Add up the like terms Expand produces (#164)

[Rafael-SOWNet]

Closes #164 properly. My
earlier PR #666 added a test for it that asserted Simplify, and the issue is about
Expand — that was the wrong test, and the issue rightly stayed open.

(x+1)^2 * (x+2-1)^2

  -3 + 4 + 2*x*(-1) + 2*x*2 + x^2 + 2*x + 2*x*(-4) + 2*x*4 + 2*x*2*x*(-1)
  + 2*x*2*x*2 + 2*x*x^2 + x^2 + x^2*(-4) + x^2*4 + x^2*2*x*(-1)
  + x^2*2*x*2 + x^2^2

Sixteen terms, of which five are distinct. "Cannot be expanded" is exactly right.

What it does

You had both already worked out what this needs, on the issue itself: each term of the
result wants simplifying, but Simplify is far too expensive to call from inside
Expand — "some kind of fast Simplify that unites terms by their common multiplier".

That is what this is, and no more. Each term is reduced to a coefficient and a product of
powers; PowerRules folds (x^2)^2 into x^4 so the two are not counted as different
monomials; terms whose products agree are added together.

(x+1)^2 * (x+2-1)^2   →  1 + 4x + 6x² + 4x³ + x⁴
((x-1)*(x-2))^3       →  8 − 36x + 66x² − 63x³ + 33x⁴ − 9x⁵ + x⁶

The second is the case @Happypig375 asked for in the same thread.

The part that looks wasteful and is not

A term whose coefficients cancel is written out rather than dropped. That is the whole
of the correctness here. The monomial may carry a domain condition:

(4a - 2)/(2x) + (1 - 2a)/x

is zero only where x is not. Both terms are c · x⁻¹, the coefficients cancel, and
dropping the term collected the whole thing to a plain 0 — throwing the condition away.
Written out, InnerSimplified turns 0 · x⁻¹ back into 0 provided not x = 0. An
existing test caught this, and there is now one of my own pinning it.

Four existing expectations change

All to the collected form, which is the point of the change:

Test	was	now
ExpandCollapseTest.Factorial	x^2 + x*3 + 2*x + 6	x^2 + 5*x + 6
PolyParser.TestLinear3	a + 3 + 1	a + 4
IntegrationTest.TestIndefinite		same terms, C first
IntegrationTest.TestLnAbsSquared	`((ln	x

The last two are the same expression written differently; the first two are the
collection working.

Testing

UnitTests 3892 passed / 0 failed, FSharpWrapperUnitTests 127 passed / 0 failed, both
netstandard2.0 and net7.0 build.

12 new tests: both cases from the issue, ordinary products, unlike terms across different
variables being left apart, value preservation at points, and the cancelling-terms case
above.

🤖 Generated with Claude Code

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 96.00000% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 79.99%. Comparing base (90c00a8) to head (91f976d).
⚠️ Report is 51 commits behind head on master.

Files with missing lines	Patch %	Lines
...Math/Functions/Evaluation/Evaluation.Definition.cs	96.00%	1 Missing and 1 partial ⚠️
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #676      +/-   ##
==========================================
- Coverage   80.99%   79.99%   -1.01%     
==========================================
  Files         155      156       +1     
  Lines       13687    12687    -1000     
  Branches     1957     2075     +118     
==========================================
- Hits        11086    10149     -937     
+ Misses       1990     1928      -62     
+ Partials      611      610       -1     

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

🚀 New features to boost your workflow:

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