Backport PR #6911: don't use text mode in mkstemp

causes double-encoding of newlines, preventing newline arg from having desired effect,
and resulting in CRCRLF line endings on Windows.

closes  #6599

IPython/utils/io.py

@@ -1,345 +1,345 b''
 # encoding: utf-8
 """
 IO related utilities.
 """
 
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2008-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 from __future__ import print_function
 from __future__ import absolute_import
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 import codecs
 from contextlib import contextmanager
 import io
 import os
 import shutil
 import stat
 import sys
 import tempfile
 from .capture import CapturedIO, capture_output
 from .py3compat import string_types, input, PY3
 
 #-----------------------------------------------------------------------------
 # Code
 #-----------------------------------------------------------------------------
 
 
 class IOStream:
 
     def __init__(self,stream, fallback=None):
         if not hasattr(stream,'write') or not hasattr(stream,'flush'):
             if fallback is not None:
                 stream = fallback
             else:
                 raise ValueError("fallback required, but not specified")
         self.stream = stream
         self._swrite = stream.write
 
         # clone all methods not overridden:
         def clone(meth):
             return not hasattr(self, meth) and not meth.startswith('_')
         for meth in filter(clone, dir(stream)):
             setattr(self, meth, getattr(stream, meth))
 
     def __repr__(self):
         cls = self.__class__
         tpl = '{mod}.{cls}({args})'
         return tpl.format(mod=cls.__module__, cls=cls.__name__, args=self.stream)
 
     def write(self,data):
         try:
             self._swrite(data)
         except:
             try:
                 # print handles some unicode issues which may trip a plain
                 # write() call.  Emulate write() by using an empty end
                 # argument.
                 print(data, end='', file=self.stream)
             except:
                 # if we get here, something is seriously broken.
                 print('ERROR - failed to write data to stream:', self.stream,
                       file=sys.stderr)
 
     def writelines(self, lines):
         if isinstance(lines, string_types):
             lines = [lines]
         for line in lines:
             self.write(line)
 
     # This class used to have a writeln method, but regular files and streams
     # in Python don't have this method. We need to keep this completely
     # compatible so we removed it.
 
     @property
     def closed(self):
         return self.stream.closed
 
     def close(self):
         pass
 
 # setup stdin/stdout/stderr to sys.stdin/sys.stdout/sys.stderr
 devnull = open(os.devnull, 'w') 
 stdin = IOStream(sys.stdin, fallback=devnull)
 stdout = IOStream(sys.stdout, fallback=devnull)
 stderr = IOStream(sys.stderr, fallback=devnull)
 
 class IOTerm:
     """ Term holds the file or file-like objects for handling I/O operations.
 
     These are normally just sys.stdin, sys.stdout and sys.stderr but for
     Windows they can can replaced to allow editing the strings before they are
     displayed."""
 
     # In the future, having IPython channel all its I/O operations through
     # this class will make it easier to embed it into other environments which
     # are not a normal terminal (such as a GUI-based shell)
     def __init__(self, stdin=None, stdout=None, stderr=None):
         mymodule = sys.modules[__name__]
         self.stdin  = IOStream(stdin, mymodule.stdin)
         self.stdout = IOStream(stdout, mymodule.stdout)
         self.stderr = IOStream(stderr, mymodule.stderr)
 
 
 class Tee(object):
     """A class to duplicate an output stream to stdout/err.
 
     This works in a manner very similar to the Unix 'tee' command.
 
     When the object is closed or deleted, it closes the original file given to
     it for duplication.
     """
     # Inspired by:
     # http://mail.python.org/pipermail/python-list/2007-May/442737.html
 
     def __init__(self, file_or_name, mode="w", channel='stdout'):
         """Construct a new Tee object.
 
