Feature Request: Parallel Analysis with Ray and Incremental Caching

[rahlk]

Is your feature request related to a problem? Please describe.
The current implementation processes files sequentially, leading to slow analysis times for large Python codebases. Unchanged files are re-analyzed on every run, and minor code changes trigger full project re-analysis. These issues result in inefficiencies, especially for projects with hundreds or thousands of files.

Describe the solution you'd like
Introduce Ray-based parallelization to leverage multiple CPU cores and potentially multiple machines. Add CLI options for controlling the number of processes:

• --use-ray: Enable Ray-based parallel analysis.
• --nproc=%: Specify the percentage of available CPU cores to use.
• --nproc=all: Use all available CPU cores.

For incremental analysis, add a new argument:

• --file-name=<file>: Analyze only the specified file, leveraging cached results for unchanged files.

Describe alternatives you've considered

• Threading/Multiprocessing: Less scalable than Ray and does not support distributed computing.
• File-level caching only: Would still re-analyze entire files when only one function changes.
• Simple timestamp-based caching: Less reliable than content-based hashing for detecting changes.

Additional context
Expected performance improvements:

• Significant speedup with Ray parallelization on multi-core systems.
• Faster subsequent runs with incremental caching.
• Efficient handling of minor code changes with SHA-based updates.

Implementation should maintain backward compatibility and integrate with the existing cache directory structure.