Backport fast alternate io.BytesIO implementation.
Merged r62778, r62779, r62802, r62806, r62807, r62808, r62809, r62844,
r62846, r62952, r62956.
diff --git a/Lib/io.py b/Lib/io.py
index 14178d7..0b6907a 100644
--- a/Lib/io.py
+++ b/Lib/io.py
@@ -494,6 +494,7 @@
files, the newlines argument to open can be used to select the line
terminator(s) recognized.
"""
+ self._checkClosed()
if hasattr(self, "peek"):
def nreadahead():
readahead = self.peek(1)
@@ -508,6 +509,8 @@
return 1
if limit is None:
limit = -1
+ if not isinstance(limit, (int, long)):
+ raise TypeError("limit must be an integer")
res = bytearray()
while limit < 0 or len(res) < limit:
b = self.read(nreadahead())
@@ -536,6 +539,10 @@
lines so far exceeds hint.
"""
if hint is None:
+ hint = -1
+ if not isinstance(hint, (int, long)):
+ raise TypeError("hint must be an integer")
+ if hint <= 0:
return list(self)
n = 0
lines = []
@@ -686,7 +693,7 @@
import array
if not isinstance(b, array.array):
raise err
- b[:n] = array.array('b', data)
+ b[:n] = array.array(b'b', data)
return n
def write(self, b):
@@ -729,6 +736,8 @@
if pos is None:
pos = self.tell()
+ # XXX: Should seek() be used, instead of passing the position
+ # XXX directly to truncate?
return self.raw.truncate(pos)
### Flush and close ###
@@ -768,7 +777,7 @@
return self.raw.isatty()
-class BytesIO(BufferedIOBase):
+class _BytesIO(BufferedIOBase):
"""Buffered I/O implementation using an in-memory bytes buffer."""
@@ -777,20 +786,28 @@
def __init__(self, initial_bytes=None):
buf = bytearray()
if initial_bytes is not None:
- buf += initial_bytes
+ buf += bytearray(initial_bytes)
self._buffer = buf
self._pos = 0
def getvalue(self):
"""Return the bytes value (contents) of the buffer
"""
+ if self.closed:
+ raise ValueError("getvalue on closed file")
return bytes(self._buffer)
def read(self, n=None):
+ if self.closed:
+ raise ValueError("read from closed file")
if n is None:
n = -1
+ if not isinstance(n, (int, long)):
+ raise TypeError("argument must be an integer")
if n < 0:
n = len(self._buffer)
+ if len(self._buffer) <= self._pos:
+ return b""
newpos = min(len(self._buffer), self._pos + n)
b = self._buffer[self._pos : newpos]
self._pos = newpos
@@ -807,6 +824,8 @@
if isinstance(b, unicode):
raise TypeError("can't write unicode to binary stream")
n = len(b)
+ if n == 0:
+ return 0
newpos = self._pos + n
if newpos > len(self._buffer):
# Inserts null bytes between the current end of the file
@@ -818,28 +837,38 @@
return n
def seek(self, pos, whence=0):
+ if self.closed:
+ raise ValueError("seek on closed file")
try:
pos = pos.__index__()
except AttributeError as err:
raise TypeError("an integer is required") # from err
if whence == 0:
- self._pos = max(0, pos)
+ if pos < 0:
+ raise ValueError("negative seek position %r" % (pos,))
+ self._pos = pos
elif whence == 1:
self._pos = max(0, self._pos + pos)
elif whence == 2:
self._pos = max(0, len(self._buffer) + pos)
else:
- raise IOError("invalid whence value")
+ raise ValueError("invalid whence value")
return self._pos
def tell(self):
+ if self.closed:
+ raise ValueError("tell on closed file")
return self._pos
def truncate(self, pos=None):
+ if self.closed:
+ raise ValueError("truncate on closed file")
if pos is None:
pos = self._pos
+ elif pos < 0:
+ raise ValueError("negative truncate position %r" % (pos,))
del self._buffer[pos:]
- return pos
+ return self.seek(pos)
def readable(self):
return True
@@ -850,6 +879,16 @@
def seekable(self):
return True
+# Use the faster implementation of BytesIO if available
+try:
+ import _bytesio
+
+ class BytesIO(_bytesio._BytesIO, BufferedIOBase):
+ __doc__ = _bytesio._BytesIO.__doc__
+
+except ImportError:
+ BytesIO = _BytesIO
+
class BufferedReader(_BufferedIOMixin):
@@ -983,6 +1022,12 @@
raise BlockingIOError(e.errno, e.strerror, overage)
return written
+ def truncate(self, pos=None):
+ self.flush()
+ if pos is None:
+ pos = self.raw.tell()
+ return self.raw.truncate(pos)
+
def flush(self):
if self.closed:
raise ValueError("flush of closed file")
@@ -1102,6 +1147,13 @@
else:
return self.raw.tell() - len(self._read_buf)
+ def truncate(self, pos=None):
+ if pos is None:
+ pos = self.tell()
+ # Use seek to flush the read buffer.