         Parameters
         ----------
         file_or_name : filename or open filehandle (writable)
           File that will be duplicated
 
         mode : optional, valid mode for open().
           If a filename was give, open with this mode.
 
         channel : str, one of ['stdout', 'stderr']
         """
         if channel not in ['stdout', 'stderr']:
             raise ValueError('Invalid channel spec %s' % channel)
 
         if hasattr(file_or_name, 'write') and hasattr(file_or_name, 'seek'):
             self.file = file_or_name
         else:
             self.file = open(file_or_name, mode)
         self.channel = channel
         self.ostream = getattr(sys, channel)
         setattr(sys, channel, self)
         self._closed = False
 
     def close(self):
         """Close the file and restore the channel."""
         self.flush()
         setattr(sys, self.channel, self.ostream)
         self.file.close()
         self._closed = True
 
     def write(self, data):
         """Write data to both channels."""
         self.file.write(data)
         self.ostream.write(data)
         self.ostream.flush()
 
     def flush(self):
         """Flush both channels."""
         self.file.flush()
         self.ostream.flush()
 
     def __del__(self):
         if not self._closed:
             self.close()
 
 
 def ask_yes_no(prompt, default=None, interrupt=None):
     """Asks a question and returns a boolean (y/n) answer.
 
     If default is given (one of 'y','n'), it is used if the user input is
     empty. If interrupt is given (one of 'y','n'), it is used if the user
     presses Ctrl-C. Otherwise the question is repeated until an answer is
     given.
 
     An EOF is treated as the default answer.  If there is no default, an
     exception is raised to prevent infinite loops.
 
     Valid answers are: y/yes/n/no (match is not case sensitive)."""
 
     answers = {'y':True,'n':False,'yes':True,'no':False}
     ans = None
     while ans not in answers.keys():
         try:
             ans = input(prompt+' ').lower()
             if not ans:  # response was an empty string
                 ans = default
         except KeyboardInterrupt:
             if interrupt:
                 ans = interrupt
         except EOFError:
             if default in answers.keys():
                 ans = default
                 print()
             else:
                 raise
 
     return answers[ans]
 
 
 def temp_pyfile(src, ext='.py'):
     """Make a temporary python file, return filename and filehandle.
 
     Parameters
     ----------
     src : string or list of strings (no need for ending newlines if list)
       Source code to be written to the file.
 
     ext : optional, string
       Extension for the generated file.
 
     Returns
     -------
     (filename, open filehandle)
       It is the caller's responsibility to close the open file and unlink it.
     """
     fname = tempfile.mkstemp(ext)[1]
     f = open(fname,'w')
     f.write(src)
     f.flush()
     return fname, f
 
 def _copy_metadata(src, dst):
     """Copy the set of metadata we want for atomic_writing.
 
     Permission bits and flags. We'd like to copy file ownership as well, but we
     can't do that.
     """
     shutil.copymode(src, dst)
     st = os.stat(src)
     if hasattr(os, 'chflags') and hasattr(st, 'st_flags'):
         os.chflags(dst, st.st_flags)
 
 @contextmanager
 def atomic_writing(path, text=True, encoding='utf-8', **kwargs):
     """Context manager to write to a file only if the entire write is successful.
 
     This works by creating a temporary file in the same directory, and renaming
     it over the old file if the context is exited without an error. If the
     target file is a symlink or a hardlink, this will not be preserved: it will
     be replaced by a new regular file.
 
     On Windows, there is a small chink in the atomicity: the target file is
     deleted before renaming the temporary file over it. This appears to be
     unavoidable.
 
     Parameters
     ----------
     path : str
       The target file to write to.
 
     text : bool, optional
       Whether to open the file in text mode (i.e. to write unicode). Default is
       True.
 
     encoding : str, optional
       The encoding to use for files opened in text mode. Default is UTF-8.
 
     **kwargs
       Passed to :func:`io.open`.
     """
     # realpath doesn't work on Windows: http://bugs.python.org/issue9949
     # Luckily, we only need to resolve the file itself being a symlink, not
     # any of its directories, so this will suffice:
     if os.path.islink(path):
         path = os.path.join(os.path.dirname(path), os.readlink(path))
 
     dirname, basename = os.path.split(path)
-    handle, tmp_path = tempfile.mkstemp(prefix=basename, dir=dirname, text=text)
+    handle, tmp_path = tempfile.mkstemp(prefix=basename, dir=dirname)
     if text:
         fileobj = io.open(handle, 'w', encoding=encoding, **kwargs)
     else:
         fileobj = io.open(handle, 'wb', **kwargs)
 
     try:
         yield fileobj
     except:
         fileobj.close()
         os.remove(tmp_path)
         raise
 
