From 118fc8520ade4e3f6c2d21330521dc28a45447ea Mon Sep 17 00:00:00 2001 From: Ray Donnelly Date: Sun, 29 Apr 2018 01:25:53 +0100 Subject: [PATCH] Add compatibility for AST changes in Python 3.7 --- datashape/lexer.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/datashape/lexer.py b/datashape/lexer.py index b3e1935..1e20c6e 100644 --- a/datashape/lexer.py +++ b/datashape/lexer.py @@ -17,7 +17,9 @@ def _str_val(s): # Use the Python parser via the ast module to parse the string, # since the string_escape and unicode_escape codecs do the wrong thing - return ast.parse('u' + s).body[0].value.s + # mode='single' is necessary since Python 3.7, see: + # https://bugs.python.org/issue32911#msg315581 + return ast.parse('u' + s, mode='single').body[0].value.s # A list of the token names, corresponding regex, and value extraction function _tokens = [