#455 Fix bug with tricky unicode symbols (#1047)

[etc/vim.git] / blib2to3 / pgen2 / tokenize.py

diff --git a/blib2to3/pgen2/tokenize.py b/blib2to3/pgen2/tokenize.py
index 1f51ff0039fd0227df0a5d7f691bbcd684fea795..977548912c00f584ef135036d7185b67f289f47d 100644 (file)
--- a/blib2to3/pgen2/tokenize.py
+++ b/blib2to3/pgen2/tokenize.py
@@ -29,7 +29,7 @@ __author__ = 'Ka-Ping Yee <ping@lfw.org>'
 __credits__ = \
     'GvR, ESR, Tim Peters, Thomas Wouters, Fred Drake, Skip Montanaro'
 
-import re
+import regex as re
 from codecs import BOM_UTF8, lookup
 from blib2to3.pgen2.token import *
 
@@ -89,7 +89,7 @@ String = group(_litprefix + r"'[^\n'\\]*(?:\\.[^\n'\\]*)*'",
 # recognized as two instances of =).
 Operator = group(r"\*\*=?", r">>=?", r"<<=?", r"<>", r"!=",
                  r"//=?", r"->",
-                 r"[+\-*/%&@|^=<>]=?",
+                 r"[+\-*/%&@|^=<>:]=?",
                  r"~")
 
 Bracket = '[][(){}]'
@@ -334,7 +334,7 @@ def untokenize(iterable):
     ut = Untokenizer()
     return ut.untokenize(iterable)
 
-def generate_tokens(readline):
+def generate_tokens(readline, grammar=None):
     """
     The generate_tokens() generator requires one argument, readline, which
     must be a callable object which provides the same interface as the
@@ -356,6 +356,9 @@ def generate_tokens(readline):
     contline = None
     indents = [0]
 
+    # If we know we're parsing 3.7+, we can unconditionally parse `async` and
+    # `await` as keywords.
+    async_keywords = False if grammar is None else grammar.async_keywords
     # 'stashed' and 'async_*' are used for async/await parsing
     stashed = None
     async_def = False
@@ -506,7 +509,7 @@ def generate_tokens(readline):
                         yield (STRING, token, spos, epos, line)
                 elif initial.isidentifier():               # ordinary name
                     if token in ('async', 'await'):
-                        if async_def:
+                        if async_keywords or async_def:
                             yield (ASYNC if token == 'async' else AWAIT,
                                    token, spos, epos, line)
                             continue