Improve handling of MongoDB ObjectIDs as OPTIMADE immutable_id

[ml-evs]

Currently, we loop over the entire cursor and check whether the _id projection was requested, which is very inefficient. This PR makes some fixes around that:

• Use native MongoDB toString projection if _id is requested (and only do the inefficient loop if using mongomock).
• Tidy up the Mongo-specific param handling into a method
• Loosen the validator to not require comparison operators on the immutable_id field (except equality).
• Automatically cast the immutable_id field to a string in the models.
• Use the generated MongoDB ObjectID as the immutable_id in our test cases.

[codecov]

Codecov Report

Merging #1142 (3a03ca3) into master (d9f0ab0) will increase coverage by 0.09%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #1142      +/-   ##
==========================================
+ Coverage   92.32%   92.41%   +0.09%     
==========================================
  Files          68       68              
  Lines        3894     3904      +10     
==========================================
+ Hits         3595     3608      +13     
+ Misses        299      296       -3     

Flag	Coverage Δ
project	92.41% <100.00%> (+0.09%)	⬆️
validator	92.41% <100.00%> (+0.09%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...made/server/entry_collections/entry_collections.py	97.60% <ø> (-0.04%)	⬇️
optimade/validator/config.py	100.00% <ø> (ø)
optimade/models/entries.py	97.77% <100.00%> (+0.27%)	⬆️
optimade/server/entry_collections/mongo.py	96.77% <100.00%> (+2.22%)	⬆️
optimade/validator/validator.py	84.03% <0.00%> (+0.36%)	⬆️

---

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 d9f0ab0...3a03ca3. Read the comment docs.

[ml-evs]

@JPBergsma would you be able to take a look at this at some point today? I think we should try to get it into v0.17 as it could be quite significant speed-up in certain scenarios

[JPBergsma]

Sorry, I did not see your request to review it sooner.
It got lost in the flood of github messages in my inbox.
I did not see anything wrong with it.

[JPBergsma]

I think you accidentally added a tab here

Suggested change

	# ">=" is fine as all microsecond trimming will round times down
	# ">=" is fine as all microsecond trimming will round times down

[ml-evs]

This was actually intentional to stop the operator in the second line flowing with the inline comments underneath (which are just commented out functionality). Maybe a linebreak/moving the comment above would be clearer

[ml-evs]

Thanks @JPBergsma and np, lots of notifications yesterday 🙃 I'll address your comment and merge without a futher review.