Description:

Issue plotly/plotly.py#5583 points out that trace x_axis and y_axis setters are not self-explanatory because users have to discover values like "x2" and "y3" from examples.

This PR keeps the existing string-based axis ids, but gives them named types:

• XAxisId
• YAxisId

The trace fields now use those named axis-id string types, and the trace setters accept impl Into<XAxisId> / impl Into<YAxisId> so existing string inputs continue to work without introducing an enum or a hardcoded axis limit.

It also adds regression tests for both code paths:

• derive-generated trace setters (Scatter)
• manual trace setters (Contour)

This PR does not include the CI fixes from #397 which have been moved there unlike earlier.

Checklist:

• I have reviewed all changes in this PR myself.
• I have added relevant regression test cases for this fix.
• I have run formatting checks in WSL using cargo +nightly fmt --all -- --check and cd examples && cargo +nightly fmt --all -- --check.
• I have run linting checks in WSL using cargo clippy -p plotly -p plotly_derive -- -D warnings -A deprecated.
• I have run focused tests in WSL using cargo test -p plotly axis_id --features static_export_chromedriver.
• I have run package tests in WSL using cargo test -p plotly --features plotly_ndarray,plotly_image,static_export_chromedriver,debug.