Implement Base.instantiate - take 2

[tiemvanderdeure]

I'm having another go at this, I'll run Rasters tests as well this time to make totally sure we're not breaking anything.

#917 introduced a BasicDimensionalStyle which I don't think we need at all because we can just let similar dispatch do the work.

[codecov]

Codecov Report

❌ Patch coverage is 95.12195% with 2 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (breaking@fd8f0ef). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
src/array/broadcast.jl	97.50%	1 Missing ⚠️
src/opaque.jl	0.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##             breaking    #1118   +/-   ##
===========================================
  Coverage            ?   82.66%           
===========================================
  Files               ?       57           
  Lines               ?     5631           
  Branches            ?        0           
===========================================
  Hits                ?     4655           
  Misses              ?      976           
  Partials            ?        0           

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[tiemvanderdeure]

I tried Rasters tests on this branch locally and it doesn't seem to break anything. But Rasters tests are broken right now, and there are some other things on this branch that broke it, so I'm not 100% sure.

[rafaqz]

yeah, maybe lets fix those tests first. I guess its the threading on 1.12?

[tiemvanderdeure]

yeah, maybe lets fix those tests first. I guess its the threading on 1.12?

The threading is one thing and then the other breaking changes, like removing index and constructors for AbstractDimArray

[tiemvanderdeure]

I just ran Rasters test to locally with this branch, rafaqz/Rasters.jl#1030, rafaqz/Rasters.jl#1029 and some very small local changes to dispatch on Raster{T}(undef, ...).

The only failure is with some I/O things with Rasters filed with strings - I think that might be because I'm on windows and I am sure it has nothing to do with this PR.

So this is safe to merge AFAIC

[rafaqz]

Hmm I will wait until you check on linux we do use broadcasts in write

[tiemvanderdeure]

Okay actually the same test does pass on main, so I have to investigate a little more.

The thing is I branched this of breaking so there are a whole bunch of breaking changes in here and any of them might cause an error.

This is the test that is failing:

https://github.com/rafaqz/Rasters.jl/blob/9404d72372ca7ddff65097ecc34c61cf21ce2826/test/sources/ncdatasets.jl#L446-L451

[tiemvanderdeure]

It was actually a bug (though pretty niche) - see JuliaIO/DiskArrays.jl#284