+ self.seek(pos)
+ return BufferedWriter.truncate(self)
+
def read(self, n=None):
if n is None:
n = -1
@@ -1150,11 +1202,7 @@
def truncate(self, pos = None):
"""Truncate size to pos."""
- self.flush()
- if pos is None:
- pos = self.tell()
- self.seek(pos)
- return self.buffer.truncate()
+ self._unsupported("truncate")
def readline(self):
"""Read until newline or EOF.
@@ -1351,6 +1399,12 @@
def seekable(self):
return self._seekable
+ def readable(self):
+ return self.buffer.readable()
+
+ def writable(self):
+ return self.buffer.writable()
+
def flush(self):
self.buffer.flush()
self._telling = self._seekable
@@ -1542,7 +1596,16 @@
finally:
decoder.setstate(saved_state)
+ def truncate(self, pos=None):
+ self.flush()
+ if pos is None:
+ pos = self.tell()
+ self.seek(pos)
+ return self.buffer.truncate()
+
def seek(self, cookie, whence=0):
+ if self.closed:
+ raise ValueError("tell on closed file")
if not self._seekable:
raise IOError("underlying stream is not seekable")
if whence == 1: # seek relative to current position
@@ -1629,8 +1692,12 @@
return line
def readline(self, limit=None):
+ if self.closed:
+ raise ValueError("read from closed file")
if limit is None:
limit = -1
+ if not isinstance(limit, (int, long)):
+ raise TypeError("limit must be an integer")
# Grab all the decoded text (we will rewind any extra bits later).
line = self._get_decoded_chars()
diff --git a/Lib/test/test_io.py b/Lib/test/test_io.py
index e193834..b93ce02 100644
--- a/Lib/test/test_io.py
+++ b/Lib/test/test_io.py
@@ -97,7 +97,7 @@
self.assertEqual(f.seek(-1, 2), 13)
self.assertEqual(f.tell(), 13)
self.assertEqual(f.truncate(12), 12)
- self.assertEqual(f.tell(), 13)
+ self.assertEqual(f.tell(), 12)
self.assertRaises(TypeError, f.seek, 0.0)
def read_ops(self, f, buffered=False):
@@ -142,7 +142,7 @@
self.assertEqual(f.tell(), self.LARGE + 2)
self.assertEqual(f.seek(0, 2), self.LARGE + 2)
self.assertEqual(f.truncate(self.LARGE + 1), self.LARGE + 1)
- self.assertEqual(f.tell(), self.LARGE + 2)
+ self.assertEqual(f.tell(), self.LARGE + 1)
self.assertEqual(f.seek(0, 2), self.LARGE + 1)
self.assertEqual(f.seek(-1, 2), self.LARGE)
self.assertEqual(f.read(2), b"x")
@@ -726,6 +726,7 @@
txt.write("BB\nCCC\n")
txt.write("X\rY\r\nZ")
txt.flush()
+ self.assertEquals(buf.closed, False)
self.assertEquals(buf.getvalue(), expected)
def testNewlines(self):
@@ -806,7 +807,8 @@
txt = io.TextIOWrapper(buf, encoding="ascii", newline=newline)
txt.write(data)
txt.close()
- self.assertEquals(buf.getvalue(), expected)
+ self.assertEquals(buf.closed, True)
+ self.assertRaises(ValueError, buf.getvalue)
finally:
os.linesep = save_linesep
diff --git a/Lib/test/test_memoryio.py b/Lib/test/test_memoryio.py
new file mode 100644
index 0000000..0b5ec9f
--- /dev/null
+++ b/Lib/test/test_memoryio.py
@@ -0,0 +1,416 @@
+"""Unit tests for memory-based file-like objects.
