add the possibility to use custom hash function

[studyproject-mlds]

Add the possibility to use custom hash function

joblib.Memory("cache_dir", hash_func=my_custom_hash_func)

[codecov]

Codecov Report

Merging #1130 (a7037e3) into master (53a8173) will increase coverage by 0.37%.
The diff coverage is 85.71%.

@@            Coverage Diff             @@
##           master    #1130      +/-   ##
==========================================
+ Coverage   93.93%   94.30%   +0.37%     
==========================================
  Files          47       47              
  Lines        6956     6958       +2     
==========================================
+ Hits         6534     6562      +28     
+ Misses        422      396      -26     

Impacted Files	Coverage Δ
joblib/memory.py	95.58% <85.71%> (+0.02%)	⬆️
joblib/test/test_parallel.py	95.62% <0.00%> (-1.31%)	⬇️
joblib/test/test_memory.py	98.56% <0.00%> (+0.14%)	⬆️
joblib/pool.py	87.80% <0.00%> (+0.81%)	⬆️
joblib/func_inspect.py	92.26% <0.00%> (+1.19%)	⬆️
joblib/logger.py	86.84% <0.00%> (+1.31%)	⬆️
joblib/disk.py	92.06% <0.00%> (+1.58%)	⬆️
joblib/test/test_memmapping.py	99.23% <0.00%> (+1.90%)	⬆️
joblib/_memmapping_reducer.py	96.60% <0.00%> (+2.26%)	⬆️
joblib/backports.py	84.21% <0.00%> (+39.47%)	⬆️

---

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 53a8173...a7037e3. Read the comment docs.

[tomMoral]

Just a few formatting comments.

The change seems quite small so I would say it is okay but could you precise the use case you see for this change?

Most of the when I need to change the hashing, it is for some custom objects and I implement a custom __reduce__ which is sufficient to change the hash behavior. I am curious of use cases that would not be handled by this.

[studyproject-mlds]

Just a few formatting comments.

The change seems quite small so I would say it is okay but could you precise the use case you see for this change?

Most of the when I need to change the hashing, it is for some custom objects and I implement a custom __reduce__ which is sufficient to change the hash behavior. I am curious of use cases that would not be handled by this.

I have a lot of custom objects and I can't implement __reduce__ for all of them.

[judahrand]

I have an alternate implementation in #1232 which I believe adheres more closely to how other custom 'plugin' type functionality works in Joblib. Examples being store backends and compressors.