Issue #22221: Backported fixes from Python 3 (issue #18960).
* Now the source encoding declaration on the second line isn't effective if
the first line contains anything except a comment. This affects compile(),
eval() and exec() too.
* IDLE now ignores the source encoding declaration on the second line if the
first line contains anything except a comment.
* 2to3 and the findnocoding.py script now ignore the source encoding
declaration on the second line if the first line contains anything except
a comment.
diff --git a/Lib/idlelib/IOBinding.py b/Lib/idlelib/IOBinding.py
index aedd372..e3affa8 100644
--- a/Lib/idlelib/IOBinding.py
+++ b/Lib/idlelib/IOBinding.py
@@ -72,6 +72,7 @@
encoding = encoding.lower()
coding_re = re.compile(r'^[ \t\f]*#.*coding[:=][ \t]*([-\w.]+)')
+blank_re = re.compile(r'^[ \t\f]*(?:[#\r\n]|$)')
class EncodingMessage(SimpleDialog):
"Inform user that an encoding declaration is needed."
@@ -130,6 +131,8 @@
match = coding_re.match(line)
if match is not None:
break
+ if not blank_re.match(line):
+ return None
else:
return None
name = match.group(1)
diff --git a/Lib/lib2to3/pgen2/tokenize.py b/Lib/lib2to3/pgen2/tokenize.py
index 4cb2a41..d64a3e6 100644
--- a/Lib/lib2to3/pgen2/tokenize.py
+++ b/Lib/lib2to3/pgen2/tokenize.py
@@ -237,6 +237,7 @@
toks_append(tokval)
cookie_re = re.compile(r'^[ \t\f]*#.*coding[:=][ \t]*([-\w.]+)')
+blank_re = re.compile(r'^[ \t\f]*(?:[#\r\n]|$)')
def _get_normal_name(orig_enc):
"""Imitates get_normal_name in tokenizer.c."""
@@ -309,6 +310,8 @@
encoding = find_cookie(first)
if encoding:
return encoding, [first]
+ if not blank_re.match(first):
+ return default, [first]
second = read_or_stop()
if not second:
diff --git a/Lib/test/test_compile.py b/Lib/test/test_compile.py
index ca396a9..cfc6389 100644
--- a/Lib/test/test_compile.py
+++ b/Lib/test/test_compile.py
@@ -412,9 +412,24 @@
l = lambda: "foo"
self.assertIsNone(l.__doc__)
- def test_unicode_encoding(self):
+ @test_support.requires_unicode
+ def test_encoding(self):
+ code = b'# -*- coding: badencoding -*-\npass\n'
+ self.assertRaises(SyntaxError, compile, code, 'tmp', 'exec')
code = u"# -*- coding: utf-8 -*-\npass\n"
self.assertRaises(SyntaxError, compile, code, "tmp", "exec")
+ code = 'u"\xc2\xa4"\n'
+ self.assertEqual(eval(code), u'\xc2\xa4')
+ code = u'u"\xc2\xa4"\n'
+ self.assertEqual(eval(code), u'\xc2\xa4')
+ code = '# -*- coding: latin1 -*-\nu"\xc2\xa4"\n'
+ self.assertEqual(eval(code), u'\xc2\xa4')
+ code = '# -*- coding: utf-8 -*-\nu"\xc2\xa4"\n'
+ self.assertEqual(eval(code), u'\xa4')
+ code = '# -*- coding: iso8859-15 -*-\nu"\xc2\xa4"\n'
+ self.assertEqual(eval(code), test_support.u(r'\xc2\u20ac'))
+ code = 'u"""\\\n# -*- coding: utf-8 -*-\n\xc2\xa4"""\n'
+ self.assertEqual(eval(code), u'# -*- coding: utf-8 -*-\n\xc2\xa4')
def test_subscripts(self):
# SF bug 1448804
diff --git a/Misc/NEWS b/Misc/NEWS
index e50de99..91e1db4 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -10,6 +10,9 @@
Core and Builtins
-----------------
+- Issue #22221: Now the source encoding declaration on the second line isn't
+ effective if the first line contains anything except a comment.
+
- Issue #22023: Fix ``%S``, ``%R`` and ``%V`` formats of
:c:func:`PyUnicode_FromFormat`.
@@ -124,6 +127,9 @@
IDLE
----
+- Issue #22221: IDLE now ignores the source encoding declaration on the second
+ line if the first line contains anything except a comment.
+
- Issue #17390: Adjust Editor window title; remove 'Python',
move version to end.
@@ -140,6 +146,10 @@
Tools/Demos
-----------
+- Issue #22221: 2to3 and the findnocoding.py script now ignore the source
+ encoding declaration on the second line if the first line contains anything
+ except a comment.
+
- Issue #22201: Command-line interface of the zipfile module now correctly
extracts ZIP files with directory entries. Patch by Ryan Wilson.
diff --git a/Parser/tokenizer.c b/Parser/tokenizer.c
index 3e4af53..d0e4a0e 100644
--- a/Parser/tokenizer.c
+++ b/Parser/tokenizer.c
@@ -259,11 +259,25 @@
char * cs;
int r = 1;
- if (tok->cont_line)
+ if (tok->cont_line) {
/* It's a continuation line, so it can't be a coding spec. */
+ tok->read_coding_spec = 1;
return 1;
+ }
cs = get_coding_spec(line, size);
- if (cs != NULL) {
+ if (!cs) {
+ Py_ssize_t i;
+ for (i = 0; i < size; i++) {
+ if (line[i] == '#' || line[i] == '\n' || line[i] == '\r')
+ break;
+ if (line[i] != ' ' && line[i] != '\t' && line[i] != '\014') {
+ /* Stop checking coding spec after a line containing
+ * anything except a comment. */
+ tok->read_coding_spec = 1;
+ break;
+ }
+ }
+ } else {
tok->read_coding_spec = 1;
if (tok->encoding == NULL) {
assert(tok->decoding_state == 1); /* raw */
@@ -688,7 +702,7 @@
if (newl[0]) {
if (!check_coding_spec(str, newl[0] - str, tok, buf_setreadl))
return error_ret(tok);
- if (tok->enc == NULL && newl[1]) {
+ if (tok->enc == NULL && !tok->read_coding_spec && newl[1]) {
if (!check_coding_spec(newl[0]+1, newl[1] - newl[0],
tok, buf_setreadl))
return error_ret(tok);
diff --git a/Tools/scripts/findnocoding.py b/Tools/scripts/findnocoding.py
index 5d93290..70b1a66 100755
--- a/Tools/scripts/findnocoding.py
+++ b/Tools/scripts/findnocoding.py
@@ -33,6 +33,7 @@
decl_re = re.compile(r'^[ \t\f]*#.*coding[:=][ \t]*([-\w.]+)')
+blank_re = re.compile(r'^[ \t\f]*(?:[#\r\n]|$)')
def get_declaration(line):
match = decl_re.match(line)
@@ -57,7 +58,8 @@
line1 = infile.readline()
line2 = infile.readline()
- if get_declaration(line1) or get_declaration(line2):
+ if (get_declaration(line1) or
+ blank_re.match(line1) and get_declaration(line2)):
# the file does have an encoding declaration, so trust it
infile.close()
return False