Fix: cannot extract text from elements without children

3 files changed, 27 insertions, 1 deletions

diff --git a/NEWS b/NEWS
index cff7b21..98f27c0 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,7 @@
+# 0.3.19
+
+* Fix: cannot extract raw text from elements without children
+
 # 0.3.18
 
 * Support links and images in footnotes and endnotes.
diff --git a/mammoth/raw_text.py b/mammoth/raw_text.py
index fb39acf..ccf50ed 100644
--- a/mammoth/raw_text.py
+++ b/mammoth/raw_text.py
@@ -5,7 +5,7 @@ def extract_raw_text_from_element(element):
     if isinstance(element, documents.Text):
         return element.value
     else:
-        text = "".join(map(extract_raw_text_from_element, element.children))
+        text = "".join(map(extract_raw_text_from_element, getattr(element, "children", [])))
         if isinstance(element, documents.Paragraph):
             return text + "\n\n"
         else:
diff --git a/tests/raw_text_tests.py b/tests/raw_text_tests.py
new file mode 100644
index 0000000..c2537c5
--- /dev/null
+++ b/tests/raw_text_tests.py
@@ -0,0 +1,22 @@
+from nose.tools import istest, assert_equal
+
+from mammoth.raw_text import extract_raw_text_from_element
+from mammoth import documents
+
+
+@istest
+def raw_text_of_text_element_is_value():
+    assert_equal("Hello", extract_raw_text_from_element(documents.Text("Hello")))
+
+
+@istest
+def raw_text_of_paragraph_is_terminated_with_newlines():
+    paragraph = documents.paragraph(children=[documents.Text("Hello")])
+    assert_equal("Hello\n\n", extract_raw_text_from_element(paragraph))
+
+
+@istest
+def non_text_element_without_children_has_no_raw_text():
+    tab = documents.Tab()
+    assert not hasattr(tab, "children")
+    assert_equal("", extract_raw_text_from_element(documents.Tab()))