Retain paragraph type during import

Usfm2osis.py generates paragraphs having many different values for type (such as x-center, x-right, x-noindent etc.). Currently this information is lost when importing to SWORD because during the transformation of <p> to milestone <div type="x-p"> the paragraph type attributes are just dropped. Instead, can they be copied to the div's subType so they can be retained and accessed later, something like:

SWBuf tagName = t.getName();
if (!t.isEndTag()) {
// Transform <p> into <div type="x-p"> and milestone it
if (tagName == "p") {
+ SWBuf subType = t.getAttribute("type");
t.setText("<div type=\"x-p\" />");
sprintf(buf, "gen%d", sID++);
t.setAttribute("sID", buf);
+ if (subType.length())

{t.setAttribute("subType", subType);}

}