     # Flush to disk
     fileobj.flush()
     os.fsync(fileobj.fileno())
 
     # Written successfully, now rename it
     fileobj.close()
 
     # Copy permission bits, access time, etc.
     try:
         _copy_metadata(path, tmp_path)
     except OSError:
         # e.g. the file didn't already exist. Ignore any failure to copy metadata
         pass
 
     if os.name == 'nt' and os.path.exists(path):
         # Rename over existing file doesn't work on Windows
         os.remove(path)
 
     os.rename(tmp_path, path)
 
 
 def raw_print(*args, **kw):
     """Raw print to sys.__stdout__, otherwise identical interface to print()."""
 
     print(*args, sep=kw.get('sep', ' '), end=kw.get('end', '\n'),
           file=sys.__stdout__)
     sys.__stdout__.flush()
 
 
 def raw_print_err(*args, **kw):
     """Raw print to sys.__stderr__, otherwise identical interface to print()."""
 
     print(*args, sep=kw.get('sep', ' '), end=kw.get('end', '\n'),
           file=sys.__stderr__)
     sys.__stderr__.flush()
 
 
 # Short aliases for quick debugging, do NOT use these in production code.
 rprint = raw_print
 rprinte = raw_print_err
 
 def unicode_std_stream(stream='stdout'):
     u"""Get a wrapper to write unicode to stdout/stderr as UTF-8.
 
     This ignores environment variables and default encodings, to reliably write
     unicode to stdout or stderr.
 
     ::
 
         unicode_std_stream().write(u'ł@e¶ŧ←')
     """
     assert stream in ('stdout', 'stderr')
     stream  = getattr(sys, stream)
     if PY3:
         try:
             stream_b = stream.buffer
         except AttributeError:
             # sys.stdout has been replaced - use it directly
             return stream
     else:
         stream_b = stream
 
     return codecs.getwriter('utf-8')(stream_b)

IPython/utils/tests/test_io.py

@@ -1,178 +1,215 b''
 # encoding: utf-8
 """Tests for io.py"""
 
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2008-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 from __future__ import print_function
 from __future__ import absolute_import
 
 import io as stdlib_io
 import os.path
 import stat
 import sys
 
 from subprocess import Popen, PIPE
 import unittest
 
 import nose.tools as nt
 
 from IPython.testing.decorators import skipif
 from IPython.utils.io import (Tee, capture_output, unicode_std_stream,
                               atomic_writing,
                               )
 from IPython.utils.py3compat import doctest_refactor_print, PY3
 from IPython.utils.tempdir import TemporaryDirectory
 
 if PY3:
     from io import StringIO
 else:
     from StringIO import StringIO
 
 #-----------------------------------------------------------------------------
 # Tests
 #-----------------------------------------------------------------------------
 
 
 def test_tee_simple():
     "Very simple check with stdout only"
     chan = StringIO()
     text = 'Hello'
     tee = Tee(chan, channel='stdout')
     print(text, file=chan)
     nt.assert_equal(chan.getvalue(), text+"\n")
 
 
 class TeeTestCase(unittest.TestCase):
 
     def tchan(self, channel, check='close'):
         trap = StringIO()
         chan = StringIO()
         text = 'Hello'
         
         std_ori = getattr(sys, channel)
         setattr(sys, channel, trap)
 
         tee = Tee(chan, channel=channel)
         print(text, end='', file=chan)
         setattr(sys, channel, std_ori)
         trap_val = trap.getvalue()
         nt.assert_equal(chan.getvalue(), text)
         if check=='close':
             tee.close()
         else:
             del tee
 
     def test(self):
         for chan in ['stdout', 'stderr']:
             for check in ['close', 'del']:
                 self.tchan(chan, check)
 
 def test_io_init():
     """Test that io.stdin/out/err exist at startup"""
     for name in ('stdin', 'stdout', 'stderr'):
         cmd = doctest_refactor_print("from IPython.utils import io;print io.%s.__class__"%name)
         p = Popen([sys.executable, '-c', cmd],
                     stdout=PIPE)
         p.wait()
         classname = p.stdout.read().strip().decode('ascii')
         # __class__ is a reference to the class object in Python 3, so we can't
         # just test for string equality.
         assert 'IPython.utils.io.IOStream' in classname, classname
 
 def test_capture_output():
     """capture_output() context works"""
     
     with capture_output() as io:
         print('hi, stdout')
         print('hi, stderr', file=sys.stderr)
     
