feat: Support custom and default titles in both graphics and plot objects.

[prmukherj]

All plots now come with a default title. These titles can be over-written by a user-defined title passed as argument during adding a graphics or plot as shown.

    win = GraphicsWindow()
    win.add_graphics(pathlines, position=(0, 0))
    win.add_plot(residual, position=(0, 1), title="Residual History")
    win.show()

AND,

    graphics_window = GraphicsWindow()

    mesh = Mesh(show_edges=True, surfaces=WallBoundaries())
    graphics_window.add_graphics(mesh, position=(0, 0), title="Default Mesh")

    mesh = Mesh(surfaces=WallBoundaries())
    graphics_window.add_graphics(mesh, position=(0, 1))

    graphics_window.show()

[seanpearsonuk]

@prmukherj For no title (i.e. overriding the default), does the caller just pass title = ""?

[seanpearsonuk]

Note that Variable Descriptors contain a human-friendly name string (a quality not necessarily guaranteed via the Fluent names). By basing names on Variable Descriptors, we might also be able to take advantage of that framework's localisation features in future.

[prmukherj]

@prmukherj For no title (i.e. overriding the default), does the caller just pass title = ""?

title=None, can also be passed. So we should update the type-hints, right?

[mkundu1]

Not related to the window title, but is the truncated colormap title in the attached images a known issue? Maybe add a github tracker for it

[prmukherj]

Not related to the window title, but is the truncated colormap title in the attached images a known issue? Maybe add a github tracker for it

Created an issue: #638

Should be a simple investigative fix from PyVista.

[seanpearsonuk]

@prmukherj For no title (i.e. overriding the default), does the caller just pass title = ""?

title=None, can also be passed. So we should update the type-hints, right?

@prmukherj, Please could you answer the question, which was 'For no title (i.e. overriding the default), does the caller just pass title = ""?'. You gave the answer "title=None, can also be passed.", which is unsatisfactory, in my opinion, as you didn't say anything about what title=None is meant to indicate, and that leaves me confused. Moreover, since the implementation is able to generate default titles and the title argument is provided as an optional override, that would seem to imply that the title argument itself is probably chosen to be None by default in order for the implementation to cleanly distinguish between the different use cases. Are you now actually suggesting instead that title has some other default? Either way, I don't understand.

[prmukherj]

@prmukherj For no title (i.e. overriding the default), does the caller just pass title = ""?

title=None, can also be passed. So we should update the type-hints, right?

@prmukherj, Please could you answer the question, which was 'For no title (i.e. overriding the default), does the caller just pass title = ""?'. You gave the answer "title=None, can also be passed.", which is unsatisfactory, in my opinion, as you didn't say anything about what title=None is meant to indicate, and that leaves me confused. Moreover, since the implementation is able to generate default titles and the title argument is provided as an optional override, that would seem to imply that the title argument itself is probably chosen to be None by default in order for the implementation to cleanly distinguish between the different use cases. Are you now actually suggesting instead that title has some other default? Either way, I don't understand.

ok, I got your point now. Passing title="", is same as passing title=None, i.e. it will fall back to the default title. Do we want the feature to behave like passing title="" should overload default?
@seanpearsonuk

[prmukherj]

@prmukherj For no title (i.e. overriding the default), does the caller just pass title = ""?

title=None, can also be passed. So we should update the type-hints, right?

@prmukherj, Please could you answer the question, which was 'For no title (i.e. overriding the default), does the caller just pass title = ""?'. You gave the answer "title=None, can also be passed.", which is unsatisfactory, in my opinion, as you didn't say anything about what title=None is meant to indicate, and that leaves me confused. Moreover, since the implementation is able to generate default titles and the title argument is provided as an optional override, that would seem to imply that the title argument itself is probably chosen to be None by default in order for the implementation to cleanly distinguish between the different use cases. Are you now actually suggesting instead that title has some other default? Either way, I don't understand.

ok, I got your point now. Passing title="", is same as passing title=None, i.e. it will fall back to the default title. Do we want the feature to behave like passing title="" should overload default? @seanpearsonuk

Just to summarize the behavior now:

1. If user don't pass the title argument (it is optional), they get the default title.
2. If user passes None or "" as title still they will get default title.
3. If user passes a custom title then the title will be updated. So, overwriting the title (to show no title) now is only possible by passing " ".

We can handle this by updating the checks so that if user pass "", then also the default title will be over-written with no title if that is required.

[seanpearsonuk]

We can handle this by updating the checks so that if user pass "", then also the default title will be over-written with no title if that is required.

Yes, please. That's what I had in mind.

[prmukherj]

We can handle this by updating the checks so that if user pass "", then also the default title will be over-written with no title if that is required.

Yes, please. That's what I had in mind.

Done:

graphics_window = GraphicsWindow()
graphics_window.add_graphics(mesh, position=(0, 0), title="")
graphics_window.add_graphics(mesh, position=(0, 1))
graphics_window.show()

[seanpearsonuk]

I don’t think I had a clear picture of the state of the code in this repo. I’m approving these changes, but I’ve found them hard to validate through code review alone. I also noticed a considerable amount of duplication in the recent changes.