Consistent naming conventions for plural collection objects

[ktaletsk]

Description
Currently, naming conventions endpoints and JSON keys for different collection types are slightly inconsistent when it comes to plural of the collection object (images, csvs, datas). To demonstrate the issue, take a look at the table below:

Collection type	API endpoint	Collection element endpoint	JSON key containing elements of collection
Image Collections	/api/imagesCollections	/api/imagesCollections/{imagesCollectionId}/images	_embedded -> images
CSV Collections	/api/csvCollections	/api/csvCollections/{csvCollectionId}/csv	_embedded -> csvs
Generic Data Collections	/api/genericDatas	/api/genericDatas/{genericDataId}/genericFile	_embedded -> genericFiles

So, you can see that the same word images is used in both the endpoint and JSON key in this case, but different words (csv and csvs, genericFile and genericFiles) are used in other cases.

Proposal
Use the plural form for objects; use the same form in all places in the API. The following naming scheme would be more consistent:

Collection type	API endpoint	Collection element endpoint	JSON key containing elements of collection
Image Collections	/api/imagesCollections	/api/imagesCollections/{imagesCollectionId}/images	_embedded -> images
CSV Collections	/api/csvCollections	/api/csvCollections/{csvCollectionId}/csvs	_embedded -> csvs
Generic Data Collections	/api/genericDataCollections	/api/genericDatas/{genericDataId}/genericFiles	_embedded -> genericFiles

[MyleneSimon]

That's a good point, it seems that it happening for the most recent additions to the API, I will fix the endpoints.
Did you notice any other places besides CSVs and GenericData?
GenericData could probably use another name as well...

[ktaletsk]

I'll probably look through the most of the API endpoints in the near future as I'm working on the Python client for the API (here: PolusAI/data-tools#1). I will report any findings of the inconsistent names in this thread

[mohamedOuladi]

That's a good point, it seems that it happening for the most recent additions to the API, I will fix the endpoints. Did you notice any other places besides CSVs and GenericData? GenericData could probably use another name as well...

In this case, I should rename GenericData in my PR #184. Should we go for GenericDataCollection?

[MyleneSimon]

I am good with GenericDataCollection, this is how it is displayed in the UI for now so it would keep it consistent (and avoid dealing with singular/plural for the word data). If you are updating your PR could you check that all Swagger annotations are properly added/updated? (see #177, you also need to merge #178 in your branch). Thanks!

[MyleneSimon]

If anyone has an idea for a better name though, now is a good time to voice it :)

[MyleneSimon]

I think at some point someone floated the idea of introducing the word raw instead of generic, I would not be opposed to RawDataCollection for example. Another option could be UnstructuredDataCollection.

[pbajcsy]

Following the comment about using plurals, the names should be XXXCollections. I would vote for RawDataCollections (GenericDataCollections would be fine too) since UnstructuredDataCollections is used by text mining researchers for unstructured text.

[mohamedOuladi]

I am good with GenericDataCollection, this is how it is displayed in the UI for now so it would keep it consistent (and avoid dealing with singular/plural for the word data). If you are updating your PR could you check that all Swagger annotations are properly added/updated? (see #177, you also need to merge #178 in your branch). Thanks!

@MyleneSimon Yeah I will make sure that Swagger annotations are updated.

[mohamedOuladi]

I think at some point someone floated the idea of introducing the word raw instead of generic, I would not be opposed to RawDataCollection for example. Another option could be UnstructuredDataCollection.

I think @NHotaling was the one who suggested RawDataCollection. So, I thought I should include him in the conversation.

I have already renamed GenericData to GenericDataCollection in the WIPP-frontend Labshare fork (see PolusAI/WIPP-frontend@94fa379). The reason why I opted for GenericDataCollection and not RawDataCollection was because of the addition of image conversion options that @MyleneSimon was working on. Having a conversion option raw might confuse the user, especially if the raw tags are created for image collections.