fix: preserve empty spaces in XML parsing and handle temporary wrappers in text node importer (#775)

Co-authored-by: clarencepalmer <cole@rollprogramcole.com>

Author: palmer-cl

packages/super-editor/src/core/super-converter/SuperConverter.js

@@ -163,7 +163,9 @@ class SuperConverter {
   }
 
   parseXmlToJson(xml) {
-    return JSON.parse(xmljs.xml2json(xml, null, 2));
+    // We need to preserve nodes with xml:space="preserve" and only have empty spaces
+    const newXml = xml.replace(/(<w:t xml:space="preserve">)(\s+)(<\/w:t>)/g, '$1[[sdspace]]$2[[sdspace]]$3');
+    return JSON.parse(xmljs.xml2json(newXml, null, 2));
   }
 
   static getStoredSuperdocVersion(docx) {

packages/super-editor/src/core/super-converter/v2/importer/textNodeImporter.js

@@ -16,7 +16,11 @@ export const handleTextNode = (params) => {
 
   // Text nodes have no children. Only text, and there should only be one child
   let text;
-  if (elements.length === 1) text = elements[0].text;
+  if (elements.length === 1) {
+    text = elements[0].text;
+    // Handle the removal of a temporary wrapper that we added to preserve empty spaces
+    text = text.replace(/\[\[sdspace\]\]/g, '');
+  }
   // Word sometimes will have an empty text node with a space attribute, in that case it should be a space
   else if (!elements.length && 'attributes' in node && node.attributes['xml:space'] === 'preserve') {
     text = ' ';