Implement hermitian polynomials

[lijas]

Me and @fredrikekre had som fun and implemented hermitian polynomials 😬

Hermitian polynomials cant use identity-mapping for every second shape functions; instead we have that dNdx = dNd\xi. This means that it is actually the shape functions N that are modified/mapped.

Not sure how to extend to 2d-element :P
Nice to see that the interpolation and cellvalues interface is quite generic :)

Edit: this is only draft PR, will not finish anytime soon since there is currently no need for it.

[codecov]

Codecov Report

Attention: Patch coverage is 88.88889% with 5 lines in your changes missing coverage. Please review.

Project coverage is 93.63%. Comparing base (241d181) to head (ff3465e).

Files with missing lines	Patch %	Lines
src/interpolations.jl	73.33%	4 Missing ⚠️
src/FEValues/FunctionValues.jl	96.55%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1053      +/-   ##
==========================================
- Coverage   93.67%   93.63%   -0.04%     
==========================================
  Files          39       39              
  Lines        6004     6047      +43     
==========================================
+ Hits         5624     5662      +38     
- Misses        380      385       +5     

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

[termi-official]

Not sure how to extend to 2d-element :P

$H(x,y) = H(x) \otimes H(y)$ 🥴

I would suggest that we also should check analytical convergence for harmonic and biharmonic problems in L2 and H1.