Read nested HTML element paths

2 files changed, 48 insertions, 6 deletions

diff --git a/mammoth/style_reader.py b/mammoth/style_reader.py
index 0c27bd8..7a7474f 100644
--- a/mammoth/style_reader.py
+++ b/mammoth/style_reader.py
@@ -4,19 +4,53 @@ from . import html_paths
 
 
 def read_html_path(string):
-    root_node = _grammar.parse(string)
-    tag_names = [root_node.children[0].text]
+    path_node = _grammar.parse(string)
+    return _read_path_node(path_node)
     
-    for choice_node in root_node.children[1].children:
-        tag_names.append(choice_node.children[1].text)
+
+
+def _read_path_node(path_node):
+    elements = [
+        _read_element_node(child)
+        for child in _repeated_children_with_separator(path_node, has_whitespace=True)
+    ]
+    return html_paths.path(elements)
+
+
+def _read_element_node(node):
+    return html_paths.element(_read_tag_names_node(node))
+
+
+def _read_tag_names_node(node):
+    return [
+        child.text
+        for child in _repeated_children_with_separator(node, has_whitespace=False)
+    ]
+
+
+def _repeated_children_with_separator(node, has_whitespace):
+    yield node.children[0]
+    
+    if has_whitespace:
+        sequence_node_index = 2
+    else:
+        sequence_node_index = 1
     
-    return html_paths.path([html_paths.element(tag_names)])
+    sequence_node = node.children[sequence_node_index]
+    for child in sequence_node.children:
+        yield child.children[sequence_node_index]
 
 
-_grammar = Grammar("""
+_grammar = Grammar(r"""
+
+path = element whitespace* (">" whitespace* element)*
+
+element = tag_names
 
 tag_names = identifier ("|" identifier)*
 
 identifier = ~"[A-Z0-9]*"i
 
+whitespace = ~"\s"*
+
 """)
diff --git a/tests/style_reader_tests.py b/tests/style_reader_tests.py
index 43bb3db..595ece1 100644
--- a/tests/style_reader_tests.py
+++ b/tests/style_reader_tests.py
@@ -27,3 +27,11 @@ class ReadHtmlPathTests(object):
             html_paths.path([html_paths.element(["ul", "ol", "p"])]),
             read_html_path("ul|ol|p")
         )
+
+    
+    @istest
+    def can_read_nested_elements(self):
+        assert_equal(
+            html_paths.path([html_paths.element(["ul"]), html_paths.element(["li"])]),
+            read_html_path("ul > li")
+        )