Fix a few minor contour()-related issues

[adeak]

This is a collection of minor things I came across while debugging #2901.

1. The rng range in contour() should be a sorted (min, max) pair, otherwise we raise.
2. The contour examples in the Flying Edges example should use a single contour value specified explicitly instead of specifying a degenerate range containing the same value twice (the original unsorted values probably only worked due to how VTK handles non-sorted cases).
3. Explicit scalars passed to contour() should be allowed to be lists and other array-likes.
4. The original error in case of missing point scalars was confusing, because it did this:

>>> pv.Sphere().contour(1, scalars='potato')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/adeak/pyvista/pyvista/core/filters/data_set.py", line 1588, in contour
    raise TypeError(
TypeError: Contour filter only works on Point data. Array (potato) is in the Cell data.

Since the same error appears when (i) the scalar name is invalid, and (ii) when it refers to cell scalars, and (iii) when the explicit scalars array passed is consistent with cell scalars, I decided to simplify it just to say that we need point scalars.

[codecov]

Codecov Report

Merging #2951 (d38b8fc) into main (47ddcd9) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #2951   +/-   ##
=======================================
  Coverage   94.30%   94.30%           
=======================================
  Files          76       76           
  Lines       16527    16535    +8     
=======================================
+ Hits        15586    15594    +8     
  Misses        941      941           

[akaszynski]

This is good to go as-is. Coverage will be fixed in #2952.

Recommending immediate merge once all but codecov is passing.

[adeak]

This is good to go as-is. Coverage will be fixed in #2952.

Recommending immediate merge once all but codecov is passing.

I was about to recommend merging yours first because there's probably overlap with this one for the scalars type checking things :) Whichever direction works for you.

(Even codecov should be good now.)

[akaszynski]

I'll merge mine first once approved.

[akaszynski]

Merging this with the hope to get 0.35.0 out by the end of the day. This breaks the 24H rule, but as this is a small bug fix, I'm sure we'll be forgiven.

[adeak]

For what it's worth I consider "impending release" to be an exception to review pedantry, see #2245 (reply in thread).