I am writing a autocompleter for a content editable DIV (need to render html content in the text box. So preferred to use contenteditable DIV over TEXTAREA). Now I need to find the cursor position when there is a keyup/keydown/click event in the DIV. So that I can insert the html/text at that position. I am clueless how I can find it by some computation or is there a native browser functionality that would help me find the cursor position in a contententeditable DIV.
If all you want to do is insert some content at the cursor, there's no need to find its position explicitly. The following function will insert a DOM node (element or text node) at the cursor position in all the mainstream desktop browsers:
function insertNodeAtCursor(node) {
var range, html;
if (window.getSelection && window.getSelection().getRangeAt) {
range = window.getSelection().getRangeAt(0);
range.insertNode(node);
} else if (document.selection && document.selection.createRange) {
range = document.selection.createRange();
html = (node.nodeType == 3) ? node.data : node.outerHTML;
range.pasteHTML(html);
}
}
If you would rather insert an HTML string:
function insertHtmlAtCursor(html) {
var range, node;
if (window.getSelection && window.getSelection().getRangeAt) {
range = window.getSelection().getRangeAt(0);
node = range.createContextualFragment(html);
range.insertNode(node);
} else if (document.selection && document.selection.createRange) {
document.selection.createRange().pasteHTML(html);
}
}
UPDATE
Following the OP's comments, I suggest using my own Rangy library, which adds a wrapper to IE TextRange
object that behaves like a DOM Range. A DOM Range consists of a start and end boundary, each of which is expressed in terms of a node and an offset within that node, and a bunch of methods for manipulating the Range. The MDC article should provide some introduction.