors` are not available until Python...

ValueFlow: start splitting it into multiple files

12

`valueflow.cpp` is very monolithic which makes it hard to manage. This is splitting several parts into separate files as discussed in #4642.

removed unused `Token::until()` and `tokenrange.h`

4

This essentially reverts a5664c3e4900f46df5ad774bdc2181ebdb4dc56c. Since we still have no users of this within our code base it doesn't make much sense to keep it. So if there were issues in...