* Polished the interface and APIs of C tracer to fully eliminate the middleware `_VizTracer`
* Utilized `__rdtsc()` for timestamp when available, reducing the overhead
* Made it possible to align with Unix timestamp at us level
* Optimized import speed so `viztracer` command can be much faster to invoke
* Allowed custom repr function for function args and return value
* Used `sys.monitoring` for Python3.12+ for better performance
* Updated the documentation to reflect the current implementation