1
:mod:`io` --- Core tools for working with streams
2
=================================================
5
:synopsis: Core tools for working with streams.
6
.. moduleauthor:: Guido van Rossum <guido@python.org>
7
.. moduleauthor:: Mike Verdone <mike.verdone@gmail.com>
8
.. moduleauthor:: Mark Russell <mark.russell@zen.co.uk>
9
.. moduleauthor:: Antoine Pitrou <solipsis@pitrou.net>
10
.. moduleauthor:: Amaury Forgeot d'Arc <amauryfa@gmail.com>
11
.. sectionauthor:: Benjamin Peterson <benjamin@python.org>
13
The :mod:`io` module provides the Python interfaces to stream handling. The
14
builtin :func:`open` function is defined in this module.
16
At the top of the I/O hierarchy is the abstract base class :class:`IOBase`. It
17
defines the basic interface to a stream. Note, however, that there is no
18
separation between reading and writing to streams; implementations are allowed
19
to throw an :exc:`IOError` if they do not support a given operation.
21
Extending :class:`IOBase` is :class:`RawIOBase` which deals simply with the
22
reading and writing of raw bytes to a stream. :class:`FileIO` subclasses
23
:class:`RawIOBase` to provide an interface to files in the machine's
26
:class:`BufferedIOBase` deals with buffering on a raw byte stream
27
(:class:`RawIOBase`). Its subclasses, :class:`BufferedWriter`,
28
:class:`BufferedReader`, and :class:`BufferedRWPair` buffer streams that are
29
readable, writable, and both readable and writable.
30
:class:`BufferedRandom` provides a buffered interface to random access
31
streams. :class:`BytesIO` is a simple stream of in-memory bytes.
33
Another :class:`IOBase` subclass, :class:`TextIOBase`, deals with
34
streams whose bytes represent text, and handles encoding and decoding
35
from and to strings. :class:`TextIOWrapper`, which extends it, is a
36
buffered text interface to a buffered raw stream
37
(:class:`BufferedIOBase`). Finally, :class:`StringIO` is an in-memory
40
Argument names are not part of the specification, and only the arguments of
41
:func:`open` are intended to be used as keyword arguments.
47
.. data:: DEFAULT_BUFFER_SIZE
49
An int containing the default buffer size used by the module's buffered I/O
50
classes. :func:`open` uses the file's blksize (as obtained by
51
:func:`os.stat`) if possible.
53
.. function:: open(file[, mode[, buffering[, encoding[, errors[, newline[, closefd=True]]]]]])
55
Open *file* and return a stream. If the file cannot be opened, an
56
:exc:`IOError` is raised.
58
*file* is either a string giving the name (and the path if the file isn't in
59
the current working directory) of the file to be opened or a file
60
descriptor of the file to be opened. (If a file descriptor is given,
61
for example, from :func:`os.fdopen`, it is closed when the returned
62
I/O object is closed, unless *closefd* is set to ``False``.)
64
*mode* is an optional string that specifies the mode in which the file is
65
opened. It defaults to ``'r'`` which means open for reading in text mode.
66
Other common values are ``'w'`` for writing (truncating the file if it
67
already exists), and ``'a'`` for appending (which on *some* Unix systems,
68
means that *all* writes append to the end of the file regardless of the
69
current seek position). In text mode, if *encoding* is not specified the
70
encoding used is platform dependent. (For reading and writing raw bytes use
71
binary mode and leave *encoding* unspecified.) The available modes are:
73
========= ===============================================================
75
--------- ---------------------------------------------------------------
76
``'r'`` open for reading (default)
77
``'w'`` open for writing, truncating the file first
78
``'a'`` open for writing, appending to the end of the file if it exists
80
``'t'`` text mode (default)
81
``'+'`` open a disk file for updating (reading and writing)
82
``'U'`` universal newline mode (for backwards compatibility; should
83
not be used in new code)
84
========= ===============================================================
86
The default mode is ``'rt'`` (open for reading text). For binary random
87
access, the mode ``'w+b'`` opens and truncates the file to 0 bytes, while
88
``'r+b'`` opens the file without truncation.