     nt.assert_equal(io.stdout, 'hi, stdout\n')
     nt.assert_equal(io.stderr, 'hi, stderr\n')
 
 def test_UnicodeStdStream():
     # Test wrapping a bytes-level stdout
     if PY3:
         stdoutb = stdlib_io.BytesIO()
         stdout = stdlib_io.TextIOWrapper(stdoutb, encoding='ascii')
     else:
         stdout = stdoutb = stdlib_io.BytesIO()
 
     orig_stdout = sys.stdout
     sys.stdout = stdout
     try:
         sample = u"@łe¶ŧ←"
         unicode_std_stream().write(sample)
 
         output = stdoutb.getvalue().decode('utf-8')
         nt.assert_equal(output, sample)
         assert not stdout.closed
     finally:
         sys.stdout = orig_stdout
 
 @skipif(not PY3, "Not applicable on Python 2")
 def test_UnicodeStdStream_nowrap():
     # If we replace stdout with a StringIO, it shouldn't get wrapped.
     orig_stdout = sys.stdout
     sys.stdout = StringIO()
     try:
         nt.assert_is(unicode_std_stream(), sys.stdout)
         assert not sys.stdout.closed
     finally:
         sys.stdout = orig_stdout
 
 def test_atomic_writing():
     class CustomExc(Exception): pass
 
     with TemporaryDirectory() as td:
         f1 = os.path.join(td, 'penguin')
         with stdlib_io.open(f1, 'w') as f:
             f.write(u'Before')
 
         if os.name != 'nt':
             os.chmod(f1, 0o701)
             orig_mode = stat.S_IMODE(os.stat(f1).st_mode)
 
         f2 = os.path.join(td, 'flamingo')
         try:
             os.symlink(f1, f2)
             have_symlink = True
         except (AttributeError, NotImplementedError, OSError):
             # AttributeError: Python doesn't support it
             # NotImplementedError: The system doesn't support it
             # OSError: The user lacks the privilege (Windows)
             have_symlink = False
 
         with nt.assert_raises(CustomExc):
             with atomic_writing(f1) as f:
                 f.write(u'Failing write')
                 raise CustomExc
 
         # Because of the exception, the file should not have been modified
         with stdlib_io.open(f1, 'r') as f:
             nt.assert_equal(f.read(), u'Before')
 
         with atomic_writing(f1) as f:
             f.write(u'Overwritten')
 
         with stdlib_io.open(f1, 'r') as f:
             nt.assert_equal(f.read(), u'Overwritten')
 
         if os.name != 'nt':
             mode = stat.S_IMODE(os.stat(f1).st_mode)
             nt.assert_equal(mode, orig_mode)
 
         if have_symlink:
             # Check that writing over a file preserves a symlink
             with atomic_writing(f2) as f:
                 f.write(u'written from symlink')
 
             with stdlib_io.open(f1, 'r') as f:
-                nt.assert_equal(f.read(), u'written from symlink')
  No newline at end of file
+                nt.assert_equal(f.read(), u'written from symlink')
+
+def test_atomic_writing_newlines():
+    with TemporaryDirectory() as td:
+        path = os.path.join(td, 'testfile')
+
+        lf = u'a\nb\nc\n'
+        plat = lf.replace(u'\n', os.linesep)
+        crlf = lf.replace(u'\n', u'\r\n')
+
+        # test default
+        with stdlib_io.open(path, 'w') as f:
+            f.write(lf)
+        with stdlib_io.open(path, 'r', newline='') as f:
+            read = f.read()
+        nt.assert_equal(read, plat)
+
+        # test newline=LF
+        with stdlib_io.open(path, 'w', newline='\n') as f:
+            f.write(lf)
+        with stdlib_io.open(path, 'r', newline='') as f:
+            read = f.read()
+        nt.assert_equal(read, lf)
+
+        # test newline=CRLF
+        with atomic_writing(path, newline='\r\n') as f:
+            f.write(lf)
+        with stdlib_io.open(path, 'r', newline='') as f:
+            read = f.read()
+        nt.assert_equal(read, crlf)
+
+        # test newline=no convert
+        text = u'crlf\r\ncr\rlf\n'
+        with atomic_writing(path, newline='') as f:
+            f.write(text)
+        with stdlib_io.open(path, 'r', newline='') as f:
+            read = f.read()
+        nt.assert_equal(read, text)