> How did you capture and render text input, and how did you handle new lines when the text wrapped? How do you handle paragraphs, especially when a user would delete text?
For rendering, I had a component that rendered the block element by splitting the text string at the indicies specified by the text ranges, inserting the appropriate tags, and dangerouslySetInnerHTMLing the text. I'd probably just manually generate vDOM nodes if I were to revisit it.
I didn't do line breaking, moving up/down was done using the caretPositionFromPoint code: create range, get it's bounding box to get initial x/y, getClientRects to get the position for the next line, caretPositionFromPoint, get block text index.
In this case (I did mention it was basic) I wound up not having to do cross-block operations because the UI never had more than one. I had a plan but not having to do it makes me happy.
In conclusion, you're probably better off just using an existing RTE.
For rendering, I had a component that rendered the block element by splitting the text string at the indicies specified by the text ranges, inserting the appropriate tags, and dangerouslySetInnerHTMLing the text. I'd probably just manually generate vDOM nodes if I were to revisit it.
I didn't do line breaking, moving up/down was done using the caretPositionFromPoint code: create range, get it's bounding box to get initial x/y, getClientRects to get the position for the next line, caretPositionFromPoint, get block text index.
In this case (I did mention it was basic) I wound up not having to do cross-block operations because the UI never had more than one. I had a plan but not having to do it makes me happy.
In conclusion, you're probably better off just using an existing RTE.