More docx source cleaning

Signed-off-by: kthoden <kthoden@gwdg.de>

fix_tei.py

@@ -316,9 +316,13 @@ def cleanup_xml(xml_tree):
 
     metypeset_attrib = xml_tree.findall("//t:*[@meTypesetSize]", namespaces=NS_MAP)
     color_attrib = xml_tree.xpath("//t:hi[contains(@rend, 'color') or contains(@rend, 'background')]", namespaces=NS_MAP)
+    hi_style_attrib = xml_tree.xpath("//t:hi[contains(@style,'font-size')]", namespaces=NS_MAP)
+    xml_preserve = xml_tree.xpath("//*[@xml:space]")
 
     logging.info("Found %s metypesets." % len(metypeset_attrib))
     logging.info("Found %s colour attributes." % len(color_attrib))
+    logging.info(f"Found {len(hi_style_attrib)} hi style attributes.")
+    logging.info(f"Found {len(xml_preserve)} xml:space attributes.")
 
     for attribute in metypeset_attrib:
         logging.info("Number of attributes: %s" % len(attribute.attrib))
@@ -327,6 +331,18 @@ def cleanup_xml(xml_tree):
     for attribute in color_attrib:
         attribute.attrib.pop("rend")
 
+    for attribute in hi_style_attrib:
+        attribute.attrib.pop("style")
+
+    for attribute in xml_preserve:
+        attribute.attrib.pop("{http://www.w3.org/XML/1998/namespace}space")
+
+    empty_rend = xml_tree.xpath("//t:hi[(@rend='')]", namespaces=NS_MAP)
+    logging.info(f"Found {len(empty_rend)} empty rend attributes.")
+
+    for attribute in empty_rend:
+        attribute.attrib.pop("rend")
+
     hi_without_attrib2 = xml_tree.findall("//t:hi", namespaces=NS_MAP)
 
     for attribute in hi_without_attrib2: