Avoid parsing URI when replacing fragment

2 files changed, 9 insertions, 3 deletions

diff --git a/mammoth/docx/body_xml.py b/mammoth/docx/body_xml.py
index 0724ac8..cc4e98f 100644
--- a/mammoth/docx/body_xml.py
+++ b/mammoth/docx/body_xml.py
@@ -4,11 +4,10 @@ import re
 from .. import documents
 from .. import results
 from .. import lists
-from ..compat import urldefrag
 from . import complex_fields
 from .xmlparser import node_types, XmlElement
 from .styles_xml import Styles
-from .uris import uri_to_zip_entry_name
+from .uris import replace_fragment, uri_to_zip_entry_name
 
 
 def reader(
@@ -296,7 +295,7 @@ def _create_reader(numbering, content_types, relationships, styles, docx_file, f
         if relationship_id is not None:
             href = relationships[relationship_id].target
             if anchor is not None:
-                href = urldefrag(href)[0] + "#" + anchor
+                href = replace_fragment(href, anchor)
             
             return children_result.map(lambda children: documents.hyperlink(href=href, children=children))
         elif anchor is not None:
diff --git a/mammoth/docx/uris.py b/mammoth/docx/uris.py
index 59f3ffe..892078e 100644
--- a/mammoth/docx/uris.py
+++ b/mammoth/docx/uris.py
@@ -3,3 +3,10 @@ def uri_to_zip_entry_name(base, uri):
         return uri[1:]
     else:
         return base + "/" + uri
+
+
+def replace_fragment(uri, fragment):
+    hash_index = uri.find("#")
+    if hash_index != -1:
+        uri = uri[:hash_index]
+    return uri + "#" + fragment