mirror of
https://github.com/Sneed-Group/Poodletooth-iLand
synced 2024-12-27 13:42:28 -06:00
108 lines
3.8 KiB
Python
108 lines
3.8 KiB
Python
|
# -*- coding: utf-8 -*-
|
||
|
#
|
||
|
# Util/py3compat.py : Compatibility code for handling Py3k / Python 2.x
|
||
|
#
|
||
|
# Written in 2010 by Thorsten Behrens
|
||
|
#
|
||
|
# ===================================================================
|
||
|
# The contents of this file are dedicated to the public domain. To
|
||
|
# the extent that dedication to the public domain is not available,
|
||
|
# everyone is granted a worldwide, perpetual, royalty-free,
|
||
|
# non-exclusive license to exercise all rights associated with the
|
||
|
# contents of this file for any purpose whatsoever.
|
||
|
# No rights are reserved.
|
||
|
#
|
||
|
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
||
|
# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
||
|
# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
||
|
# NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
|
||
|
# BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
|
||
|
# ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
|
||
|
# CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
||
|
# SOFTWARE.
|
||
|
# ===================================================================
|
||
|
|
||
|
"""Compatibility code for handling string/bytes changes from Python 2.x to Py3k
|
||
|
|
||
|
In Python 2.x, strings (of type ''str'') contain binary data, including encoded
|
||
|
Unicode text (e.g. UTF-8). The separate type ''unicode'' holds Unicode text.
|
||
|
Unicode literals are specified via the u'...' prefix. Indexing or slicing
|
||
|
either type always produces a string of the same type as the original.
|
||
|
Data read from a file is always of '''str'' type.
|
||
|
|
||
|
In Python 3.x, strings (type ''str'') may only contain Unicode text. The u'...'
|
||
|
prefix and the ''unicode'' type are now redundant. A new type (called
|
||
|
''bytes'') has to be used for binary data (including any particular
|
||
|
''encoding'' of a string). The b'...' prefix allows one to specify a binary
|
||
|
literal. Indexing or slicing a string produces another string. Slicing a byte
|
||
|
string produces another byte string, but the indexing operation produces an
|
||
|
integer. Data read from a file is of '''str'' type if the file was opened in
|
||
|
text mode, or of ''bytes'' type otherwise.
|
||
|
|
||
|
Since PyCrypto aims at supporting both Python 2.x and 3.x, the following helper
|
||
|
functions are used to keep the rest of the library as independent as possible
|
||
|
from the actual Python version.
|
||
|
|
||
|
In general, the code should always deal with binary strings, and use integers
|
||
|
instead of 1-byte character strings.
|
||
|
|
||
|
b(s)
|
||
|
Take a text string literal (with no prefix or with u'...' prefix) and
|
||
|
make a byte string.
|
||
|
bchr(c)
|
||
|
Take an integer and make a 1-character byte string.
|
||
|
bord(c)
|
||
|
Take the result of indexing on a byte string and make an integer.
|
||
|
tobytes(s)
|
||
|
Take a text string, a byte string, or a sequence of character taken from
|
||
|
a byte string and make a byte string.
|
||
|
"""
|
||
|
|
||
|
__revision__ = "$Id$"
|
||
|
|
||
|
import sys
|
||
|
|
||
|
if sys.version_info[0] == 2:
|
||
|
def b(s):
|
||
|
return s
|
||
|
def bchr(s):
|
||
|
return chr(s)
|
||
|
def bstr(s):
|
||
|
return str(s)
|
||
|
def bord(s):
|
||
|
return ord(s)
|
||
|
if sys.version_info[1] == 1:
|
||
|
def tobytes(s):
|
||
|
try:
|
||
|
return s.encode('latin-1')
|
||
|
except:
|
||
|
return ''.join(s)
|
||
|
else:
|
||
|
def tobytes(s):
|
||
|
if isinstance(s, unicode):
|
||
|
return s.encode("latin-1")
|
||
|
else:
|
||
|
return ''.join(s)
|
||
|
else:
|
||
|
def b(s):
|
||
|
return s.encode("latin-1") # utf-8 would cause some side-effects we don't want
|
||
|
def bchr(s):
|
||
|
return bytes([s])
|
||
|
def bstr(s):
|
||
|
if isinstance(s,str):
|
||
|
return bytes(s,"latin-1")
|
||
|
else:
|
||
|
return bytes(s)
|
||
|
def bord(s):
|
||
|
return s
|
||
|
def tobytes(s):
|
||
|
if isinstance(s,bytes):
|
||
|
return s
|
||
|
else:
|
||
|
if isinstance(s,str):
|
||
|
return s.encode("latin-1")
|
||
|
else:
|
||
|
return bytes(s)
|
||
|
|
||
|
# vim:set ts=4 sw=4 sts=4 expandtab:
|