Improve performance of openeing files.

[dncnmcdougall]

This update works to improve the reading performance of large files.
See #5.
It does this simply by reading the sqlite databases in a lazy manner.
This also provdes some utilities, like join_frame and range on the analyse object that can be used for more complex queries.
In particular join_frame can be used to join another table to the Frames table along the Frame Id.
For example when reading MALDI data using frame_info = tof_data.analysis.join_frame("MaldiFrameInfo") gives the frame infor and the x and y all at the same time, reducing iteration complexity.

Secondly this adds the tims_read_scans_v2_as_array method which returns the results as a single numpy array. This is slightly faster than tims_read_scans_v2 and yields the data in a manner that is potentially faster for processing.

Unfortunately this PR also has a lot of collateral changes:

1. The code was reformatted with Ruff
2. A uv pyproject.toml was added
3. The licensing was moved over to REUSE