diff --git a/CHANGELOG b/CHANGELOG
index 854cca9..7f6a437 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -2,6 +2,12 @@
Changelog
=========
+* 0.1.7
+ * If the indentation level of a set of lists (with the same list id) were
+ mangled (Starting off with a higher indentation level followed by a
+ lower) then the entire sub list (the list with the lower indentation
+ level) would not be added to the root list. This would result in removing
+ the mangled list from the final output. This issue has been addressed.
* 0.1.6
* Header detection was relying on case. However it is possible for a lower
case version of headers to show up. Those are now handled correctly.
diff --git a/docx2html/core.py b/docx2html/core.py
index 72e1efb..b020cc5 100644
--- a/docx2html/core.py
+++ b/docx2html/core.py
@@ -165,8 +165,7 @@ def is_natural_header(el, styles_dict):
if (
style_id in styles_dict and
'header' in styles_dict[style_id] and
- styles_dict[style_id]['header']
- ):
+ styles_dict[style_id]['header']):
return styles_dict[style_id]['header']
@@ -287,6 +286,7 @@ def get_li_nodes(li, meta_data):
yield li
w_namespace = get_namespace(li, 'w')
current_numId = get_numId(li, w_namespace)
+ starting_ilvl = get_ilvl(li, w_namespace)
el = li
while True:
el = el.getnext()
@@ -301,6 +301,11 @@ def get_li_nodes(li, meta_data):
if _is_top_level_upper_roman(el, meta_data):
break
+ if (
+ is_li(el, meta_data) and
+ (starting_ilvl > get_ilvl(el, w_namespace))):
+ break
+
# If the list id of the next tag is different that the previous that
# means a new list being made (not nested)
if is_last_li(el, meta_data, current_numId):
@@ -309,7 +314,6 @@ def get_li_nodes(li, meta_data):
# Not a subsequent list.
yield el
break
-
yield el
@@ -1018,8 +1022,7 @@ def get_tr_data(tr, meta_data, row_spans):
# ignored.
if (
v_merge is not None and
- v_merge.get('%sval' % w_namespace) != 'restart'
- ):
+ v_merge.get('%sval' % w_namespace) != 'restart'):
continue
# Loop through each and build a list of all the content.
@@ -1072,8 +1075,7 @@ def get_tr_data(tr, meta_data, row_spans):
# here.
if (
v_merge is not None and
- v_merge.get('%sval' % w_namespace) == 'restart'
- ):
+ v_merge.get('%sval' % w_namespace) == 'restart'):
rowspan = next(row_spans)
td_el.set('rowspan', '%d' % rowspan)
diff --git a/docx2html/tests/test_docx.py b/docx2html/tests/test_docx.py
index bb2a566..5863353 100644
--- a/docx2html/tests/test_docx.py
+++ b/docx2html/tests/test_docx.py
@@ -33,8 +33,7 @@ def test_extract_html():
'simple.docx',
)
actual_html = convert(file_path)
- assert_html_equal(actual_html,
- '''
+ assert_html_equal(actual_html, '''
Simple text
@@ -66,8 +65,7 @@ def test_nested_list():
'nested_lists.docx',
)
actual_html = convert(file_path)
- assert_html_equal(actual_html,
- '''
+ assert_html_equal(actual_html, '''
- one
@@ -111,8 +109,7 @@ def test_simple_list():
'simple_lists.docx',
)
actual_html = convert(file_path)
- assert_html_equal(actual_html,
- '''
+ assert_html_equal(actual_html, '''
- One
@@ -132,8 +129,7 @@ def test_inline_tags():
'inline_tags.docx',
)
actual_html = convert(file_path)
- assert_html_equal(actual_html,
- '''
+ assert_html_equal(actual_html, '''
This sentence has some bold, some italics and some underline, as well as a hyperlink.
''') # noqa
diff --git a/docx2html/tests/test_xml.py b/docx2html/tests/test_xml.py
index 586b42e..52bca28 100644
--- a/docx2html/tests/test_xml.py
+++ b/docx2html/tests/test_xml.py
@@ -593,3 +593,32 @@ def test_get_headings(self):
styles_xml = etree.fromstring(xml)
styles_dict = get_style_dict(styles_xml)
self.assertEqual(styles_dict['heading 1']['header'], 'h2')
+
+
+class MangledIlvlTestCase(_TranslationTestCase):
+ expected_output = '''
+
+
+ - AAA
+
+
+ - BBB
+
+
+ - CCC
+
+
+ '''
+
+ def get_xml(self):
+ li_text = [
+ ('AAA', 0, 2),
+ ('BBB', 1, 1),
+ ('CCC', 0, 1),
+ ]
+ lis = ''
+ for text, ilvl, numId in li_text:
+ lis += DXB.li(text=text, ilvl=ilvl, numId=numId)
+
+ xml = DXB.xml(lis)
+ return etree.fromstring(xml)