90
Python distinguishes between files opened in binary and text modes, even when
91
the underlying operating system doesn't. Files opened in binary mode
92
(including ``'b'`` in the *mode* argument) return contents as ``bytes``
93
objects without any decoding. In text mode (the default, or when ``'t'`` is
94
included in the *mode* argument), the contents of the file are returned as
95
strings, the bytes having been first decoded using a platform-dependent
96
encoding or using the specified *encoding* if given.
98
*buffering* is an optional integer used to set the buffering policy. By
99
default full buffering is on. Pass 0 to switch buffering off (only allowed
100
in binary mode), 1 to set line buffering, and an integer > 1 for full
103
*encoding* is the name of the encoding used to decode or encode the file.
104
This should only be used in text mode. The default encoding is platform
105
dependent, but any encoding supported by Python can be used. See the
106
:mod:`codecs` module for the list of supported encodings.
108
*errors* is an optional string that specifies how encoding and decoding
109
errors are to be handled. Pass ``'strict'`` to raise a :exc:`ValueError`
110
exception if there is an encoding error (the default of ``None`` has the same
111
effect), or pass ``'ignore'`` to ignore errors. (Note that ignoring encoding
112
errors can lead to data loss.) ``'replace'`` causes a replacement marker
113
(such as ``'?'``) to be inserted where there is malformed data. When
114
writing, ``'xmlcharrefreplace'`` (replace with the appropriate XML character
115
reference) or ``'backslashreplace'`` (replace with backslashed escape
116
sequences) can be used. Any other error handling name that has been
117
registered with :func:`codecs.register_error` is also valid.
119
*newline* controls how universal newlines works (it only applies to text
120
mode). It can be ``None``, ``''``, ``'\n'``, ``'\r'``, and ``'\r\n'``. It
123
* On input, if *newline* is ``None``, universal newlines mode is enabled.
124
Lines in the input can end in ``'\n'``, ``'\r'``, or ``'\r\n'``, and these
125
are translated into ``'\n'`` before being returned to the caller. If it is
126
``''``, universal newline mode is enabled, but line endings are returned to
127
the caller untranslated. If it has any of the other legal values, input
128
lines are only terminated by the given string, and the line ending is
129
returned to the caller untranslated.
131
* On output, if *newline* is ``None``, any ``'\n'`` characters written are
132
translated to the system default line separator, :data:`os.linesep`. If
133
*newline* is ``''``, no translation takes place. If *newline* is any of
134
the other legal values, any ``'\n'`` characters written are translated to
137
If *closefd* is ``False`` and a file descriptor rather than a
138
filename was given, the underlying file descriptor will be kept open
139
when the file is closed. If a filename is given *closefd* has no
140
effect but must be ``True`` (the default).
142
The type of file object returned by the :func:`open` function depends
143
on the mode. When :func:`open` is used to open a file in a text mode
144
(``'w'``, ``'r'``, ``'wt'``, ``'rt'``, etc.), it returns a
145
:class:`TextIOWrapper`. When used to open a file in a binary mode,
146
the returned class varies: in read binary mode, it returns a
147
:class:`BufferedReader`; in write binary and append binary modes, it
148
returns a :class:`BufferedWriter`, and in read/write mode, it returns
149
a :class:`BufferedRandom`.
151
It is also possible to use a string or bytearray as a file for both reading
152
and writing. For strings :class:`StringIO` can be used like a file opened in
153
a text mode, and for bytearrays a :class:`BytesIO` can be used like a
154
file opened in a binary mode.
157
.. exception:: BlockingIOError
159
Error raised when blocking would occur on a non-blocking stream. It inherits
162
In addition to those of :exc:`IOError`, :exc:`BlockingIOError` has one
165
.. attribute:: characters_written
167
An integer containing the number of characters written to the stream
171
.. exception:: UnsupportedOperation
173
An exception inheriting :exc:`IOError` and :exc:`ValueError` that is raised
174
when an unsupported operation is called on a stream.
182
The abstract base class for all I/O classes, acting on streams of bytes.
183
There is no public constructor.
