fix(grammar): drop unit self-loop rules to prevent LALR/CYK infinite loop

lark/load_grammar.py

@@ -766,6 +766,9 @@ def compile(self, start, terminals_to_keep) -> Tuple[List[TerminalDef], List[Rul
                 else:
                     exp_options = options
 
+                if len(expansion) == 1 and expansion[0] == NonTerminal(name):
+                    continue
+
                 for sym in expansion:
                     assert isinstance(sym, Symbol)
                     if sym.is_term and exp_options and exp_options.keep_all_tokens:

tests/test_grammar.py

@@ -302,6 +302,11 @@ def test_symbol_eq(self):
 
         self.assertNotEqual(a, b)
 
+    def test_issue_1585_cyk(self):
+        l = Lark('start.1: "a" | start start*', parser='cyk')
+        for s in ('aa', 'aaaa'):
+            self.assertEqual(l.parse(s).data, 'start')
+
 
 if __name__ == '__main__':
     main()

tests/test_parser.py

@@ -72,11 +72,15 @@ def test_infinite_recurse(self):
                a: a | "a"
             """
 
-        self.assertRaises(GrammarError, Lark, g, parser='lalr')
+        for parser in ('lalr', 'earley'):
+            l = Lark(g, parser=parser)
+            tree = l.parse('a')
+            self.assertEqual(tree, Tree('start', [Tree('a', [])]))
 
-        # TODO: should it? shouldn't it?
-        # l = Lark(g, parser='earley', lexer='dynamic')
-        # self.assertRaises(ParseError, l.parse, 'a')
+        g = """start: a
+               a: a
+            """
+        self.assertRaises(GrammarError, Lark, g, parser='lalr')
 
     def test_propagate_positions(self):
         g = Lark("""start: a
@@ -2256,6 +2260,11 @@ def test_prioritization_sum(self):
             self.assertEqual(''.join(child.data for child in res.children), 'indirection')
 
 
+        def test_issue_1585_priority_recursive_star(self):
+            l = _Lark('start.1: "a" | start start*')
+            for s in ('a', 'aa', 'aaaa'):
+                self.assertEqual(l.parse(s).data, 'start')
+
         def test_utf8(self):
             g = u"""start: a
                    a: "±a"