Split ev_op/init and ev_op/ome

[AkshatRai07]

A step for #518
In this PR, we:

• Move all quad_ker_* functions to quad_ker.py.
• Update patch file such that all 3 cfuncs are a part of the patch file and do not contribute to dead code.

[Copilot]

Pull request overview

This PR continues the refactor to separate evolution-operator integration kernels (quad_ker*) from operator-matrix-element logic, and updates the “rustify” patch workflow so the low-level (numba.cfunc) callbacks live only in patch files (avoiding dead code in the normal Python sources).

Changes:

• Moved build_ome (and related OME quad kernel plumbing) into src/eko/evolution_operator/quad_ker.py and updated the OME tests to import from the new location.
• Added/updated quad_ker.py.patch and adjusted rustify.sh to apply it, consolidating the 3 callback cfuncs into the patch flow.
• Updated the patch files for __init__.py and operator_matrix_element.py to match the new split.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 10 comments.

Show a summary per file

File	Description
tests/eko/evolution_operator/test_ome.py	Switches build_ome import to come from quad_ker after the move.
src/eko/evolution_operator/quad_ker.py	Hosts build_ome and OME/evolution integration kernels after the split; docstrings updated/need cleanup.
src/eko/evolution_operator/quad_ker.py.patch	New rustify patch that injects the numba.cfunc callbacks for QCD/QED/OME; currently contains invalid sv.exponentiated/sv.expanded references.
src/eko/evolution_operator/operator_matrix_element.py	Removes local OME quad-kernel/build helpers and imports them from quad_ker.
src/eko/evolution_operator/operator_matrix_element.py.patch	Updates the rustify patch for the OME operator module to use the callback approach.
src/eko/evolution_operator/init.py	Drops duplicated kernel code and delegates to quad_ker for integration kernels.
src/eko/evolution_operator/init.py.patch	Updates rustify patch to use low-level callbacks for integration.
rustify.sh	Applies the new quad_ker.py.patch in addition to the existing patches.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[AkshatRai07]

@felixhekhorn most of copilot's suggestions are about doc changes. Should we do that in this PR or leave it for some another day?

[felixhekhorn]

@felixhekhorn most of copilot's suggestions are about doc changes. Should we do that in this PR or leave it for some another day?

Let's fix them while we are on it

[felixhekhorn]

main comment is to reduce the patch file to the minimal amount possible - since that is a dirty hack in the first place, let's keep it minimal. This can be achieved by trivially harmonizing the variable names and copy docstrings/comments

[felixhekhorn]

apart from the pyproject.toml.patch hickup this is fine

the additional patch file is not pretty, I admit, but we still reduce the number of lines by ~500 ...