Add parsing for separators

3 files changed, 28 insertions, 3 deletions

diff --git a/mammoth/html_paths.py b/mammoth/html_paths.py
index e201fef..e4f6c31 100644
--- a/mammoth/html_paths.py
+++ b/mammoth/html_paths.py
@@ -7,12 +7,12 @@ def path(elements):
     return HtmlPath(elements)
 
 
-def element(names, class_names=None, fresh=None):
+def element(names, class_names=None, fresh=None, separator=None):
     if class_names is None:
         class_names = []
     if fresh is None:
         fresh = False
-    return HtmlPathElement(names, class_names, fresh)
+    return HtmlPathElement(names, class_names, fresh, separator=separator)
 
 
 @cobble.data
@@ -32,6 +32,7 @@ class HtmlPathElement(object):
     names = cobble.field()
     class_names = cobble.field()
     fresh = cobble.field()
+    separator = cobble.field()
 
     def wrap(self, generate_nodes):
         return self.wrap_nodes(generate_nodes())
diff --git a/mammoth/styles/parser/html_path_parser.py b/mammoth/styles/parser/html_path_parser.py
index e8f448e..9e155b6 100644
--- a/mammoth/styles/parser/html_path_parser.py
+++ b/mammoth/styles/parser/html_path_parser.py
@@ -1,6 +1,6 @@
 from ... import html_paths
 from .tokeniser import TokenType
-from .token_parser import parse_identifier, try_parse_class_name
+from .token_parser import parse_identifier, parse_string, try_parse_class_name
 
 
 def parse_html_path(tokens):
@@ -27,11 +27,13 @@ def _parse_element(tokens):
     tag_names = _parse_tag_names(tokens)
     class_names = _parse_class_names(tokens)
     is_fresh = _parse_is_fresh(tokens)
+    separator = _parse_separator(tokens)
     
     return html_paths.element(
         tag_names,
         class_names=class_names,
         fresh=is_fresh,
+        separator=separator,
     )
 
 
@@ -62,3 +64,17 @@ def _parse_is_fresh(tokens):
         (TokenType.SYMBOL, ":"),
         (TokenType.IDENTIFIER, "fresh"),
     ))
+
+
+def _parse_separator(tokens):
+    is_separator = tokens.try_skip_many((
+        (TokenType.SYMBOL, ":"),
+        (TokenType.IDENTIFIER, "separator"),
+    ))
+    if is_separator:
+        tokens.skip(TokenType.SYMBOL, "(")
+        value = parse_string(tokens)
+        tokens.skip(TokenType.SYMBOL, ")")
+        return value
+    else:
+        return None
diff --git a/tests/styles/parser/html_path_parser_tests.py b/tests/styles/parser/html_path_parser_tests.py
index feaa343..1e34ea1 100644
--- a/tests/styles/parser/html_path_parser_tests.py
+++ b/tests/styles/parser/html_path_parser_tests.py
@@ -70,6 +70,14 @@ def can_read_when_element_must_be_fresh():
 
 
 @istest
+def can_read_separator_for_elements():
+    assert_equal(
+        html_paths.path([html_paths.element(["p"], separator="x")]),
+        read_html_path("p:separator('x')")
+    )
+
+
+@istest
 def can_read_ignore_element():
     assert_equal(
         html_paths.ignore,