Handle SDTs representing checkboxes wrapped in other elements

5 files changed, 132 insertions, 9 deletions

diff --git a/NEWS b/NEWS
index 6324c8c..68b497d 100644
--- a/NEWS
+++ b/NEWS
@@ -3,6 +3,10 @@
 * Add "Heading" and "Body" styles, as found in documents created by Apple Pages,
   to the default style map.
 
+* Handle structured document tags representing checkboxes wrapped in other
+  elements, such as table cells. Previously, the wrapping elements would have
+  been ignored.
+
 # 1.9.1
 
 * Ignore AlternateContent elements when there is no Fallback element.
diff --git a/mammoth/documents.py b/mammoth/documents.py
index f9c0665..e874775 100644
--- a/mammoth/documents.py
+++ b/mammoth/documents.py
@@ -92,6 +92,9 @@ class TableCellUnmerged:
     rowspan = cobble.field()
     vmerge = cobble.field()
 
+    def copy(self, **kwargs):
+        return cobble.copy(self, **kwargs)
+
 @cobble.data
 class Break(Element):
     break_type = cobble.field()
diff --git a/mammoth/docx/body_xml.py b/mammoth/docx/body_xml.py
index 3de2c60..5cef312 100644
--- a/mammoth/docx/body_xml.py
+++ b/mammoth/docx/body_xml.py
@@ -5,6 +5,7 @@ import sys
 from .. import documents
 from .. import results
 from .. import lists
+from .. import transforms
 from . import complex_fields
 from .dingbats import dingbats
 from .xmlparser import node_types, XmlElement, null_xml_element
@@ -577,17 +578,49 @@ def _create_reader(numbering, content_types, relationships, styles, docx_file, f
         return read_child_elements(element.find_child_or_null("mc:Fallback"))
 
     def read_sdt(element):
-        checkbox = element.find_child_or_null("w:sdtPr").find_child("wordml:checkbox")
+        content_result = read_child_elements(element.find_child_or_null("w:sdtContent"))
+
+        def handle_content(content):
+            # From the WordML standard: https://learn.microsoft.com/en-us/openspecs/office_standards/ms-docx/3350cb64-931f-41f7-8824-f18b2568ce66
+            #
+            # > A CT_SdtCheckbox element that specifies that the parent
+            # > structured document tag is a checkbox when displayed in the
+            # > document. The parent structured document tag contents MUST
+            # > contain a single character and optionally an additional
+            # > character in a deleted run.
+            checkbox = element.find_child_or_null("w:sdtPr").find_child("wordml:checkbox")
+
+            if checkbox is None:
+                return content
 
-        if checkbox is not None:
             checked_element = checkbox.find_child("wordml:checked")
             is_checked = (
                 checked_element is not None and
                 read_boolean_attribute_value(checked_element.attributes.get("wordml:val"))
             )
-            return _success(documents.checkbox(checked=is_checked))
-        else:
-            return read_child_elements(element.find_child_or_null("w:sdtContent"))
+            document_checkbox = documents.checkbox(checked=is_checked)
+
+            has_checkbox = False
+
+            def transform_text(text):
+                nonlocal has_checkbox
+                if len(text.value) > 0 and not has_checkbox:
+                    has_checkbox = True
+                    return document_checkbox
+                else:
+                    return text
+
+            replaced_content = list(map(
+                transforms.element_of_type(documents.Text, transform_text),
+                content,
+            ))
+
+            if has_checkbox:
+                return replaced_content
+            else:
+                return document_checkbox
+
+        return content_result.map(handle_content)
 
     handlers = {
         "w:t": text,
diff --git a/mammoth/transforms.py b/mammoth/transforms.py
index 62bfc1f..cf96cd3 100644
--- a/mammoth/transforms.py
+++ b/mammoth/transforms.py
@@ -15,18 +15,18 @@ def element_of_type(element_type, transform):
             return transform(element)
         else:
             return element
-    
+
     return _each_element(transform_element)
 
 
 def _each_element(transform_element):
     def transform_element_and_children(element):
-        if isinstance(element, documents.HasChildren):
+        if isinstance(element, (documents.HasChildren, documents.TableCellUnmerged)):
             children = list(map(transform_element_and_children, element.children))
             element = element.copy(children=children)
-        
+
         return transform_element(element)
-    
+
     return transform_element_and_children
 
 
diff --git a/tests/docx/body_xml_tests.py b/tests/docx/body_xml_tests.py
index 424c2f5..994583e 100644
--- a/tests/docx/body_xml_tests.py
+++ b/tests/docx/body_xml_tests.py
@@ -788,6 +788,89 @@ class CheckboxTests:
 
         assert_that(result, is_checkbox(checked=True))
 
+    def test_when_structured_document_tag_checkbox_has_sdt_content_then_checkbox_replaces_single_character(self):
+        element = xml_element("w:tbl", {}, [
+            w_tr(
+                xml_element("w:sdt", {}, [
+                    xml_element("w:sdtPr", {}, [
+                        xml_element("wordml:checkbox", {}, [
+                            xml_element("wordml:checked", {"wordml:val": "1"}),
+                        ]),
+                    ]),
+                    xml_element("w:sdtContent", {}, [
+                        xml_element("w:tc", {}, [
+                            xml_element("w:p", {}, [
+                                xml_element("w:r", {}, [
+                                    xml_element("w:t", {}, [
+                                        xml_text("☐"),
+                                    ]),
+                                ]),
+                            ]),
+                        ]),
+                    ]),
+                ]),
+            ),
+        ])
+
+        result = _read_and_get_document_xml_element(element)
+
+        assert_equal(result, documents.table([
+            documents.table_row([
+                documents.table_cell([
+                    documents.paragraph([
+                        documents.run([
+                            documents.checkbox(checked=True),
+                        ]),
+                    ]),
+                ]),
+            ]),
+        ]))
+
+    def test_when_structured_document_tag_checkbox_has_sdt_content_then_deleted_content_is_ignored(self):
+        element = xml_element("w:tbl", {}, [
+            w_tr(
+                xml_element("w:sdt", {}, [
+                    xml_element("w:sdtPr", {}, [
+                        xml_element("wordml:checkbox", {}, [
+                            xml_element("wordml:checked", {"wordml:val": "1"}),
+                        ]),
+                    ]),
+                    xml_element("w:sdtContent", {}, [
+                        xml_element("w:tc", {}, [
+                            xml_element("w:p", {}, [
+                                xml_element("w:r", {}, [
+                                    xml_element("w:t", {}, [
+                                        xml_text("☐"),
+                                    ]),
+                                ]),
+                                xml_element("w:del", {}, [
+                                    xml_element("w:r", {}, [
+                                        xml_element("w:t", {}, [
+                                            xml_text("☐")
+                                        ])
+                                    ])
+                                ]),
+                            ]),
+                        ]),
+                    ]),
+                ]),
+            ),
+        ])
+
+        result = _read_and_get_document_xml_element(element)
+
+        assert_equal(result, documents.table([
+            documents.table_row([
+                documents.table_cell([
+                    documents.paragraph([
+                        documents.run([
+                            documents.checkbox(checked=True),
+                        ]),
+                    ]),
+                ]),
+            ]),
+        ]))
+
     def _complex_field_checkbox_paragraph(self, ff_data_children):
         return xml_element("w:p", {}, [
             xml_element("w:r", {}, [