185
This class provides empty abstract implementations for many methods
186
that derived classes can override selectively; the default
187
implementations represent a file that cannot be read, written or
190
Even though :class:`IOBase` does not declare :meth:`read`, :meth:`readinto`,
191
or :meth:`write` because their signatures will vary, implementations and
192
clients should consider those methods part of the interface. Also,
193
implementations may raise a :exc:`IOError` when operations they do not
196
The basic type used for binary data read from or written to a file is
197
:class:`bytes`. :class:`bytearray`\s are accepted too, and in some cases
198
(such as :class:`readinto`) required. Text I/O classes work with
201
Note that calling any method (even inquiries) on a closed stream is
202
undefined. Implementations may raise :exc:`IOError` in this case.
204
IOBase (and its subclasses) support the iterator protocol, meaning that an
205
:class:`IOBase` object can be iterated over yielding the lines in a stream.
207
IOBase is also a context manager and therefore supports the
208
:keyword:`with` statement. In this example, *file* is closed after the
209
:keyword:`with` statement's suite is finished---even if an exception occurs::
211
with open('spam.txt', 'w') as file:
212
file.write('Spam and eggs!')
214
:class:`IOBase` provides these data attributes and methods:
218
Flush and close this stream. This method has no effect if the file is
219
already closed. Once the file is closed, any operation on the file
220
(e.g. reading or writing) will raise an :exc:`IOError`. The internal
221
file descriptor isn't closed if *closefd* was False.
223
.. attribute:: closed
225
True if the stream is closed.
229
Return the underlying file descriptor (an integer) of the stream if it
230
exists. An :exc:`IOError` is raised if the IO object does not use a file
235
Flush the write buffers of the stream if applicable. This does nothing
236
for read-only and non-blocking streams.
240
Return ``True`` if the stream is interactive (i.e., connected to
241
a terminal/tty device).
243
.. method:: readable()
245
Return ``True`` if the stream can be read from. If False, :meth:`read`
246
will raise :exc:`IOError`.
248
.. method:: readline([limit])
250
Read and return one line from the stream. If *limit* is specified, at
251
most *limit* bytes will be read.
253
The line terminator is always ``b'\n'`` for binary files; for text files,
254
the *newlines* argument to :func:`open` can be used to select the line
255
terminator(s) recognized.
257
.. method:: readlines([hint])
259
Read and return a list of lines from the stream. *hint* can be specified
260
to control the number of lines read: no more lines will be read if the
261
total size (in bytes/characters) of all lines so far exceeds *hint*.
263
.. method:: seek(offset[, whence])
265
Change the stream position to the given byte *offset*. *offset* is
266
interpreted relative to the position indicated by *whence*. Values for
269
* ``0`` -- start of the stream (the default); *offset* should be zero or positive
270
* ``1`` -- current stream position; *offset* may be negative
271
* ``2`` -- end of the stream; *offset* is usually negative
273
Return the new absolute position.
275
.. method:: seekable()
277
Return ``True`` if the stream supports random access. If ``False``,
278
:meth:`seek`, :meth:`tell` and :meth:`truncate` will raise :exc:`IOError`.
282
Return the current stream position.
284
.. method:: truncate([size])
286
Truncate the file to at most *size* bytes. *size* defaults to the current
287
file position, as returned by :meth:`tell`.
289
.. method:: writable()
291
Return ``True`` if the stream supports writing. If ``False``,
292
:meth:`write` and :meth:`truncate` will raise :exc:`IOError`.
294
.. method:: writelines(lines)
296
Write a list of lines to the stream. Line separators are not added, so it
297
is usual for each of the lines provided to have a line separator at the
303
Base class for raw binary I/O. It inherits :class:`IOBase`. There is no
306
In addition to the attributes and methods from :class:`IOBase`,
307
RawIOBase provides the following methods:
309
.. method:: read([n])
311
Read and return all the bytes from the stream until EOF, or if *n* is
312
specified, up to *n* bytes. Only one system call is ever made. An empty
313
bytes object is returned on EOF; ``None`` is returned if the object is set
314
not to block and has no data to read.