+StringIO -- for unicode strings
+BytesIO -- for bytes
+"""
+
+from __future__ import unicode_literals
+
+import unittest
+from test import test_support
+
+import io
+import sys
+import array
+
+try:
+ import _bytesio
+ has_c_implementation = True
+except ImportError:
+ has_c_implementation = False
+
+
+class MemoryTestMixin:
+
+ def write_ops(self, f, t):
+ self.assertEqual(f.write(t("blah.")), 5)
+ self.assertEqual(f.seek(0), 0)
+ self.assertEqual(f.write(t("Hello.")), 6)
+ self.assertEqual(f.tell(), 6)
+ self.assertEqual(f.seek(5), 5)
+ self.assertEqual(f.tell(), 5)
+ self.assertEqual(f.write(t(" world\n\n\n")), 9)
+ self.assertEqual(f.seek(0), 0)
+ self.assertEqual(f.write(t("h")), 1)
+ self.assertEqual(f.truncate(12), 12)
+ self.assertEqual(f.tell(), 12)
+
+ def test_write(self):
+ buf = self.buftype("hello world\n")
+ memio = self.ioclass(buf)
+
+ self.write_ops(memio, self.buftype)
+ self.assertEqual(memio.getvalue(), buf)
+ memio = self.ioclass()
+ self.write_ops(memio, self.buftype)
+ self.assertEqual(memio.getvalue(), buf)
+ self.assertRaises(TypeError, memio.write, None)
+ memio.close()
+ self.assertRaises(ValueError, memio.write, self.buftype(""))
+
+ def test_writelines(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass()
+
+ self.assertEqual(memio.writelines([buf] * 100), None)
+ self.assertEqual(memio.getvalue(), buf * 100)
+ memio.writelines([])
+ self.assertEqual(memio.getvalue(), buf * 100)
+ memio = self.ioclass()
+ self.assertRaises(TypeError, memio.writelines, [buf] + [1])
+ self.assertEqual(memio.getvalue(), buf)
+ self.assertRaises(TypeError, memio.writelines, None)
+ memio.close()
+ self.assertRaises(ValueError, memio.writelines, [])
+
+ def test_writelines_error(self):
+ memio = self.ioclass()
+ def error_gen():
+ yield self.buftype('spam')
+ raise KeyboardInterrupt
+
+ self.assertRaises(KeyboardInterrupt, memio.writelines, error_gen())
+
+ def test_truncate(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertRaises(ValueError, memio.truncate, -1)
+ memio.seek(6)
+ self.assertEqual(memio.truncate(), 6)
+ self.assertEqual(memio.getvalue(), buf[:6])
+ self.assertEqual(memio.truncate(4), 4)
+ self.assertEqual(memio.getvalue(), buf[:4])
+ self.assertEqual(memio.tell(), 4)
+ memio.write(buf)
+ self.assertEqual(memio.getvalue(), buf[:4] + buf)
+ pos = memio.tell()
+ self.assertEqual(memio.truncate(None), pos)
+ self.assertEqual(memio.tell(), pos)
+ self.assertRaises(TypeError, memio.truncate, '0')
+ memio.close()
+ self.assertRaises(ValueError, memio.truncate, 0)
+
+ def test_init(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+ self.assertEqual(memio.getvalue(), buf)
+ memio = self.ioclass(None)
+ self.assertEqual(memio.getvalue(), self.EOF)
+ memio.__init__(buf * 2)
+ self.assertEqual(memio.getvalue(), buf * 2)
+ memio.__init__(buf)
+ self.assertEqual(memio.getvalue(), buf)
+
+ def test_read(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertEqual(memio.read(0), self.EOF)
+ self.assertEqual(memio.read(1), buf[:1])
