Skip to content

Python tool for converting files and office documents to Markdown.

fix: handle math run with no text child in OMML->LaTeX conversion (#2189)

S1MS4556695de0c2b
diff --git a/packages/markitdown/src/markitdown/converter_utils/docx/math/omml.py b/packages/markitdown/src/markitdown/converter_utils/docx/math/omml.py
index e8158209..f566f7d3 100644
--- a/packages/markitdown/src/markitdown/converter_utils/docx/math/omml.py
+++ b/packages/markitdown/src/markitdown/converter_utils/docx/math/omml.py
@@ -378,7 +378,7 @@ class oMath2Latex(Tag2Method):
@todo \text (latex pure text support)
"""
_str = []
- for s in elm.findtext("./{0}t".format(OMML_NS)):
+ for s in elm.findtext("./{0}t".format(OMML_NS)) or "":
# s = s if isinstance(s,unicode) else unicode(s,'utf-8')
_str.append(self._t_dict.get(s, s))
return escape_latex(BLANK.join(_str))
diff --git a/packages/markitdown/tests/test_docx_omml.py b/packages/markitdown/tests/test_docx_omml.py
new file mode 100644
index 00000000..98574108
--- /dev/null
+++ b/packages/markitdown/tests/test_docx_omml.py
@@ -0,0 +1,48 @@
+# -*- coding: utf-8 -*-
+"""
+Regression test for a crash in the OMML -> LaTeX converter when a math run
+(<m:r>) has no <m:t> text child (e.g. a run that only carries formatting
+properties). Previously `do_r()` called `elm.findtext(...)` directly and
+iterated over the result, which raised `TypeError: 'NoneType' object is not
+iterable` when the run had no text, aborting equation conversion for the
+entire document.
+"""
+
+from xml.etree import ElementTree as ET
+
+from markitdown.converter_utils.docx.math.omml import OMML_NS, oMath2Latex
+
+MATH_NS_DECL = f'xmlns:m="{OMML_NS[1:-1]}"'
+
+
+def _parse_omath(xml_fragment: str):
+ wrapped = f"<m:oMath {MATH_NS_DECL}>{xml_fragment}</m:oMath>"
+ return ET.fromstring(wrapped)
+
+
+def test_run_without_text_child_does_not_crash():
+ # <m:r> with only <m:rPr>, no <m:t> child.
+ element = _parse_omath("<m:r><m:rPr/></m:r>")
+ # Should not raise TypeError: 'NoneType' object is not iterable
+ result = oMath2Latex(element)
+ assert result.latex == ""
+
+
+def test_run_with_text_still_converts():
+ element = _parse_omath("<m:r><m:t>x</m:t></m:r>")
+ result = oMath2Latex(element)
+ assert result.latex == "x"
+
+
+def test_subscript_with_missing_text_run_does_not_crash():
+ # Mirrors a real-world document: a subscript expression where one of the
+ # runs involved has no text (e.g. produced by some Word equation editors).
+ element = _parse_omath(
+ "<m:sSub>"
+ "<m:e><m:r><m:t>l</m:t></m:r></m:e>"
+ "<m:sub><m:r><m:rPr/></m:r><m:r><m:t>1</m:t></m:r></m:sub>"
+ "</m:sSub>"
+ )
+ result = oMath2Latex(element)
+ assert "l" in result.latex
+ assert "1" in result.latex