Modulate model and view

[JordanMartinez]

A few questions I have

• What should StyledTextAreaModel access-privilege be: public, protected, or package-private? Others could build on top of the model and the view.
• I'm guessing the style-related methods for StyledTextAreaModel should be public (e.g. getInitialStyle(), getInitialParagraphStyle(), getApplyStyle(), getApplyParagraphStyle(), isPreserveStyle()). Is that correct?
• Should the view be only concerned with view-related things? Or do you want to use it as a Proxy to StyledTextAreaModel? In other words, should I remove all the model-related methods from StyledTextArea so that one must access the model to change the model and the view to change the view?

[TomasMikula]

The refactoring looks mostly good, except for some minor things (see below). I'm not sure though that the fix should be to change the Paragraph#equals method. Let's first focus on getting in the refactoring and discuss the fix then.

applyStyle and applyParagraphStyle are not needed in and don't belong to the model. Also, I think the model should not depend on anything JavaFX. Note that, however, applyStyle and applyParagraphStyle do depend on Text and TextFlow.

I also think the interaction with the clipboard should not be in the model (and as a consequence, codecs will not be necessary in the model either).

Why do constructors now take model instead of document? I thought each area will create its own model, given a document.

The Behavior does not have to take model as an argument, but it should get the model from the given area. You are giving the caller of new StyledTextAreaBehavior(...) a chance to screw up by giving area and model that are unrelated.

[JordanMartinez]

The refactoring looks mostly good, except for some minor things (see below). I'm not sure though that the fix should be to change the Paragraph#equals method. Let's first focus on getting in the refactoring and discuss the fix then.

Gotcha. I'll work in the changes tomorrow.

Why do constructors now take model instead of document? I thought each area will create its own model, given a document.

Since it's been a few days since I did this, I don't recall perfectly why I did some things. However, I think that since the initialStyle and applyStyle-related items were placed into StyledTextAreaModel, I thought it'd be easier/more convenient for cloning to use the model itself to create a copy of itself (meaning, it passes the EditableStyledDocument to the newly-constructed model but allows the new model to create its own caret/selection values) and then just use that. In light of your comments about applyStyle, and Behavior, I see that's not what should be done.

Additionally, I think I started to see StyledTextAreaModel more like the model used for an area's content (i.e. EditableStyledDocument) and less like the model used for the specific view. So, that distinction might have also been why I refactored things the way I did.

I also think the interaction with the clipboard should not be in the model (and as a consequence, codecs will not be necessary in the model either).

Ok.

[JordanMartinez]

Ok. I've accounted for your comments in a new branch. See #243.

Closing this one.