Parse ℯ^x as exp(x) instead of 2.718281828459045^x

[klamike]

First, happy new year! 🍾 Wishing you all the best in 2026 😄

---

This PR adds a special case for ^ where the LHS is ℯ (\euler) and the RHS is an AbstractJuMPScalar. It avoids a conversion from Irrational{:ℯ} to Float64. jump-dev/MathOptInterface.jl#2776 is sort of related.

Before this PR:

julia> m = GenericModel{Float32}();

julia> @variable m x;

julia> expr = ℯ^x
2.718281828459045 ^ x

julia> expr.head, expr.args[1], typeof(expr.args[1])
(:^, 2.718281828459045, Float64)

julia> typeof(ℯ)
Irrational{:ℯ}

After this PR:

julia> m = GenericModel{Float32}();

julia> @variable m x;

julia> expr = ℯ^x
exp(x)

julia> expr.head, expr.args[1]
(:exp, x)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.83%. Comparing base (7ebe98c) to head (e28cdfb).
⚠️ Report is 2 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #4101   +/-   ##
=======================================
  Coverage   99.83%   99.83%           
=======================================
  Files          43       43           
  Lines        5910     5915    +5     
=======================================
+ Hits         5900     5905    +5     
  Misses         10       10           

☔ View full report in Codecov by Sentry.
📢 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.

[odow]

Are there any other methods like this that might cause similar troubles?

[klamike]

Are there any other methods like this that might cause similar troubles?

Not that I've run into