fix: allow renaming when rebuilding dask-awkward Arrays

[pfackeldey]

This follows the implementation of dask.array.Array: https://github.com/dask/dask/blob/main/dask/array/core.py#L1442

With this PR it is possible to dask.graph_manipulation.clone graphs containing dask_awkward.Arrays.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.47%. Comparing base (8cb8994) to head (7fe9611).
Report is 187 commits behind head on main.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #554      +/-   ##
==========================================
- Coverage   93.06%   92.47%   -0.60%     
==========================================
  Files          23       22       -1     
  Lines        3290     3429     +139     
==========================================
+ Hits         3062     3171     +109     
- Misses        228      258      +30     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[martindurant]

I am attending pydata global right now, but I'll get to this. I notice that coverage reports that your code isn't touched.

[martindurant]

It looks like this is ready? It maybe should have an explicit test for rename. The failures are on a nearly-equal that seems to pop up occasionally (based on random input). I don't see why anything here would particularly trigger them; but maybe it would be worth while making the assert-nearly-equal give us more information about what doesn't match, to see if there's something systematic we should fix.

[pfackeldey]

Yes, it's ready 👍
I'll add a test later, I'm just not sure what you mean with 'assert-nearly-equal'?

[martindurant]

what you mean with 'assert-nearly-equal'?

I meant calling assert ak.almost_equal within assert_eq_arrays .

...but actually I see that a_comp is all zeros, so something else is going on?