v.geometry: new tool (v.to.db wrapper) for more obvious access to vector geometry computation

[petrasovaa]

When somebody wants to compute perimeter/lengths/bbox of vector features, it's difficult to find the tool because v.to.db's name doesn't convey that. Plus for many data science workflows printing it and putting in dataframe is easier than dealing with attribute table. The v.to.db's interface is centered around the table, e.g. column is required even when just printing. Here is a quick summary of the tool:

• New tool v.geometry: read-only, JSON-first frontend to v.to.db for printing vector geometry metrics (area, perimeter, length, count, compactness, fractal_dimension, slope, sinuosity, azimuth, coordinates, start, end, bbox).
• Multiple metrics in one call are dispatched in parallel (ThreadPoolExecutor, controlled by nprocs) and merged by category; results stay deterministic in metric order.
• Output formats: json (default), csv (comma default), plain (pipe default); positional per-metric units.
• Rejects mixing metrics from different feature-type families (area/line/point) to avoid silent cat-collision merges; count is universal.

It includes functions resolving number of cores, this is something that will need to go into a library I think, but I kept it here for now.

Before:

v.to.db map=geology option=bbox column=unused -p format=json

Instead using v.geometry:

v.to.db map=geology option=bbox