Add Cylinder function

[TiagoCavalcante]

Rebase of #1450.

Closes #1294.

[TiagoCavalcante]

@rocky could you review my changes? I put a FIXME here I don't know to transform the current data in a draw around this axis.

[rocky]

@rocky could you review my changes? I put a FIXME here I don't know to transform the current data in a draw around this axis.

Right now there isn't a transform function, but there should be one. In #1373 there is code that needs to be adapted here

[TiagoCavalcante]

@rocky you think it's better we do that before we merge?

Also, what do you think about the odd number of coordinates handling?

[rocky]

Here is a different idea for how to approach this and at the same time go in the direction we want to go in....

Write a to_json() routine for this. But actually, I'd start out with sphere, since that might be easier and here we want to focus on the JSON aspect.

Then we should start a graphics-backend module. (In the past, it was erroneously called a pymathics module, but I realize this was wrong. This is a pure module of any kind that takes JSON with our simple graphics API and converts it into threejs. You could start out with graphics3d.js and/or mathics.js from the Django repository and turn that into a nodejs module (or however you want to package it).

[rocky]

@rocky you think it's better we do that before we merge?

Also, what do you think about the odd number of coordinates handling?

Release 3.0.0 of Mathics core and Django are currently out. Look on PyPI or in "releases" of this project.

[rocky]

Release 3.0.0 of Mathics core and Django are currently out. Look on PyPI or in "releases" of this project.

I am still working on mathics omnibus though.

[TiagoCavalcante]

@rocky you think it's better we do that before we merge?
Also, what do you think about the odd number of coordinates handling?

Release 3.0.0 of Mathics core and Django are currently out. Look on PyPI or in "releases" of this project.

@rocky I didn't find the information I need here.

Write a to_json() routine for this. But actually, I'd start out with sphere, since that might be easier and here we want to focus on the JSON aspect.

Ok, I guess it should be in another PR, isn't it?

Then we should start a graphics-backend module. (In the past, it was erroneously called a pymathics module, but I realize this was wrong. This is a pure module of any kind that takes JSON with our simple graphics API and converts it into threejs. You could start out with graphics3d.js and/or mathics.js from the Django repository and turn that into a nodejs module (or however you want to package it).

@rocky what are we going to use to get the package JavaScript files, Git submodules?

[rocky]

Release 3.0.0 of Mathics core and Django are currently out. Look on PyPI or in "releases" of this project.

@rocky I didn't find the information I need here.

I don't understand what information you need.

All I was saying is that there are new releases: https://pypi.org/project/Mathics3/ (pypi) and https://github.com/mathics/Mathics/releases/tag/3.0.0

What

Write a to_json() routine for this. But actually, I'd start out with sphere, since that might be easier and here we want to focus on the JSON aspect.

Ok, I guess it should be in another PR, isn't it?

Yes, all of this is a new PR. New approach, new PR. And let's try to start simple to work up to cylinders which add this additional wrinkle.

[TiagoCavalcante]

I don't understand what information you need.

Sorry for the confusion.

The information I was needing was what we should do when the user does something like:

Cylinder[{{0,0,0}}] (* it should raise an error because the user didn't provide an end point, just the start one *)

But you already answered here.

[TiagoCavalcante]

I left a FIXME saying we need to do the following transformation_vector * (radius, radius, radius) instead of (1, 1, 1) * (radius, radius, radius). The current version does the plot range be bigger than it should.

As we are going to do the transformation part in other PR I guess this should be merged before that.

Also, the following is the behavior of an odd number of points:

[rocky]

LGTM - I wonder why we get the spurious black style wars.

Did you run black . ?

[TiagoCavalcante]

@rocky no, I'm doing that (I've forgotten).

[rocky]

I suspect then it is the black version that is used. What does black --version say?

[TiagoCavalcante]

@rocky I don't have git pre-commit installed, I guess it's that.

My black version is 21.6b0.

Also, I already pushed a formatted version to this branch.

[rocky]

Ok - Let's set everything to 21.6b0.

I will add a check to the workflows too.