🛠️ fix(api): restore diff semantics and parser compat

Author: philippb

sdiff/__init__.py

@@ -8,8 +8,6 @@ def diff(md1, md2, renderer=TextRenderer(), parser_cls: type[MdParser] = MdParse
     tree2 = parse(md2, parser_cls)
 
     tree1, tree2, struct_errors = diff_struct(tree1, tree2)
-    tree1, tree2, links_errors = diff_links(tree1, tree2)
-
-    errors = struct_errors + links_errors
+    errors = struct_errors
 
     return renderer.render(tree1), renderer.render(tree2), errors

sdiff/compare.py

@@ -44,9 +44,7 @@ def _diff(tree1, tree2, include_symbols=None, exclude_symbols=None):
 
 
 def diff_links(tree1, tree2):
-    tree1, tree2, errors = _diff(tree1, tree2, exclude_symbols=['t', 'i'])
-    link_errors = [error for error in errors if error.node.symbol == 'a']
-    return tree1, tree2, link_errors
+    return _diff(tree1, tree2, include_symbols=['p', 'h', 'l', 'a'])
 
 
 def diff_struct(tree1, tree2):

sdiff/parser.py

@@ -22,6 +22,8 @@
 
 
 class MdParser:
+    list_rules = None
+
     @classmethod
     def get_lexer(cls):
         return cls()
@@ -30,9 +32,9 @@ def __init__(self):
         self._markdown = mistune.create_markdown(renderer='ast')
         self._reference_definitions = {}
 
-    def parse(self, text):
+    def parse(self, text, rules=None):
         tokens = self._markdown(text)
-        return Root(self._convert_block_tokens(tokens))
+        return self._convert_block_tokens(tokens)
 
     def _set_reference_definitions(self, definitions):
         self._reference_definitions = definitions
@@ -250,9 +252,9 @@ class ZendeskHelpMdParser(MdParser):
     _STEPS_PATTERN = re.compile(r'(?s)<steps>(?P<content>.*?)</steps>')
     _TABS_PATTERN = re.compile(r'(?s)<tabs>(?P<content>.*?)</tabs>')
 
-    def parse(self, text):
+    def parse(self, text, rules=None):
         nodes = self._parse_nodes(text)
-        return Root(nodes)
+        return nodes
 
     def _parse_nodes(self, text: str):
         nodes = []
@@ -425,9 +427,9 @@ def _remove_ltr_rtl_marks(text):
 def parse(text, parser_cls: type[MdParser] = MdParser):
     text = _remove_spaces_from_empty_lines(text)
     text = _remove_ltr_rtl_marks(text)
-    text, reference_definitions = _extract_reference_definitions(text)
     parser = parser_cls()
     if hasattr(parser, '_set_reference_definitions'):
+        text, reference_definitions = _extract_reference_definitions(text)
         parser._set_reference_definitions(reference_definitions)
     result = parser.parse(text)
     if isinstance(result, list):

tests/test_compare.py

@@ -19,6 +19,10 @@ def test_not_equal_links(self):
         _, _, actual = diff_links(trees.pa(), trees.paa())
         self.assertEqual('dummy link 2', actual[0].node.text)
 
+    def test_non_link_structure_diffs_returned(self):
+        _, _, errors = diff_links(trees.r2t(), trees.pt())
+        self.assertTrue(any(error.node.name == 'header' for error in errors))
+
 
 class TestEqual(TestCase):
 

tests/test_parser.py

@@ -263,7 +263,10 @@ def test_remove_ltr_rtl_marks(self):
 
 
 class DummyParser:
-    def parse(self, text):
+    last_text = None
+
+    def parse(self, text, rules=None):
+        DummyParser.last_text = text
         return [Paragraph([Text(text)])]
 
 
@@ -272,3 +275,13 @@ def test_wraps_list_parser_output(self):
         tree = parser.parse('hello', parser_cls=DummyParser)
         self.assertIsInstance(tree, Root)
         self.assertEqual('pt', tree.print_all())
+
+    def test_custom_parser_input_not_mutated_by_ref_defs(self):
+        data = 'See [API][id].\n\n[id]: https://example.com'
+        parser.parse(data, parser_cls=DummyParser)
+        self.assertIn('[id]: https://example.com', DummyParser.last_text)
+
+    def test_mdparser_parse_accepts_rules_argument(self):
+        md_parser = MdParser()
+        nodes = md_parser.parse('1. one', MdParser.list_rules)
+        self.assertIsInstance(nodes, list)

tests/test_sdiff.py

@@ -46,7 +46,9 @@ def test_ignores_link_content(self):
     def test_missing_link_is_reported(self):
         left = 'text [Link](http://example.com)'
         right = 'text'
-        _, _, errors = sdiff.diff(left, right)
+        tree1 = sdiff.parse(left)
+        tree2 = sdiff.parse(right)
+        _, _, errors = sdiff.diff_links(tree1, tree2)
         self.assertTrue(any(error.node.name == 'link' for error in errors))
 
     def test_extra_paragraph_has_paragraph_error(self):