feat(utils): include underlines and small-caps in word parser

webern-unibas-ch · May 7, 2024 · 3067408 · 3067408
1 parent 1d2dcdc
commit 3067408
Showing 1 changed file with 4 additions and 0 deletions.
diff --git a/convert_source_description/file_utils.py b/convert_source_description/file_utils.py
@@ -34,6 +34,10 @@ def read_html_from_word_file(self, file_path: str) -> str:
 
         with open(docx_file_name, "rb") as docx_file:
             try:
+                style_map = """
+                u => u
+                small-caps => span.small-caps
+                """
                 result = mammoth.convert_to_html(
                     docx_file, style_map=style_map)