+ self.assertEqual(memio.read(4), buf[1:5])
+ self.assertEqual(memio.read(900), buf[5:])
+ self.assertEqual(memio.read(), self.EOF)
+ memio.seek(0)
+ self.assertEqual(memio.read(), buf)
+ self.assertEqual(memio.read(), self.EOF)
+ self.assertEqual(memio.tell(), 10)
+ memio.seek(0)
+ self.assertEqual(memio.read(-1), buf)
+ memio.seek(0)
+ self.assertEqual(type(memio.read()), type(buf))
+ memio.seek(100)
+ self.assertEqual(type(memio.read()), type(buf))
+ memio.seek(0)
+ self.assertEqual(memio.read(None), buf)
+ self.assertRaises(TypeError, memio.read, '')
+ memio.close()
+ self.assertRaises(ValueError, memio.read)
+
+ def test_readline(self):
+ buf = self.buftype("1234567890\n")
+ memio = self.ioclass(buf * 2)
+
+ self.assertEqual(memio.readline(0), self.EOF)
+ self.assertEqual(memio.readline(), buf)
+ self.assertEqual(memio.readline(), buf)
+ self.assertEqual(memio.readline(), self.EOF)
+ memio.seek(0)
+ self.assertEqual(memio.readline(5), buf[:5])
+ self.assertEqual(memio.readline(5), buf[5:10])
+ self.assertEqual(memio.readline(5), buf[10:15])
+ memio.seek(0)
+ self.assertEqual(memio.readline(-1), buf)
+ memio.seek(0)
+ self.assertEqual(memio.readline(0), self.EOF)
+
+ buf = self.buftype("1234567890\n")
+ memio = self.ioclass((buf * 3)[:-1])
+ self.assertEqual(memio.readline(), buf)
+ self.assertEqual(memio.readline(), buf)
+ self.assertEqual(memio.readline(), buf[:-1])
+ self.assertEqual(memio.readline(), self.EOF)
+ memio.seek(0)
+ self.assertEqual(type(memio.readline()), type(buf))
+ self.assertEqual(memio.readline(None), buf)
+ self.assertRaises(TypeError, memio.readline, '')
+ memio.close()
+ self.assertRaises(ValueError, memio.readline)
+
+ def test_readlines(self):
+ buf = self.buftype("1234567890\n")
+ memio = self.ioclass(buf * 10)
+
+ self.assertEqual(memio.readlines(), [buf] * 10)
+ memio.seek(5)
+ self.assertEqual(memio.readlines(), [buf[5:]] + [buf] * 9)
+ memio.seek(0)
+ self.assertEqual(memio.readlines(15), [buf] * 2)
+ memio.seek(0)
+ self.assertEqual(memio.readlines(-1), [buf] * 10)
+ memio.seek(0)
+ self.assertEqual(memio.readlines(0), [buf] * 10)
+ memio.seek(0)
+ self.assertEqual(type(memio.readlines()[0]), type(buf))
+ memio.seek(0)
+ self.assertEqual(memio.readlines(None), [buf] * 10)
+ self.assertRaises(TypeError, memio.readlines, '')
+ memio.close()
+ self.assertRaises(ValueError, memio.readlines)
+
+ def test_iterator(self):
+ buf = self.buftype("1234567890\n")
+ memio = self.ioclass(buf * 10)
+
+ self.assertEqual(iter(memio), memio)
+ self.failUnless(hasattr(memio, '__iter__'))
+ self.failUnless(hasattr(memio, 'next'))
+ i = 0
+ for line in memio:
+ self.assertEqual(line, buf)
+ i += 1
+ self.assertEqual(i, 10)
+ memio.seek(0)
+ i = 0
+ for line in memio:
+ self.assertEqual(line, buf)
+ i += 1
+ self.assertEqual(i, 10)
+ memio = self.ioclass(buf * 2)
+ memio.close()
+ self.assertRaises(ValueError, memio.next)
+
+ def test_getvalue(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertEqual(memio.getvalue(), buf)
+ memio.read()
+ self.assertEqual(memio.getvalue(), buf)
