Fix inconsistent hue color mapping in plot.line

[Mr-Neutr0n]

Summary

Fixes #10998

When plot.line receives a hue coordinate with duplicate values (e.g., ["a", "a", "b", "b"]), lines sharing the same hue value now get the same color — matching the behavior of plot.scatter.

Before: ax.plot was called with the full 2D array, so matplotlib assigned each column a new color from the property cycle regardless of hue grouping. With categories = ["a", "a", "b", "b", "c", "c"], all 6 lines got different colors even though there are only 3 unique hue values.

After: Lines are grouped by unique hue value and plotted per-group with a consistent color. The legend is also deduplicated to show one entry per unique hue value.

Changes

• xarray/plot/dataarray_plot.py — In the line() function, instead of passing the full 2D array to ax.plot, iterate over unique hue values and plot each group with a color from the property cycle. Handles both x= and y= orientations (where either xplt_val or yplt_val may be 2D). Respects user-specified color/c kwargs.

Test plan

• All 519 existing tests in test_plot.py pass (including test_line_plot_along_1d_coord, test_2d_line, and all line-related tests)
• Verified with the exact reproduction script from the issue: lines with hue "a" share one color, "b" another, "c" another, and the legend shows 3 deduplicated entries

[ianhi]

Hi @Mr-Neutr0n thanks for the PR! Can you please post a before an after picture of what this produces.

Also did you explore if there was a simpler way to leverage the matplotlib API here, it seems like we've pulled in a lot of logic to work around something that might be easier to fix with a matplotlib argument. It also might not be, but I'd like to be sure.