Improve arity_assign: ~2x improvement if we share data.

[ritchie46]

This continues upon #1073 (you may close that one if you prefer in one PR, otherwise I rebase).

I did a local benchmark to see what the cost of the new arity_assign would be in the case we share data. In polars expressions that is every first operation:

Gnuplot not found, using plotters backend
bench_1                 time:   [608.34 us 609.30 us 610.73 us]                    
                        change: [+44.741% +45.228% +45.698%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 7 outliers among 100 measurements (7.00%)
  2 (2.00%) high mild
  5 (5.00%) high severe

This showed, that in case of shared data, we first do a memcpy and this is ~2x slower than the kernels where we allocate a new memory buffer.

Luckily we can have best of both worlds. :)

This PR branches depending on the shared state:

• in case we share state -> we alloc (and thus choose the fastest path possible)
• in case we own -> we mutate in place (and thus choose the fastest path possible)

I also added a distinction between set_validity and with_validity as we now have mutable data, it is not always logical return a new PrimitiveArray.

[codecov]

Codecov Report

Merging #1076 (97f3d39) into main (39db6fb) will decrease coverage by 0.05%.
The diff coverage is 62.76%.

@@            Coverage Diff             @@
##             main    #1076      +/-   ##
==========================================
- Coverage   81.01%   80.95%   -0.06%     
==========================================
  Files         366      366              
  Lines       35056    35134      +78     
==========================================
+ Hits        28399    28442      +43     
- Misses       6657     6692      +35     

Impacted Files	Coverage Δ
src/array/primitive/mutable.rs	94.93% <45.45%> (-1.35%)	⬇️
src/compute/arity_assign.rs	72.05% <62.22%> (-20.54%)	⬇️
src/array/primitive/mod.rs	81.65% <66.66%> (-4.06%)	⬇️
src/array/boolean/mod.rs	86.18% <100.00%> (ø)
src/bitmap/utils/slice_iterator.rs	98.78% <0.00%> (+1.21%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 39db6fb...97f3d39. Read the comment docs.

[jorgecarleitao]

Love it! This is great!

[ritchie46]

@jorgecarleitao in light of #1078 and the possible into_iter().collect() over bits call,

I added the allocation branch for the validity as well.