+ self.assertEqual(type(memio.getvalue()), type(buf))
+ memio = self.ioclass(buf * 1000)
+ self.assertEqual(memio.getvalue()[-3:], self.buftype("890"))
+ memio = self.ioclass(buf)
+ memio.close()
+ self.assertRaises(ValueError, memio.getvalue)
+
+ def test_seek(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ memio.read(5)
+ self.assertRaises(ValueError, memio.seek, -1)
+ self.assertRaises(ValueError, memio.seek, 1, -1)
+ self.assertRaises(ValueError, memio.seek, 1, 3)
+ self.assertEqual(memio.seek(0), 0)
+ self.assertEqual(memio.seek(0, 0), 0)
+ self.assertEqual(memio.read(), buf)
+ self.assertEqual(memio.seek(3), 3)
+ self.assertEqual(memio.seek(0, 1), 3)
+ self.assertEqual(memio.read(), buf[3:])
+ self.assertEqual(memio.seek(len(buf)), len(buf))
+ self.assertEqual(memio.read(), self.EOF)
+ memio.seek(len(buf) + 1)
+ self.assertEqual(memio.read(), self.EOF)
+ self.assertEqual(memio.seek(0, 2), len(buf))
+ self.assertEqual(memio.read(), self.EOF)
+ memio.close()
+ self.assertRaises(ValueError, memio.seek, 0)
+
+ def test_overseek(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertEqual(memio.seek(len(buf) + 1), 11)
+ self.assertEqual(memio.read(), self.EOF)
+ self.assertEqual(memio.tell(), 11)
+ self.assertEqual(memio.getvalue(), buf)
+ memio.write(self.EOF)
+ self.assertEqual(memio.getvalue(), buf)
+ memio.write(buf)
+ self.assertEqual(memio.getvalue(), buf + self.buftype('\0') + buf)
+
+ def test_tell(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertEqual(memio.tell(), 0)
+ memio.seek(5)
+ self.assertEqual(memio.tell(), 5)
+ memio.seek(10000)
+ self.assertEqual(memio.tell(), 10000)
+ memio.close()
+ self.assertRaises(ValueError, memio.tell)
+
+ def test_flush(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertEqual(memio.flush(), None)
+
+ def test_flags(self):
+ memio = self.ioclass()
+
+ self.assertEqual(memio.writable(), True)
+ self.assertEqual(memio.readable(), True)
+ self.assertEqual(memio.seekable(), True)
+ self.assertEqual(memio.isatty(), False)
+ self.assertEqual(memio.closed, False)
+ memio.close()
+ self.assertEqual(memio.writable(), True)
+ self.assertEqual(memio.readable(), True)
+ self.assertEqual(memio.seekable(), True)
+ self.assertRaises(ValueError, memio.isatty)
+ self.assertEqual(memio.closed, True)
+
+ def test_subclassing(self):
+ buf = self.buftype("1234567890")
+ def test1():
+ class MemIO(self.ioclass):
+ pass
+ m = MemIO(buf)
+ return m.getvalue()
+ def test2():
+ class MemIO(self.ioclass):
+ def __init__(me, a, b):
+ self.ioclass.__init__(me, a)
+ m = MemIO(buf, None)
+ return m.getvalue()
+ self.assertEqual(test1(), buf)
+ self.assertEqual(test2(), buf)
+
+
+class PyBytesIOTest(MemoryTestMixin, unittest.TestCase):
+ @staticmethod
+ def buftype(s):
+ return s.encode("ascii")
+ ioclass = io._BytesIO
+ EOF = b""
+
+ def test_read1(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertRaises(TypeError, memio.read1)
+ self.assertEqual(memio.read(), buf)
+
+ def test_readinto(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ b = bytearray(b"hello")
