Support Numpy 1 and Numpy 2

[garrettwrong]

Numpy 2 requires more changes than I recall. I'm thinking maybe I didn't actually get this far before due to depends issues. My mistake.

Because a lot of scientific python libraries are not supporting Numpy2 yet (I can see why, don't hold your breathe!), we should plan to support both 1 and 2 for a while. (See numpy/numpy#26191) . This PR adds that compatibility for copy= behavior which was one of ASPIRE's main blockers. ( grumbles, I understand the ABI compat issues, but I truly don't understand how user API bridges could not have been built ....) Update: I was able to resolve all occurrences covered by tests without the workaround.

There are many dtype issues. I've resolved those covered by unit tests. One I've noted I wanted to come back to, as I'm not satisfied yet (I figured out where to fix it, but not where it came from 😬 ... 🔎 ). I'll look at that before review, but wanted to get this draft up now to run the CI etc.

My main concern is that I can virtually guarantee I'm missing some np2 dtype mismatches that will cause runtime crashes later. Some of these are not going to be from our code (they will likely be fluid as extern packages update). There isn't much I can do about that other than to try and fix them on our side as they come to surface. 🤷‍♂️

[codecov]

Codecov Report

Attention: Patch coverage is 95.65217% with 1 line in your changes missing coverage. Please review.

Project coverage is 90.39%. Comparing base (ab3dec1) to head (5e75e20).
Report is 14 commits behind head on develop.

Files with missing lines	Patch %	Lines
src/aspire/reconstruction/kernel.py	50.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #1194   +/-   ##
========================================
  Coverage    90.38%   90.39%           
========================================
  Files          132      132           
  Lines        13754    13761    +7     
========================================
+ Hits         12432    12439    +7     
  Misses        1322     1322           

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

---

🚨 Try these New Features:

• Flaky Tests Detection - Detect and resolve failed and flaky tests

[garrettwrong]

I'm going to enable this for review just to run the conda CI jobs. Still a draft... at this time not intending to merge.

[garrettwrong]

@j-c-c , lets get this reviewed and into develop asap so we can start running with it to shake out more issues, thanks.

[j-c-c]

LGTM

[janden]

LGTM