python-peps/peps/pep-0383.rst

PEP: 383
Title: Non-decodable Bytes in System Character Interfaces
Author: Martin von Löwis <martin@v.loewis.de>
Status: Final
Type: Standards Track
Content-Type: text/x-rst
Created: 22-Apr-2009
Python-Version: 3.1
Post-History:

Abstract
========

File names, environment variables, and command line arguments are
defined as being character data in POSIX; the C APIs however allow
passing arbitrary bytes - whether these conform to a certain encoding
or not. This PEP proposes a means of dealing with such irregularities
by embedding the bytes in character strings in such a way that allows
recreation of the original byte string.

Rationale
=========

The C char type is a data type that is commonly used to represent both
character data and bytes. Certain POSIX interfaces are specified and
widely understood as operating on character data, however, the system
call interfaces make no assumption on the encoding of these data, and
pass them on as-is. With Python 3, character strings use a
Unicode-based internal representation, making it difficult to ignore
the encoding of byte strings in the same way that the C interfaces can
ignore the encoding.

On the other hand, Microsoft Windows NT has corrected the original
design limitation of Unix, and made it explicit in its system
interfaces that these data (file names, environment variables, command
line arguments) are indeed character data, by providing a
Unicode-based API (keeping a C-char-based one for backwards
compatibility).

For Python 3, one proposed solution is to provide two sets of APIs: a
byte-oriented one, and a character-oriented one, where the
character-oriented one would be limited to not being able to represent
all data accurately. Unfortunately, for Windows, the situation would
be exactly the opposite: the byte-oriented interface cannot represent
all data; only the character-oriented API can. As a consequence,
libraries and applications that want to support all user data in a
cross-platform manner have to accept mish-mash of bytes and characters
exactly in the way that caused endless troubles for Python 2.x.

With this PEP, a uniform treatment of these data as characters becomes
possible. The uniformity is achieved by using specific encoding
algorithms, meaning that the data can be converted back to bytes on
POSIX systems only if the same encoding is used.

Being able to treat such strings uniformly will allow application
writers to abstract from details specific to the operating system, and
reduces the risk of one API failing when the other API would have
worked.

Specification
=============

On Windows, Python uses the wide character APIs to access
character-oriented APIs, allowing direct conversion of the
environmental data to Python str objects (:pep:`277`).

On POSIX systems, Python currently applies the locale's encoding to
convert the byte data to Unicode, failing for characters that cannot
be decoded. With this PEP, non-decodable bytes >= 128 will be
represented as lone surrogate codes U+DC80..U+DCFF. Bytes below
128 will produce exceptions; see the discussion below.

To convert non-decodable bytes, a new error handler (:pep:`293`)
"surrogateescape" is introduced, which produces these surrogates. On
encoding, the error handler converts the surrogate back to the
corresponding byte. This error handler will be used in any API that
receives or produces file names, command line arguments, or
environment variables.

The error handler interface is extended to allow the encode error
handler to return byte strings immediately, in addition to returning
Unicode strings which then get encoded again (also see the discussion
below).

Byte-oriented interfaces that already exist in Python 3.0 are not
affected by this specification. They are neither enhanced nor
deprecated.

External libraries that operate on file names (such as GUI file
choosers) should also encode them according to the PEP.

Discussion
==========

This surrogateescape encoding is based on Markus Kuhn's idea that
he called UTF-8b [3]_.

While providing a uniform API to non-decodable bytes, this interface
has the limitation that chosen representation only "works" if the data
get converted back to bytes with the surrogateescape error handler
also. Encoding the data with the locale's encoding and the (default)
strict error handler will raise an exception, encoding them with UTF-8
will produce nonsensical data.

Data obtained from other sources may conflict with data produced
by this PEP. Dealing with such conflicts is out of scope of the PEP.

This PEP allows the possibility of "smuggling" bytes in character
strings. This would be a security risk if the bytes are
security-critical when interpreted as characters on a target system,
such as path name separators. For this reason, the PEP rejects
smuggling bytes below 128. If the target system uses EBCDIC, such
smuggled bytes may still be a security risk, allowing smuggling of
e.g. square brackets or the backslash. Python currently does not
support EBCDIC, so this should not be a problem in practice. Anybody
porting Python to an EBCDIC system might want to adjust the error
handlers, or come up with other approaches to address the security
risks.

Encodings that are not compatible with ASCII are not supported by
this specification; bytes in the ASCII range that fail to decode
will cause an exception. It is widely agreed that such encodings
should not be used as locale charsets.

For most applications, we assume that they eventually pass data
received from a system interface back into the same system
interfaces. For example, an application invoking os.listdir() will
likely pass the result strings back into APIs like os.stat() or
open(), which then encodes them back into their original byte
representation. Applications that need to process the original byte
strings can obtain them by encoding the character strings with the
file system encoding, passing "surrogateescape" as the error handler
name. For example, a function that works like os.listdir, except for
accepting and returning bytes, would be written as::

  def listdir_b(dirname):
      fse = sys.getfilesystemencoding()
      dirname = dirname.decode(fse, "surrogateescape")
      for fn in os.listdir(dirname):
          # fn is now a str object
          yield fn.encode(fse, "surrogateescape")

The extension to the encode error handler interface proposed by this
PEP is necessary to implement the 'surrogateescape' error handler,
because there are required byte sequences which cannot be generated
from replacement Unicode.  However, the encode error handler interface
presently requires replacement Unicode to be provided in lieu of the
non-encodable Unicode from the source string.  Then it promptly
encodes that replacement Unicode.  In some error handlers, such as the
'surrogateescape' proposed here, it is also simpler and more efficient
for the error handler to provide a pre-encoded replacement byte
string, rather than forcing it to calculating Unicode from which the
encoder would create the desired bytes.

A few alternative approaches have been proposed:

* create a new string subclass that supports embedded bytes
* use different escape schemes, such as escaping with a NUL
  character, or mapping to infrequent characters.

Of these proposals, the approach of escaping each byte XX
with the sequence U+0000 U+00XX has the disadvantage that
encoding to UTF-8 will introduce a NUL byte in the UTF-8
sequence. As a consequence, C libraries may interpret this
as a string termination, even though the string continues.
In particular, the gtk libraries will truncate text in this
case; other libraries may show similar problems.

References
==========

.. [3] UTF-8b
   https://web.archive.org/web/20090830064219/http://mail.nl.linux.org/linux-utf8/2000-07/msg00040.html

Copyright
=========

This document has been placed in the public domain.