text-nodes.js

// Upper-case because we compare them to DOM nodeName
let text_nodes = {'ADDRESS': true, 'ARTICLE': true, 'ASIDE': true, 'AUDIO': true, 'BLOCKQUOTE': true, 'BODY': true, 'CANVAS': true, 'DD': true, 'DIV': true, 'DL': true, 'FIELDSET': true, 'FIGCAPTION': true, 'FIGURE': true, 'FOOTER': true, 'FORM': true, 'H1': true, 'H2': true, 'H3': true, 'H4': true, 'H5': true, 'H6': true, 'HEADER': true, 'HGROUP': true, 'HTML': true, 'HR': true, 'LI': true, 'MAIN': true, 'NAV': true, 'NOSCRIPT': true, 'OL': true, 'OUTPUT': true, 'P': true, 'PRE': true, 'SECTION': true, 'TABLE': true, 'TD': true, 'TH': true, 'UL': true, 'VIDEO': true};

function findTextNodes(nodes, filter) {
  let tns = [], wsx = /\S/;

  if (!$.isArray(nodes)) {
    nodes = [nodes];
  }

  function _findTextNodes(node) {
    if (node.nodeType == 3) {
      if (wsx.test(node.nodeValue)) {
        tns.push(node);
      }
    }
    else {
      for (let i=0 ; i < node.childNodes.length ; ++i) {
        if (typeof filter !== 'function' || filter(node.childNodes[i])) {
          _findTextNodes(node.childNodes[i]);
        }
      }
    }
  }

  for (let i=0 ; i<nodes.length ; ++i) {
    if (typeof filter !== 'function' || filter(nodes[i])) {
      _findTextNodes(nodes[i]);
    }
  }
  return tns;
}

/////////////////

      let tns = findTextNodes(dom.body);
      let ns = [];
      for (let i=0 ; i<tns.length ; ++i) {
        let n = tns[i];
        do {
          n = n.parentNode;
        } while(n && n.parentNode && !text_nodes.hasOwnProperty(n.nodeName));

        // Only add unseen parent nodes
        if (ns.indexOf(n) === -1) {
          ns.push(n);
        }
      }

      // Deduplicate found parent nodes, and mark the unique ones for tracking
      let elms = [];
      for (let i=0 ; i<ns.length ; ++i) {
        let p = ns[i];
        do {
          p = p.parentNode;
          if (ns.indexOf(p) !== -1) {
            console.log(['Skipping node with a parent already in the set', ns[i]]);
            ns[i] = null;
            break;
          }
        } while (p && p.parentNode);

        if (ns[i]) {
          ns[i].setAttribute('data-gs-id', elms.length+1);
          elms.push(ns[i]);
        }
      }

      // elms now contains unique block nodes that we want to send