DOC: add SIL parachute ejection integration example (#524)

[thatrandomasiandev]

Summary

• Add docs/user/sil_parachute_ejection.rst showing Software-in-the-Loop parachute ejection detection wired through Flight / parachute trigger callables.
• Link the page from the user guide toctree and from the parachute triggers "See Also" section.
• Include a simplified stateful barometric detector, dual-deploy attachment pattern, notes on wrapping compiled recovery firmware, and a cautious HIL mention; points to the Valetudo Monte Carlo inspiration.

Fixes #524

Test plan

• Confirm toctree entry renders under Special Case Simulations
• Skim sil_parachute_ejection.rst for Sphinx link resolution (:doc:, :ref:, :class:)
• Optional: cd docs && make html if the docs env is available

[Gui-FernandesBR]

Two things and then this can go in.

1. The build-docs failure is not yours. It was an invalid RST sublist in StochasticModel.dict_generator's docstring that reached develop through another PR:

rocketpy/stochastic/stochastic_model.py:docstring of ...dict_generator:13: ERROR: Unexpected indentation. [docutils]

I fixed that directly on develop. Merge develop in and the docs job goes green — nothing to change on your side.

2. The "Dual path: trigger vs controller callback" section uses a Markdown table inside an .rst file. RST has no pipe-table syntax: a leading | is parsed as a line block, so this renders as a column of ragged lines rather than a table. Sphinx does not error on it, which is why CI never caught it — but the published page would be wrong. Please convert it to an RST list-table (or a simple table), e.g.

.. list-table::
   :header-rows: 1

   * - Approach
     - When it fires
   * - ``trigger``
     - ...
   * - controller callback
     - ...

[Gui-FernandesBR]

I pushed the table fix to your branch myself (67c262f9) so this doesn't have to wait on another round trip — your rebase had already cleared the build-docs failure, and the table was the only thing left.

What changed: the "Dual path: trigger vs controller callback" section is now a list-table, matching the two-column explanatory tables already in docs/user/flight.rst (:header-rows: 1, :widths: 30 70). Nothing else touched — one file, +15/-5.

For the record, since Sphinx never warned about this and it would have shipped silently, I built the page both ways with Sphinx 8.1.3 under -W --keep-going:

before:  0 <table>,  1 line-block   -> pipes and the "---------" row rendered as ragged text
after :  1 <table class="docutils"> with a real header row

No new warnings either way, which is exactly why CI stayed green on it.

I also scanned the rest of the 245-line file for other Markdown leakage — no fenced code blocks, no [text](url) links, no # headings, no other pipe tables. The one table was the only thing.

Documentation is green on the new head and this is still approved and mergeable, so it's ready to go. Thanks for the write-up — it's a genuinely useful page.