⚡ Prefer built-in conversions over extensibility for perf

[tmilnthorp]

Fixes #1069

Per @lipchev's comments in #1023, I decided to add the local conversion method back (as private TryToUnit). This will allow better performance for auto-generated conversions before falling back to the extensibility points.

[codecov]

Codecov Report

Merging #1049 (21eca5f) into master (ee058dc) will increase coverage by 0%.
The diff coverage is 97%.

@@           Coverage Diff            @@
##           master   #1049     +/-   ##
========================================
  Coverage      85%     85%             
========================================
  Files         322     322             
  Lines       49261   52319   +3058     
========================================
+ Hits        42075   44893   +2818     
- Misses       7186    7426    +240     

Impacted Files	Coverage Δ
...t/GeneratedCode/Quantities/LinearPowerDensity.g.cs	89% <ø> (+<1%)	⬆️
UnitsNet/GeneratedCode/Quantities/Luminance.g.cs	85% <ø> (+<1%)	⬆️
UnitsNet/GeneratedCode/Quantities/Luminosity.g.cs	86% <ø> (+<1%)	⬆️
...nitsNet/GeneratedCode/Quantities/LuminousFlux.g.cs	77% <ø> (-1%)	⬇️
...et/GeneratedCode/Quantities/LuminousIntensity.g.cs	77% <ø> (-1%)	⬇️
...itsNet/GeneratedCode/Quantities/MagneticField.g.cs	83% <ø> (+<1%)	⬆️
...nitsNet/GeneratedCode/Quantities/MagneticFlux.g.cs	77% <ø> (-1%)	⬇️
...itsNet/GeneratedCode/Quantities/Magnetization.g.cs	77% <ø> (-1%)	⬇️
UnitsNet/GeneratedCode/Quantities/Mass.g.cs	91% <ø> (+<1%)	⬆️
...et/GeneratedCode/Quantities/MassConcentration.g.cs	92% <ø> (+<1%)	⬆️
... and 218 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[tmilnthorp]

@angularsen I will fill in some missing test coverage in this PR.

Also, is there a way to run the branch marking suite on a PR? It would be nice to see proactively rather than after something is already merged.

[tmilnthorp]

Alright. Looks good now.

Can we add the benchmarking as an optional check that has to be manually run?

[angularsen]

Alright. Looks good now.

Can we add the benchmarking as an optional check that has to be manually run?

I don't know how to do that? :D
The checks are optional, the PR can be merged even if they are red.
I believe you can re-trigger the checks via the Checks tab.

I just reduced the codecov precision from 1 to 0, and changed rounding from up to down (default).
Hoping this reduces the pesky failed checks on -0.0 or -0.1% diff.
a2b92ab

[angularsen]

Oh, I didn't read your question right.
Benchmarking, yes I guess we can make it optional. To be honest, I haven't paid much attention to them because we found that the results varied alot from run to run due to running on different VMs I guess.

[angularsen]

@tmilnthorp I disabled the auto benchmarking action. There is another action to manually run benchmarks.
https://github.com/angularsen/UnitsNet/actions

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[angularsen]

@tmilnthorp I'm reopening this in light of #1069 .

I think maybe a fair compromise is to merge this PR, and not allow extensions to override built-in functions to mitigate performance issue. I'm not sure I see a good usecase to override built-in functions except to duct-tape fix bugs, but I would rather merge those fixes into the library.

What do you think?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[angularsen]

Release UnitsNet/4.142.0 · angularsen/UnitsNet