Hide kuzzle object from console.log

[Aschen]

What does this PR do?

Use Reflect to hide the kuzzle object from log

Other changes

• Get ride of every delete keyword and use Map for RAM caches
• Do not throw custom error when unsubscribing from unknown room (let Kuzzle trigger it's own documented error)
• Allows to pass _id to realtime.publish

[codecov]

Codecov Report

Merging #520 into 7-dev will decrease coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##            7-dev     #520      +/-   ##
==========================================
- Coverage   95.59%   95.57%   -0.03%     
==========================================
  Files          32       32              
  Lines        1318     1311       -7     
==========================================
- Hits         1260     1253       -7     
  Misses         58       58              

Impacted Files	Coverage Δ
src/controllers/MemoryStorage.js	96.05% <ø> (-0.20%)	⬇️
src/controllers/Security.js	97.03% <ø> (-0.03%)	⬇️
src/controllers/Base.js	100.00% <100.00%> (ø)
src/controllers/Realtime.js	97.87% <100.00%> (+0.04%)	⬆️
src/core/KuzzleEventEmitter.js	96.42% <100.00%> (+0.06%)	⬆️
src/core/Room.js	100.00% <100.00%> (ø)
src/core/searchResult/SearchResultBase.js	90.19% <100.00%> (ø)
src/core/security/Profile.js	100.00% <100.00%> (ø)
src/core/security/Role.js	80.00% <100.00%> (ø)
src/core/security/User.js	100.00% <100.00%> (ø)

---

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 e6d154d...d4a1d26. Read the comment docs.

[scottinet]

A couple issues:

• you should mention what versions of kuzzle accept this option
• this looks like a circular dependency in disguise; why do we need to release the support for this option before we release the version of kuzzle accepting it?

[Aschen]

A couple issues:

* you should mention what versions of kuzzle accept this option

* this looks like a circular dependency in disguise; why do we need to release the support for this option before we release the version of kuzzle accepting it?

I'm still not sure about what Kuzzle version it's gonna be. It's an enhancement since the SDK is already exposed but I still feel like being able to use it with the HotelClerc features is a big improvement and it may deserve the new-feature tag IMHO, what do you think ?

About the circular dependency, I have no choice unfortunately because I cannot use the query method since I need to use the realtime.subscribe method with the Room mechanism.. (It would have been possible with this feature: #489 😁 )

But in this case it's not a big deal since this option is not recognized by Kuzzle for now.

[scottinet]

What do you think about hiding the documentation (you can left it there, but comment it), so that once Kuzzle is released with that feature, you can uncomment the SDK documentation, adding the right Kuzzle version (and a link to the appropriate guide/API doc)?

I fear that documenting a feature that cannot be used (at the moment) is not a good idea.

[Aschen]

@scottinet I'm OK with that.
And also maybe we should not document this feature at all in the SDK but only in the plugin/application documentation since the options is useless outside a plugin, what do you think?

[Aschen]

@Leodau @scottinet
I removed all the code related to the cluster option since this option is now added by the EmbeddedSDK only.
So this PR is just about few minor stuff now

[scottinet]

Why is this property writable?