316
.. method:: readall()
318
Read and return all the bytes from the stream until EOF, using multiple
319
calls to the stream if necessary.
321
.. method:: readinto(b)
323
Read up to len(b) bytes into bytearray *b* and return the number of bytes
328
Write the given bytes or bytearray object, *b*, to the underlying raw
329
stream and return the number of bytes written (This is never less than
330
``len(b)``, since if the write fails, an :exc:`IOError` will be raised).
333
.. class:: BufferedIOBase
335
Base class for streams that support buffering. It inherits :class:`IOBase`.
336
There is no public constructor.
338
The main difference with :class:`RawIOBase` is that the :meth:`read` method
339
supports omitting the *size* argument, and does not have a default
340
implementation that defers to :meth:`readinto`.
342
In addition, :meth:`read`, :meth:`readinto`, and :meth:`write` may raise
343
:exc:`BlockingIOError` if the underlying raw stream is in non-blocking mode
344
and not ready; unlike their raw counterparts, they will never return
347
A typical implementation should not inherit from a :class:`RawIOBase`
348
implementation, but wrap one like :class:`BufferedWriter` and
349
:class:`BufferedReader`.
351
:class:`BufferedIOBase` provides or overrides these methods in addition to
352
those from :class:`IOBase`:
354
.. method:: read([n])
356
Read and return up to *n* bytes. If the argument is omitted, ``None``, or
357
negative, data is read and returned until EOF is reached. An empty bytes
358
object is returned if the stream is already at EOF.
360
If the argument is positive, and the underlying raw stream is not
361
interactive, multiple raw reads may be issued to satisfy the byte count
362
(unless EOF is reached first). But for interactive raw streams, at most
363
one raw read will be issued, and a short result does not imply that EOF is
366
A :exc:`BlockingIOError` is raised if the underlying raw stream has no
369
.. method:: read1([n])
371
Read and return up to *n* bytes, with at most one call to the underlying
372
raw stream's :meth:`~RawIOBase.read` method.
374
.. method:: readinto(b)
376
Read up to len(b) bytes into bytearray *b* and return the number of bytes
379
Like :meth:`read`, multiple reads may be issued to the underlying raw
380
stream, unless the latter is 'interactive.'
382
A :exc:`BlockingIOError` is raised if the underlying raw stream has no
387
Write the given bytes or bytearray object, *b*, to the underlying raw
388
stream and return the number of bytes written (never less than ``len(b)``,
389
since if the write fails an :exc:`IOError` will be raised).
391
A :exc:`BlockingIOError` is raised if the buffer is full, and the
392
underlying raw stream cannot accept more data at the moment.
398
.. class:: FileIO(name[, mode])
400
:class:`FileIO` represents a file containing bytes data. It implements
401
the :class:`RawIOBase` interface (and therefore the :class:`IOBase`
404
The *mode* can be ``'r'``, ``'w'`` or ``'a'`` for reading (default), writing,
405
or appending. The file will be created if it doesn't exist when opened for
406
writing or appending; it will be truncated when opened for writing. Add a
407
``'+'`` to the mode to allow simultaneous reading and writing.
409
In addition to the attributes and methods from :class:`IOBase` and
410
:class:`RawIOBase`, :class:`FileIO` provides the following data
411
attributes and methods:
415
The mode as given in the constructor.
419
The file name. This is the file descriptor of the file when no name is
420
given in the constructor.
422
.. method:: read([n])
424
Read and return at most *n* bytes. Only one system call is made, so it is
425
possible that less data than was requested is returned. Use :func:`len`
426
on the returned bytes object to see how many bytes were actually returned.
427
(In non-blocking mode, ``None`` is returned when no data is available.)
429
.. method:: readall()
431
Read and return the entire file's contents in a single bytes object. As
432
much as immediately available is returned in non-blocking mode. If the
433
EOF has been reached, ``b''`` is returned.
437
Write the bytes or bytearray object, *b*, to the file, and return
438
the number actually written. Only one system call is made, so it
439
is possible that only some of the data is written.
