Improve error message when failing to find the body element in a document

3 files changed, 33 insertions, 2 deletions

diff --git a/NEWS b/NEWS
index 84d8902..d627a85 100644
--- a/NEWS
+++ b/NEWS
@@ -2,6 +2,8 @@
 
 * Support attributes in HTML paths in style mappings.
 
+* Improve error message when failing to find the body element in a document.
+
 * Drop support for Python 2.7, Python 3.5 and Python 3.6.
 
 # 1.6.0
diff --git a/mammoth/docx/document_xml.py b/mammoth/docx/document_xml.py
index e0d19a4..45905c6 100644
--- a/mammoth/docx/document_xml.py
+++ b/mammoth/docx/document_xml.py
@@ -6,13 +6,17 @@ def read_document_xml_element(
         body_reader,
         notes=None,
         comments=None):
-    
+
     if notes is None:
         notes = []
     if comments is None:
         comments = []
-    
+
     body_element = element.find_child("w:body")
+
+    if body_element is None:
+        raise ValueError("Could not find the body element: are you sure this is a docx file?")
+
     return body_reader.read_all(body_element.children) \
         .map(lambda children: documents.document(
             children,
diff --git a/tests/docx/document_xml_tests.py b/tests/docx/document_xml_tests.py
index a044154..f869749 100644
--- a/tests/docx/document_xml_tests.py
+++ b/tests/docx/document_xml_tests.py
@@ -1,3 +1,5 @@
+import pytest
+
 from mammoth import documents
 from mammoth.docx.xmlparser import element as xml_element, text as xml_text
 from mammoth.docx.document_xml import read_document_xml_element
@@ -5,6 +7,29 @@ from mammoth.docx import body_xml
 from ..testing import assert_equal
 
 
+def test_when_body_element_is_present_then_body_is_read():
+    paragraph_xml = xml_element("w:p", {}, [])
+    body_xml = xml_element("w:body", {}, [paragraph_xml])
+    document_xml = xml_element("w:document", {}, [body_xml])
+
+    document = _read_and_get_document_xml_element(document_xml)
+
+    assert_equal(
+        documents.document([documents.paragraph([])]),
+        document
+    )
+
+
+def test_when_body_element_is_not_present_then_error_is_raised():
+    paragraph_xml = xml_element("w:p", {}, [])
+    body_xml = xml_element("w:body2", {}, [paragraph_xml])
+    document_xml = xml_element("w:document", {}, [body_xml])
+
+    error = pytest.raises(ValueError, lambda: _read_and_get_document_xml_element(document_xml))
+
+    assert_equal(str(error.value), "Could not find the body element: are you sure this is a docx file?")
+
+
 def test_can_read_text_within_document():
     element = _document_element_with_text("Hello!")
     assert_equal(