Resolve buffering issues in TeeStream and capture_output

[jsiirola]

Fixes #3084, fixes #3031, fixes #2845.

Summary/Motivation:

The HiGHS interface exposed some issues in the logic in TeeStream and capture_output where the two could end up fighting with each other. This PR should resolve all issues with console log management that we have been seeing with HiGHS.

Changes proposed in this PR:

• Ensure that flushing a TeeStream handle is propagated through tho the target output streams
• Allow passing a TeeStream to capture_output to prevent ending up with nested TeeStream instances
• TeeStream should dup all target file descriptors so that subsequent context managers (like capture_output) don't commandeer them (causing a weird circular reference loop)
• Update how capture_output handles redirection (do not let redirect_fd update the sys.stdout / sys.stderr streams)
• Add tests for buffered and unbuffered output through TeeStream (and capture_output)
• Update all solver interfaces to standardize how we use capture_output / TeeStream for managing the console output.

Legal Acknowledgement

By contributing to this software project, I have read the contribution guide and agree to the following terms and conditions for my contribution:

1. I agree my contributions are submitted under the BSD license.
2. I represent I am authorized to make the contributions and grant the license. If my employer has rights to intellectual property that includes these contributions, I represent that I have received permission to make contributions and grant the required license on behalf of that employer.

[FBumann]

Would love to see this resolved.
Kepp the good work up ;)

[mrmundt]

Would love to see this resolved. Kepp the good work up ;)

Thanks! We have been having lots of issues with our testing infrastructure that we need to fix before we can get any actually functional PRs merged. Stay tuned!