Refactor dataset collection page for ui-v2

[adkinsrs]

This is normally @jorvis's domain, but I will take first crack at this.

[adkinsrs]

Tentatively done with this. It is deployed to gear-devel for testing if anyone wants to try it out.

There are a lot of moving parts, so I certainly can imagine there are bugs or things I forgot to implement or design decisions that need to be reviewed

[jorvis]

Awesome! Does this included respecting the guser.is_admin and guser.is_curator settings yet? We need to elucidate out what those mean still.

[adkinsrs]

Still need to do that. I was more or less referring to things that mirror functionality from v1 and the gene list manager along with various other previously discussed additions

[adkinsrs]

New additions

• Layout arranger is split into single and multi-gene layouts
• Layout arranger now works for adjusting display width and height, as well as moving to different spots on a grid
• Condensed table view of datasets
• More explicit toggle to search all datasets or in the scope of the selected dataset collection
• Toggle to make dataset downloadable. Dataset will also not be downloadable if the h5ad file is not present (per the database flag)
• Various security-related fixes on what database pieces are returned to the client from the CGI calls.
• Confirmation prompt to change dataset to public if a private dataset is added to a public collection.

Known things left to do

• Adding a dataset to a layout multiple times (V2 - Allow dataset to be added to a layout multiple times #687)
• Add some indicator to how many times the dataset is present in the collection
• Layout arranger needs to pick displays instead of datasets (V2 - Allow dataset to be added to a layout multiple times #687)
• Just realized I forget to get image previews working for the layout arranger
• is_admin and is_curator -> Refactor dataset collection page for ui-v2 #671 (comment)
• bug with adding dataset to collection introduced recently
• bug where search terms + share_id input for fetchDatasets() results in an error. Amongst other things, providing a share_id is mutually exclusive to the search filters. I am going to change this, which should allow for filtering within a collection (i.e. if there are 200 datasets for some reason, heaven forbid)

[adkinsrs]

Also need to add the ability to change share IDs (permalinks) for dataset collections as well as individual datasets

This part is done

[adkinsrs]

Last bullet point in the set has been converted into #871 so closing