support shortest path layout between anchors

[marcocheungkami]

Referring to #119

[pierpo]

Thank you for the PR! It looks like it does the job 😊

I'll have a deeper look in the coming days. Thanks!

[pierpo]

Okay, I had a look! Unfortunately, it does not work properly yet:

• The code does not run. You forgot to declare the shortestPath argument in other places.

• The API is not satisfying, unfortunately. We have two booleans "noCurves", "shortestPath"... But having "noCurves" to "false" and "shortestPath" to "true" does not really make sense. My point is that we should probably have a "drawingMode" props that has multiple states: "shortest-path", "curves", "no-curves" (needs to be thought properly, this is just an idea)

• Even when the above is fixed, there is an offset once the arrow is drawn. Indeed, the path of the arrow isn't as obvious as we think it should be: the arrowhead has some logic to be drawn. We have to draw a line from the start to the beginning minus the arrowhead length (because it is not included in the path). Here the offset on the left element corresponds as the arrowhead height that has been substracted to the top anchor.

I'm afraid this issue is more complex than it looks like 😞
The drawing logic needs a deeper work.

[marcocheungkami]

I see. There is an offset problem needed to be fixed among vertical linkages. Alright, I may try to fix this issue in the coming days.
Guess I need to fine-tune the last two parameters of path d. Also, for the name, I agree with what you have mentioned on drawingModel prop and there will be three modes for people to pick. 👍

[marcocheungkami]

@pierpo I have modified the logic and seems there won't be any extra spacing.

[pierpo]

Thank you for the updates 😊
And sorry for not coming back earlier!

So, it's better indeed, thanks for handling this. But I don't think it's in a mergeable state, because it's still not looking perfect. The fact that the arrow is angled in the end is not really what one would expect 😢

I made a little drawing of what's necessary. It's way more complex.
On the left is what you made, on the right is what we'd aim at.

With the theta angle being atan2(x_starting_elem - xa, y_starting_elem - ya)
I'm not really good with geometry, so there might be simpler calculations!

Anyway, this is quite a hard issue in the end. There's no similar mechanism in the lib.
I always wanted to redraw the drawing logic from scratch to have a more flexible thing. That could've be a part of it.

You can try to add this to the existing logic if you're still motivated 😊

[pierpo]

Closed in favor of #135 😊