sdfstudio dataparser/dataset

[pablovela5620]

Breaking up adding sdfstudio models into nerfstudio from this previous PR, starting with the dataparser and datasets. A few points here, I've basically brought it over exactly as it was written on the original repo. The reason for this being that I wanted to make sure it worked with the data they provided for dtu/replica/scannet/etc... right out of the box and we don't have to do any processing

I think it may also make sense to also add via a new dataset (such as depth dataset) along with a brand new dataparser so that it's compatible with the nerfstudio-parser and can be used with real world collection, this would allow for easier integration with ns-process when using colmap/polycam/etc.

Let me know your thoughts. I think it also makes sense to differentiate depths/normals that are generated from a sensor (such as polycam) v.s. those by a neural network (like the omnidata model that sdfstudio uses)

[tancik]

Running ns-dev-test will fix some of the github actions errors.

[pablovela5620]

Running ns-dev-test will fix some of the github actions errors.

fixed the missing licensing issue but gave this error (which im slightly confused about since I thought we used black for linting?)

[tancik]

These are linting errors that don't get fixed by black. I added comments on how to fix.