445
.. class:: BytesIO([initial_bytes])
447
A stream implementation using an in-memory bytes buffer. It inherits
448
:class:`BufferedIOBase`.
450
The argument *initial_bytes* is an optional initial bytearray.
452
:class:`BytesIO` provides or overrides these methods in addition to those
453
from :class:`BufferedIOBase` and :class:`IOBase`:
455
.. method:: getvalue()
457
Return ``bytes`` containing the entire contents of the buffer.
461
In :class:`BytesIO`, this is the same as :meth:`read`.
463
.. method:: truncate([size])
465
Truncate the buffer to at most *size* bytes. *size* defaults to the
466
current stream position, as returned by :meth:`tell`.
469
.. class:: BufferedReader(raw[, buffer_size])
471
A buffer for a readable, sequential :class:`RawIOBase` object. It inherits
472
:class:`BufferedIOBase`.
474
The constructor creates a :class:`BufferedReader` for the given readable
475
*raw* stream and *buffer_size*. If *buffer_size* is omitted,
476
:data:`DEFAULT_BUFFER_SIZE` is used.
478
:class:`BufferedReader` provides or overrides these methods in addition to
479
those from :class:`BufferedIOBase` and :class:`IOBase`:
481
.. method:: peek([n])
483
Return 1 (or *n* if specified) bytes from a buffer without advancing the
484
position. Only a single read on the raw stream is done to satisfy the
485
call. The number of bytes returned may be less than requested since at
486
most all the buffer's bytes from the current position to the end are
489
.. method:: read([n])
491
Read and return *n* bytes, or if *n* is not given or negative, until EOF
492
or if the read call would block in non-blocking mode.
496
Read and return up to *n* bytes with only one call on the raw stream. If
497
at least one byte is buffered, only buffered bytes are returned.
498
Otherwise, one raw stream read call is made.
501
.. class:: BufferedWriter(raw[, buffer_size[, max_buffer_size]])
503
A buffer for a writeable sequential RawIO object. It inherits
504
:class:`BufferedIOBase`.
506
The constructor creates a :class:`BufferedWriter` for the given writeable
507
*raw* stream. If the *buffer_size* is not given, it defaults to
508
:data:`DEFAULT_BUFFER_SIZE`.
510
*max_buffer_size* is unused and deprecated.
512
:class:`BufferedWriter` provides or overrides these methods in addition to
513
those from :class:`BufferedIOBase` and :class:`IOBase`:
517
Force bytes held in the buffer into the raw stream. A
518
:exc:`BlockingIOError` should be raised if the raw stream blocks.
522
Write the bytes or bytearray object, *b*, onto the raw stream and return
523
the number of bytes written. A :exc:`BlockingIOError` is raised when the
527
.. class:: BufferedRWPair(reader, writer[, buffer_size[, max_buffer_size]])
529
A combined buffered writer and reader object for a raw stream that can be
530
written to and read from. It has and supports both :meth:`read`, :meth:`write`,
531
and their variants. This is useful for sockets and two-way pipes.
532
It inherits :class:`BufferedIOBase`.
534
*reader* and *writer* are :class:`RawIOBase` objects that are readable and
535
writeable respectively. If the *buffer_size* is omitted it defaults to
536
:data:`DEFAULT_BUFFER_SIZE`.
538
*max_buffer_size* is unused and deprecated.
540
:class:`BufferedRWPair` implements all of :class:`BufferedIOBase`\'s methods.
543
.. class:: BufferedRandom(raw[, buffer_size[, max_buffer_size]])
545
A buffered interface to random access streams. It inherits
546
:class:`BufferedReader` and :class:`BufferedWriter`.
548
The constructor creates a reader and writer for a seekable raw stream, given
549
in the first argument. If the *buffer_size* is omitted it defaults to
550
:data:`DEFAULT_BUFFER_SIZE`.
552
*max_buffer_size* is unused and deprecated.
554
:class:`BufferedRandom` is capable of anything :class:`BufferedReader` or
555
:class:`BufferedWriter` can do.
