Fix docs build: import FunctionMap/Rodas4P for OrdinaryDiffEq v7

[ChrisRackauckas-Claude]

Problem

The Documentation / Build and Deploy Documentation check on master is failing. The Documenter build terminates with [:example_block] because every @example tpp-advanced block after the intensity/integrated_intensity definitions in docs/src/applications/advanced_point_process.md throws:

UndefVarError: `FunctionMap` not defined in `Main.var"__atexample__named__tpp-advanced"`
UndefVarError: `Rodas4P` not defined in `Main.var"__atexample__named__tpp-advanced"`

(The single UndefVarError: λ further down is a cascade: λ = intensity(...) never binds because intensity itself fails on FunctionMap.)

Root cause

The docs environment resolves OrdinaryDiffEq v7.0.0. In v7 the package dropped its blanket @reexport and now only exports a curated solver set (Tsit5, Vern6-9, Rosenbrock23, Rodas5P, FBDF, ...). FunctionMap and Rodas4P are no longer reachable via using OrdinaryDiffEq — they live in the OrdinaryDiffEqFunctionMap and OrdinaryDiffEqRosenbrock subpackages. The tutorial uses both, so the build broke once v7 entered the docs resolve. The main package already migrated for v7 (it added OrdinaryDiffEqFunctionMap to its test deps); the docs environment was simply not updated.

Fix

• Add OrdinaryDiffEqFunctionMap and OrdinaryDiffEqRosenbrock to docs/Project.toml (deps + compat).
• Import the specific solvers in the tutorial: using OrdinaryDiffEqFunctionMap: FunctionMap and using OrdinaryDiffEqRosenbrock: Rodas4P.

No solver algorithm or prose change — Rodas4P is still registered/exported by OrdinaryDiffEqRosenbrock v2, so the documented behavior is preserved exactly.

Local verification

Against the exact CI resolve (OrdinaryDiffEq v7.0.0, OrdinaryDiffEqFunctionMap v2.0.0, OrdinaryDiffEqRosenbrock v2.3.0) on Julia 1.10, in the docs/ environment:

• Pkg.instantiate() resolves cleanly with the new deps.
• Reproducing the exact failing patterns:
  • solve(dprob, FunctionMap()) on a DiscreteProblem with a DiscreteCallback -> retcode = Success, 23 saved points.
  • solve(oprob, Rodas4P()) on a stiff ODEProblem -> retcode = Success.
• JuliaFormatter (sciml style, format_markdown=true) leaves the edited import block unchanged.

Please ignore until reviewed by @ChrisRackauckas.