Now supports Python3!

lark/earley.py

@@ -1,6 +1,11 @@
 "My name is Earley"
 
-from .utils import classify
+from .utils import classify, STRING_TYPE
+
+try:
+    xrange
+except NameError:
+    xrange = range
 
 class MatchFailed(object):
     pass
@@ -27,7 +32,7 @@ class State(object):
             self.is_literal = isinstance(self.expect_symbol, dict)
             if self.is_literal:
                 self.expect_symbol = self.expect_symbol['literal']
-            assert isinstance(self.expect_symbol, (str, unicode)), self.expect_symbol
+            assert isinstance(self.expect_symbol, STRING_TYPE), self.expect_symbol
 
     def next_state(self, data):
         return State(self.rule, self.expect+1, self.reference, self.data + [data])
@@ -139,6 +144,6 @@ class Parser(object):
             if (t.rule.name == self.start
                 and t.expect == len(t.rule.symbols)
                 and t.reference == 0
-                and t.data != MatchFailed):
+                and t.data is not MatchFailed):
                 yield t.data
 

lark/grammar_analysis.py

@@ -1,6 +1,7 @@
 from collections import defaultdict, deque
-from utils import classify, classify_bool, bfs, fzset
-from common import GrammarError, is_terminal
+
+from .utils import classify, classify_bool, bfs, fzset
+from .common import GrammarError, is_terminal
 
 ACTION_SHIFT = 0
 
@@ -105,7 +106,7 @@ class GrammarAnalyzer(object):
         """Calculate FOLLOW sets.
 
         Adapted from: http://lara.epfl.ch/w/cc09:algorithm_for_first_and_follow_sets"""
-        symbols = {sym for rule in self.rules for sym in rule.expansion}
+        symbols = {sym for rule in self.rules for sym in rule.expansion} | {rule.origin for rule in self.rules}
         symbols.add('$root')    # what about other unused rules?
 
         # foreach grammar rule X ::= Y(1) ... Y(k)

lark/lexer.py

@@ -1,5 +1,6 @@
 ## Lexer Implementation
-from utils import Str
+
+from .utils import Str
 
 class LexError(Exception):
     pass
@@ -52,7 +53,7 @@ class Lexer(object):
 
         self.mres = []
         self.name_from_index = []
-        x = tokens
+        x = list(tokens)
         while x:
             mre =  re.compile(u'|'.join(u'(?P<%s>%s)'%t for t in x[:LIMIT]))
             self.mres.append(mre)

lark/load_grammar.py

@@ -185,7 +185,7 @@ class SimplifyRule_Visitor(Visitor):
     expansions = _flatten
 
 def dict_update_safe(d1, d2):
-    for k, v in d2.iteritems():
+    for k, v in d2.items():
         assert k not in d1
         d1[k] = v
 
@@ -195,7 +195,8 @@ class RuleTreeToText(Transformer):
         return x
     def expansion(self, symbols):
         return [sym.value for sym in symbols], None
-    def alias(self, ((expansion, _alias), alias)):
+    def alias(self, x):
+        (expansion, _alias), alias = x
         assert _alias is None, (alias, expansion, '-', _alias)
         return expansion, alias.value
 
@@ -294,7 +295,7 @@ class GrammarLoader:
         tokens2 = []
         for name, token, flags in tokens:
             value = token.value[1:-1]
-            if '\u' in value:
+            if r'\u' in value:
                 # XXX for now, you can't mix unicode escaping and unicode characters at the same token
                 value = unicode_escape(value)[0]
             tokens2.append((name, token.type, value, flags))
@@ -406,7 +407,7 @@ def test():
     d: "+" | "-"
     """
     # print load_grammar(g)
-    print GrammarLoader().load_grammar2(g)
+    print(GrammarLoader().load_grammar2(g))
 
 
 if __name__ == '__main__':

lark/parser.py

@@ -1,4 +1,4 @@
-from grammar_analysis import ACTION_SHIFT
+from .grammar_analysis import ACTION_SHIFT
 
 class ParseError(Exception):
     pass

lark/parser_frontends.py

@@ -1,5 +1,5 @@
 from .grammar_analysis import GrammarAnalyzer
-from common import is_terminal
+from .common import is_terminal
 from . import parser, earley
 
 class LALR:

lark/tree.py

@@ -1,5 +1,6 @@
 from copy import deepcopy
-from utils import inline_args
+
+from .utils import inline_args
 
 class Tree(object):
     def __init__(self, data, children):

lark/utils.py

@@ -58,7 +58,7 @@ def inline_args(f):
         def _f_func(self, args):
             return f(self, *args)
         return _f_func
-    elif isinstance(f, (types.TypeType, types.BuiltinFunctionType)):
+    elif isinstance(f, (type, types.BuiltinFunctionType)):
         @functools.wraps(f)
         def _f_builtin(self, args):
             return f(*args)