561
.. class:: TextIOBase
563
Base class for text streams. This class provides a character and line based
564
interface to stream I/O. There is no :meth:`readinto` method because
565
Python's character strings are immutable. It inherits :class:`IOBase`.
566
There is no public constructor.
568
:class:`TextIOBase` provides or overrides these data attributes and
569
methods in addition to those from :class:`IOBase`:
571
.. attribute:: encoding
573
The name of the encoding used to decode the stream's bytes into
574
strings, and to encode strings into bytes.
576
.. attribute:: newlines
578
A string, a tuple of strings, or ``None``, indicating the newlines
583
Read and return at most *n* characters from the stream as a single
584
:class:`str`. If *n* is negative or ``None``, reads to EOF.
586
.. method:: readline()
588
Read until newline or EOF and return a single ``str``. If the stream is
589
already at EOF, an empty string is returned.
593
Write the string *s* to the stream and return the number of characters
597
.. class:: TextIOWrapper(buffer[, encoding[, errors[, newline[, line_buffering]]]])
599
A buffered text stream over a :class:`BufferedIOBase` raw stream, *buffer*.
600
It inherits :class:`TextIOBase`.
602
*encoding* gives the name of the encoding that the stream will be decoded or
603
encoded with. It defaults to :func:`locale.getpreferredencoding`.
605
*errors* is an optional string that specifies how encoding and decoding
606
errors are to be handled. Pass ``'strict'`` to raise a :exc:`ValueError`
607
exception if there is an encoding error (the default of ``None`` has the same
608
effect), or pass ``'ignore'`` to ignore errors. (Note that ignoring encoding
609
errors can lead to data loss.) ``'replace'`` causes a replacement marker
610
(such as ``'?'``) to be inserted where there is malformed data. When
611
writing, ``'xmlcharrefreplace'`` (replace with the appropriate XML character
612
reference) or ``'backslashreplace'`` (replace with backslashed escape
613
sequences) can be used. Any other error handling name that has been
614
registered with :func:`codecs.register_error` is also valid.
616
*newline* can be ``None``, ``''``, ``'\n'``, ``'\r'``, or ``'\r\n'``. It
617
controls the handling of line endings. If it is ``None``, universal newlines
618
is enabled. With this enabled, on input, the lines endings ``'\n'``,
619
``'\r'``, or ``'\r\n'`` are translated to ``'\n'`` before being returned to
620
the caller. Conversely, on output, ``'\n'`` is translated to the system
621
default line separator, :data:`os.linesep`. If *newline* is any other of its
622
legal values, that newline becomes the newline when the file is read and it
623
is returned untranslated. On output, ``'\n'`` is converted to the *newline*.
625
If *line_buffering* is ``True``, :meth:`flush` is implied when a call to
626
write contains a newline character.
628
:class:`TextIOWrapper` provides these data attributes in addition to those of
629
:class:`TextIOBase` and its parents:
631
.. attribute:: errors
633
The encoding and decoding error setting.
635
.. attribute:: line_buffering
637
Whether line buffering is enabled.
640
.. class:: StringIO([initial_value[, newline]])
642
An in-memory stream for text. It inherits :class:`TextIOWrapper`.
644
The initial value of the buffer (an empty string by default) can be set by
645
providing *initial_value*. The *newline* argument works like that of
646
:class:`TextIOWrapper`. The default is to do no newline translation.
648
:class:`StringIO` provides this method in addition to those from
649
:class:`TextIOWrapper` and its parents:
651
.. method:: getvalue()
653
Return a ``str`` containing the entire contents of the buffer at any
654
time before the :class:`StringIO` object's :meth:`close` method is
661
output = io.StringIO()
662
output.write('First line.\n')
663
print('Second line.', file=output)
665
# Retrieve file contents -- this will be
666
# 'First line.\nSecond line.\n'
667
contents = output.getvalue()
669
# Close object and discard memory buffer --
670
# .getvalue() will now raise an exception.
673
.. class:: IncrementalNewlineDecoder
675
A helper codec that decodes newlines for universal newlines mode. It
676
inherits :class:`codecs.IncrementalDecoder`.
added
removed
Doc/library/io.rst
