Allow empty HTML paths

2 files changed, 13 insertions, 2 deletions

diff --git a/mammoth/html_paths.py b/mammoth/html_paths.py
index 887288b..1428015 100644
--- a/mammoth/html_paths.py
+++ b/mammoth/html_paths.py
@@ -15,3 +15,5 @@ def element(names, class_names=None, fresh=None):
 
 HtmlPath = collections.namedtuple("HtmlPath", ["elements"])
 HtmlPathElement = collections.namedtuple("HtmlPathElement", ["names", "class_names", "fresh"])
+
+empty = path([])
diff --git a/mammoth/style_reader/html_path_reader.py b/mammoth/style_reader/html_path_reader.py
index b508f59..20666d4 100644
--- a/mammoth/style_reader/html_path_reader.py
+++ b/mammoth/style_reader/html_path_reader.py
@@ -9,6 +9,13 @@ def read_html_path(string):
 
 
 def read_html_path_node(path_node):
+    if path_node.children:
+        return _read_html_path_elements_node(path_node.children[0])
+    else:
+        return html_paths.empty
+
+
+def _read_html_path_elements_node(path_node):
     elements = [
         _read_element_node(child)
         for child in _repeated_children_with_separator(path_node, has_whitespace=True)
@@ -60,7 +67,9 @@ def _repeated_children_with_separator(node, has_whitespace):
 
 grammar_text = r"""
 
-html_path = element (whitespace* ">" whitespace* element)*
+html_path = html_path_elements?
+
+html_path_elements = element (whitespace* ">" whitespace* element)*
 
 element = tag_names class_name* fresh?
 
@@ -70,7 +79,7 @@ class_name = "." identifier
 
 fresh = ":fresh"
 
-identifier = ~"[A-Z0-9]*"i
+identifier = ~"[A-Z0-9]+"i
 
 whitespace = ~"\s"*