+ self.assertEqual(memio.readinto(b), 5)
+ self.assertEqual(b, b"12345")
+ self.assertEqual(memio.readinto(b), 5)
+ self.assertEqual(b, b"67890")
+ self.assertEqual(memio.readinto(b), 0)
+ self.assertEqual(b, b"67890")
+ b = bytearray(b"hello world")
+ memio.seek(0)
+ self.assertEqual(memio.readinto(b), 10)
+ self.assertEqual(b, b"1234567890d")
+ b = bytearray(b"")
+ memio.seek(0)
+ self.assertEqual(memio.readinto(b), 0)
+ self.assertEqual(b, b"")
+ self.assertRaises(TypeError, memio.readinto, '')
+ a = array.array(b'b', map(ord, b"hello world"))
+ memio = self.ioclass(buf)
+ memio.readinto(a)
+ self.assertEqual(a.tostring(), b"1234567890d")
+ memio.close()
+ self.assertRaises(ValueError, memio.readinto, b)
+
+ def test_relative_seek(self):
+ buf = self.buftype("1234567890")
+ memio = self.ioclass(buf)
+
+ self.assertEqual(memio.seek(-1, 1), 0)
+ self.assertEqual(memio.seek(3, 1), 3)
+ self.assertEqual(memio.seek(-4, 1), 0)
+ self.assertEqual(memio.seek(-1, 2), 9)
+ self.assertEqual(memio.seek(1, 1), 10)
+ self.assertEqual(memio.seek(1, 2), 11)
+ memio.seek(-3, 2)
+ self.assertEqual(memio.read(), buf[-3:])
+ memio.seek(0)
+ memio.seek(1, 1)
+ self.assertEqual(memio.read(), buf[1:])
+
+ def test_unicode(self):
+ memio = self.ioclass()
+
+ self.assertRaises(TypeError, self.ioclass, "1234567890")
+ self.assertRaises(TypeError, memio.write, "1234567890")
+ self.assertRaises(TypeError, memio.writelines, ["1234567890"])
+
+ def test_bytes_array(self):
+ buf = b"1234567890"
+
+ a = array.array(b'b', map(ord, buf))
+ memio = self.ioclass(a)
+ self.assertEqual(memio.getvalue(), buf)
+ self.assertEqual(memio.write(a), 10)
+ self.assertEqual(memio.getvalue(), buf)
+
+
+class PyStringIOTest(MemoryTestMixin, unittest.TestCase):
+ buftype = unicode
+ ioclass = io.StringIO
+ EOF = ""
+
+ def test_relative_seek(self):
+ memio = self.ioclass()
+
+ self.assertRaises(IOError, memio.seek, -1, 1)
+ self.assertRaises(IOError, memio.seek, 3, 1)
+ self.assertRaises(IOError, memio.seek, -3, 1)
+ self.assertRaises(IOError, memio.seek, -1, 2)
+ self.assertRaises(IOError, memio.seek, 1, 1)
+ self.assertRaises(IOError, memio.seek, 1, 2)
+
+ # XXX: For the Python version of io.StringIO, this is highly
+ # dependent on the encoding used for the underlying buffer.
+ # def test_widechar(self):
+ # buf = self.buftype("\U0002030a\U00020347")
+ # memio = self.ioclass(buf)
+ #
+ # self.assertEqual(memio.getvalue(), buf)
+ # self.assertEqual(memio.write(buf), len(buf))
+ # self.assertEqual(memio.tell(), len(buf))
+ # self.assertEqual(memio.getvalue(), buf)
+ # self.assertEqual(memio.write(buf), len(buf))
+ # self.assertEqual(memio.tell(), len(buf) * 2)
+ # self.assertEqual(memio.getvalue(), buf + buf)
+
+if has_c_implementation:
+ class CBytesIOTest(PyBytesIOTest):
+ ioclass = io.BytesIO
+
+def test_main():
+ tests = [PyBytesIOTest, PyStringIOTest]
+ if has_c_implementation:
+ tests.extend([CBytesIOTest])
+ test_support.run_unittest(*tests)
+
+if __name__ == '__main__':
+ test_main()