Fix handling of hv.dim expressions in size and color parameters.

[Azaya89]

closes #674
closes #1449

This PR fixes a bug where hv.dim() expressions were not properly handled when passed to the color (or c) and size (or s) parameters, causing errors when plotting.

• Normalized 's' to 'size' and 'c' to 'color' for consistent handling
• Introduced color_dim and color_is_dim variables to validate dim expressions as separate from string column names before handling
• Added tests

Examples

import hvplot.pandas
import numpy as np
import pandas as pd
import holoviews as hv

x = np.linspace(0.0, 10.0, 25)
y = np.sin(x)
df = pd.DataFrame({'x': x, 'y': y})
style_opts = ('size', 's', 'color', 'c')

plots = [
    df.hvplot.scatter(
        'x',
        'y',
        title=f"{opt}=hv.dim('y')",
        height=300,
        width=400,
        **{opt: hv.dim('y') * 20},
    ) for opt in style_opts]

hv.Layout(plots).opts(shared_axes=False).cols(2)