From af6742df1c876cdeb9e408d4d15c345e5c7e2161 Mon Sep 17 00:00:00 2001 From: Lysandros Nikolaou Date: Thu, 11 Jun 2020 02:56:08 +0300 Subject: [PATCH] bpo-40847: Consider a line with only a LINECONT a blank line (GH-20769) A line with only a line continuation character should be considered a blank line at tokenizer level so that only a single NEWLINE token gets emitted. The old parser was working around the issue, but the new parser threw a `SyntaxError` for valid input. For example, an empty line following a line continuation character was interpreted as a `SyntaxError`. Co-authored-by: Pablo Galindo (cherry picked from commit 896f4cf63f9ab93e30572d879a5719d5aa2499fb) Co-authored-by: Lysandros Nikolaou --- Lib/test/test_peg_parser.py | 7 +++++++ Lib/test/test_syntax.py | 14 ++++++++++++++ .../2020-06-09-23-52-32.bpo-40847.4XAACw.rst | 4 ++++ Parser/tokenizer.c | 3 ++- 4 files changed, 27 insertions(+), 1 deletion(-) create mode 100644 Misc/NEWS.d/next/Core and Builtins/2020-06-09-23-52-32.bpo-40847.4XAACw.rst diff --git a/Lib/test/test_peg_parser.py b/Lib/test/test_peg_parser.py index 6ccb2573176bb5..fae85e323da044 100644 --- a/Lib/test/test_peg_parser.py +++ b/Lib/test/test_peg_parser.py @@ -153,6 +153,13 @@ def f(): ('dict_comp', '{x:1 for x in a}'), ('dict_comp_if', '{x:1+2 for x in a if b}'), ('dict_empty', '{}'), + ('empty_line_after_linecont', + r''' + pass + \ + + pass + '''), ('for', ''' for i in a: diff --git a/Lib/test/test_syntax.py b/Lib/test/test_syntax.py index c6f53eb3c88af3..1d10a200945eb4 100644 --- a/Lib/test/test_syntax.py +++ b/Lib/test/test_syntax.py @@ -854,6 +854,20 @@ def test_kwargs_last3(self): "iterable argument unpacking follows " "keyword argument unpacking") + def test_empty_line_after_linecont(self): + # See issue-40847 + s = r"""\ +pass + \ + +pass +""" + try: + compile(s, '', 'exec') + except SyntaxError: + self.fail("Empty line after a line continuation character is valid.") + + def test_main(): support.run_unittest(SyntaxTestCase) from test import test_syntax diff --git a/Misc/NEWS.d/next/Core and Builtins/2020-06-09-23-52-32.bpo-40847.4XAACw.rst b/Misc/NEWS.d/next/Core and Builtins/2020-06-09-23-52-32.bpo-40847.4XAACw.rst new file mode 100644 index 00000000000000..0b489f24832159 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2020-06-09-23-52-32.bpo-40847.4XAACw.rst @@ -0,0 +1,4 @@ +Fix a bug where a line with only a line continuation character is not considered a blank line at tokenizer level. +In such cases, more than a single `NEWLINE` token was emitted. The old parser was working around the issue, +but the new parser threw a :exc:`SyntaxError` for valid input due to this. For example, an empty line following +a line continuation character was interpreted as a :exc:`SyntaxError`. diff --git a/Parser/tokenizer.c b/Parser/tokenizer.c index cebfadc8e89f30..d461e4e24e721c 100644 --- a/Parser/tokenizer.c +++ b/Parser/tokenizer.c @@ -1203,8 +1203,9 @@ tok_get(struct tok_state *tok, const char **p_start, const char **p_end) } } tok_backup(tok, c); - if (c == '#' || c == '\n') { + if (c == '#' || c == '\n' || c == '\\') { /* Lines with only whitespace and/or comments + and/or a line continuation character shouldn't affect the indentation and are not passed to the parser as NEWLINE tokens, except *totally* empty lines in interactive