[refactor] [lang] Restore ODOP support

[archibate]

Related issue = #1827

[Click here for the format server]

---

The remove + restore, refactored away the self.classfunc and self.classkernel property completely.

[codecov]

Codecov Report

Merging #1830 (0cf6101) into archibate-odop (a2c7ee4) will increase coverage by 0.31%.
The diff coverage is 85.00%.

@@                Coverage Diff                 @@
##           archibate-odop    #1830      +/-   ##
==================================================
+ Coverage           43.26%   43.57%   +0.31%     
==================================================
  Files                  44       44              
  Lines                5991     6024      +33     
  Branches             1073     1076       +3     
==================================================
+ Hits                 2592     2625      +33     
- Misses               3245     3248       +3     
+ Partials              154      151       -3     

Impacted Files	Coverage Δ
python/taichi/lang/kernel.py	70.69% <85.00%> (+1.08%)	⬆️

... and 2 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[k-ye]

Again, please feel free to merge to experimental branches, as they impose no risk to master...

[archibate]

No, I will only merge after your approvement.
The point of such workflow is:
each PR into experimental branch is reviewed, so that we don't need review when merging experimental branch into master.
If you don't review this carefully, you'll have to review the final PR that merge into master, with a huge changeset.

[archibate]

Hello? My point for experimental branches is break PRs down. If you don't review the sub-PRs, then the workflow has no difference than a traditional monothonic PR.

[archibate]

Hello? I'm not using this workflow for fun. It requires you to review it, otherwise it's nothing different from traditional PRs.

[CLAassistant]

All committers have signed the CLA.

[bobcao3]

Closing stale PR