reduce method invalidations and latency by remove method of converting PyObject to nothing

[songjhaha]

This PR is trying to solve the high latency when using PyCall in second time like #1052.

We found the method defination Base.convert(::Type{Nothing}, ::MyType) would create many method invalidations, which casues this latency (see JuliaLang/julia#51389), so this PR remove the method convert(::Type{Nothing}, ::PyObject), and some patches by @thautwarm solve the auto converting convert(::Type{PyAny}, o::PyObject).

before this PR:

julia> @time using PyCall
  0.478829 seconds (255.49 k allocations: 18.250 MiB, 1.91% compilation time)

julia> @time using PyCall
  1.957530 seconds (1.29 M allocations: 90.161 MiB, 5.23% gc time, 99.97% compilation time: 100% of which was recompilation)

after this PR:

julia> @time using PyCall
  0.424794 seconds (261.29 k allocations: 18.610 MiB, 2.36% compilation time)

julia> @time using PyCall
  0.000566 seconds (309 allocations: 21.836 KiB)

This latency time is estimated in Julia1.9.3 and win10 OS

[codecov-commenter]

Codecov Report

All modified lines are covered by tests ✅

Comparison is base (2a9f077) 67.30% compared to head (436b314) 67.44%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1055      +/-   ##
==========================================
+ Coverage   67.30%   67.44%   +0.13%     
==========================================
  Files          20       20              
  Lines        2025     2024       -1     
==========================================
+ Hits         1363     1365       +2     
+ Misses        662      659       -3     

Flag	Coverage Δ
unittests	67.44% <100.00%> (+0.13%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
src/conversions.jl	63.32% <100.00%> (+0.58%)	⬆️

... and 2 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[MilesCranmer]

Hey @songjhaha,
Sorry this hasn't been reviewed yet. I just saw this PR while scrolling by and was wondering if this still has an impact on v1.10 or whether JuliaLang/julia#49525 was enough to fix this issue?

I see this locally:

julia> @time using PyCall
  0.311416 seconds (367.73 k allocations: 24.860 MiB, 3.42% gc time, 66.53% compilation time: 98% of which was recompilation)

julia> @time using PyCall
  0.000159 seconds (151 allocations: 13.984 KiB)

Thanks!
Miles