Make selected img nodes selection transparent about simpla-richtext-behavior HOT 7 CLOSED

Alternative to this would be changing how to handle images. Right now they're inserted into a p tag, but this causes problems anyway as a p tag's schema allows one or more inline elements inside it, which means it can insert an img inside it and then have text either side of it.

What might be better is having embed('image') actually insert a <figure><img/></figure> block, then ProseMirror can select the figure element which we can style as it's a top-level child (or rather, ask the element e.g. simpla-article to style it). This would also fix p being able to contain images and text.

from simpla-richtext-behavior.

You mean the blue outline when you focus on something?

outline: none;

Re: embeding, is there a reason we can't just insert a plain <img> on its own? Why does it have to be wrapped?

from simpla-richtext-behavior.

FWIW can't repro this, whatever it is.

By selecting an image do you mean dragging over it with highlight? That's normal behavior. Way to focus an image should be to tap it.

from simpla-richtext-behavior.

Yeah dragging over it with highlight. If you tap it, it won't focus on the img because it's not an input. ProseMirror handles it by actually 'highlighting' it when you tap on it i.e. simulating dragging over it. Then the 'img' is selected. It's actually a really nice solution because the img is actually highlighted. And it's a general behavior for any future embed.

Doesn't have to wrapped I don't think - I'm just using the default ProseMirror schema which means images get inserted inline inside p. But I can have a play with a custom schema and see how they play sitting by themselves.

from simpla-richtext-behavior.

Fixed in 3f73355, makes image top-level nodes so can be styled through ::slotted

from simpla-richtext-behavior.

FWIW tapping to focus worked for me /shrug

from simpla-richtext-behavior.

Sorry yeah so ProseMirror automatically handles taps / clicks and selects it, but the node itself isn't 'focused' as such, just highlighted / selected by ProseMirror. At least, I'm pretty sure that's how it worked

from simpla-richtext-behavior.