in tokenize.detect_encoding(), return utf-8-sig when a BOM is found
diff --git a/Doc/library/tokenize.rst b/Doc/library/tokenize.rst
index 7017045..ac6ae36 100644
--- a/Doc/library/tokenize.rst
+++ b/Doc/library/tokenize.rst
@@ -95,7 +95,8 @@
It detects the encoding from the presence of a UTF-8 BOM or an encoding
cookie as specified in :pep:`263`. If both a BOM and a cookie are present,
- but disagree, a SyntaxError will be raised.
+ but disagree, a SyntaxError will be raised. Note that if the BOM is found,
+ ``'utf-8-sig'`` will be returned as an encoding.
If no encoding is specified, then the default of ``'utf-8'`` will be returned.
diff --git a/Lib/test/test_tokenize.py b/Lib/test/test_tokenize.py
index 7b91ab2..1bfac40 100644
--- a/Lib/test/test_tokenize.py
+++ b/Lib/test/test_tokenize.py
@@ -726,7 +726,7 @@
b'do_something(else)\n'
)
encoding, consumed_lines = detect_encoding(self.get_readline(lines))
- self.assertEquals(encoding, 'utf-8')
+ self.assertEquals(encoding, 'utf-8-sig')
self.assertEquals(consumed_lines,
[b'# something\n', b'print(something)\n'])
@@ -747,7 +747,7 @@
b'do_something(else)\n'
)
encoding, consumed_lines = detect_encoding(self.get_readline(lines))
- self.assertEquals(encoding, 'utf-8')
+ self.assertEquals(encoding, 'utf-8-sig')
self.assertEquals(consumed_lines, [b'# coding=utf-8\n'])
def test_mismatched_bom_and_cookie_first_line_raises_syntaxerror(self):
@@ -779,7 +779,7 @@
b'do_something(else)\n'
)
encoding, consumed_lines = detect_encoding(self.get_readline(lines))
- self.assertEquals(encoding, 'utf-8')
+ self.assertEquals(encoding, 'utf-8-sig')
self.assertEquals(consumed_lines,
[b'#! something\n', b'f# coding=utf-8\n'])
@@ -833,12 +833,12 @@
readline = self.get_readline((b'\xef\xbb\xbfprint(something)\n',))
encoding, consumed_lines = detect_encoding(readline)
- self.assertEquals(encoding, 'utf-8')
+ self.assertEquals(encoding, 'utf-8-sig')
self.assertEquals(consumed_lines, [b'print(something)\n'])
readline = self.get_readline((b'\xef\xbb\xbf',))
encoding, consumed_lines = detect_encoding(readline)
- self.assertEquals(encoding, 'utf-8')
+ self.assertEquals(encoding, 'utf-8-sig')
self.assertEquals(consumed_lines, [])
readline = self.get_readline((b'# coding: bad\n',))
diff --git a/Lib/tokenize.py b/Lib/tokenize.py
index f82922b..8972137 100644
--- a/Lib/tokenize.py
+++ b/Lib/tokenize.py
@@ -301,14 +301,16 @@
in.
It detects the encoding from the presence of a utf-8 bom or an encoding
- cookie as specified in pep-0263. If both a bom and a cookie are present,
- but disagree, a SyntaxError will be raised. If the encoding cookie is an
- invalid charset, raise a SyntaxError.
+ cookie as specified in pep-0263. If both a bom and a cookie are present, but
+ disagree, a SyntaxError will be raised. If the encoding cookie is an invalid
+ charset, raise a SyntaxError. Note that if a utf-8 bom is found,
+ 'utf-8-sig' is returned.
If no encoding is specified, then the default of 'utf-8' will be returned.
"""
bom_found = False
encoding = None
+ default = 'utf-8'
def read_or_stop():
try:
return readline()
@@ -340,8 +342,9 @@
if first.startswith(BOM_UTF8):
bom_found = True
first = first[3:]
+ default = 'utf-8-sig'
if not first:
- return 'utf-8', []
+ return default, []
encoding = find_cookie(first)
if encoding:
@@ -349,13 +352,13 @@
second = read_or_stop()
if not second:
- return 'utf-8', [first]
+ return default, [first]
encoding = find_cookie(second)
if encoding:
return encoding, [first, second]
- return 'utf-8', [first, second]
+ return default, [first, second]
def tokenize(readline):
@@ -394,6 +397,9 @@
indents = [0]
if encoding is not None:
+ if encoding == "utf-8-sig":
+ # BOM will already have been stripped.
+ encoding = "utf-8"
yield TokenInfo(ENCODING, encoding, (0, 0), (0, 0), '')
while True: # loop over lines in stream
try:
diff --git a/Misc/NEWS b/Misc/NEWS
index 01c37ce..f1b068b 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -283,6 +283,9 @@
Library
-------
+- ``tokenize.detect_encoding`` now returns ``'utf-8-sig'`` when a UTF-8 BOM is
+ detected.
+
- Issue #8024: Update the Unicode database to 5.2.
- Issue #6716/2: Backslash-replace error output in compilall.