Correct UnitaryForm detection, introduce ClassicalForms

[chseeger]

• Fix bug in UnitaryForm: A root computation over a sometimes incorrect field caused the unitary form of certain matrix groups not to be detected reliably.
• Remove retry loops in UnitaryForm and BilinearForm: Unless I am mistaken, they were unnecessary; the randomization in MeatAxe is not relevant at this point (or am I wrong?). The retry loops only made the functions very unpleasant to use when groups were passed that do not preserve the corresponding form. With up to 1000 loop iterations, their use in ClassicalForms (see below) was thus almost out of the question, because otherwise ClassicalForms would have been unacceptably slow.
• Add ClassicalForms: This function will be used extensively for the implementation of C9-maximal subgroups.

As noted in #86, the functions mentioned should perhaps be renamed in the future to avoid name collisions.

[codecov]

Codecov Report

❌ Patch coverage is 91.56627% with 7 lines in your changes missing coverage. Please review.
✅ Project coverage is 94.74%. Comparing base (08008d1) to head (e2eac39).

Files with missing lines	Patch %	Lines
gap/Forms.gi	91.46%	7 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #139      +/-   ##
==========================================
+ Coverage   94.47%   94.74%   +0.26%     
==========================================
  Files          14       14              
  Lines        4669     4716      +47     
==========================================
+ Hits         4411     4468      +57     
+ Misses        258      248      -10     

Files with missing lines	Coverage Δ
gap/Forms.gd	100.00% <100.00%> (ø)
gap/Forms.gi	92.09% <91.46%> (+4.81%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[chseeger]

CI fails on master due to a recent upstream PR in gap-system/gap (#6203), which changes SetInvariantQuadraticFormFromMatrix to take more than two arguments.