[project @ Arch-1:robey@lag.net--2003-public%secsh--dev--1.0--patch-164]
fix some docs remove some epydoc comments about fileno() being non-portable.
This commit is contained in:
parent
89ccac6bfa
commit
bd7911895b
|
@ -78,7 +78,7 @@ class Channel (object):
|
|||
self.name = str(chanid)
|
||||
self.logger = util.get_logger('paramiko.chan.' + str(chanid))
|
||||
self.pipe_rfd = self.pipe_wfd = None
|
||||
self.pipe_set = False
|
||||
self.pipe_set = False
|
||||
self.event = threading.Event()
|
||||
self.combine_stderr = False
|
||||
self.exit_status = -1
|
||||
|
@ -470,8 +470,6 @@ class Channel (object):
|
|||
@return: C{True} if a L{recv} call on this channel would immediately
|
||||
return at least one byte; C{False} otherwise.
|
||||
@rtype: boolean
|
||||
|
||||
@note: This method doesn't work if you've called L{fileno}.
|
||||
"""
|
||||
self.lock.acquire()
|
||||
try:
|
||||
|
@ -682,8 +680,8 @@ class Channel (object):
|
|||
sent.
|
||||
|
||||
@note: If the channel is closed while only part of the data hase been
|
||||
sent, there is no way to determine how much data (if any) was sent.
|
||||
This is irritating, but identically follows python's API.
|
||||
sent, there is no way to determine how much data (if any) was sent.
|
||||
This is irritating, but identically follows python's API.
|
||||
"""
|
||||
while s:
|
||||
if self.closed:
|
||||
|
@ -719,10 +717,9 @@ class Channel (object):
|
|||
|
||||
def makefile(self, *params):
|
||||
"""
|
||||
Return a file-like object associated with this channel, without the
|
||||
non-portable side effects of L{fileno}. The optional C{mode} and
|
||||
C{bufsize} arguments are interpreted the same way as by the built-in
|
||||
C{file()} function in python.
|
||||
Return a file-like object associated with this channel. The optional
|
||||
C{mode} and C{bufsize} arguments are interpreted the same way as by
|
||||
the built-in C{file()} function in python.
|
||||
|
||||
@return: object which can be used for python file I/O.
|
||||
@rtype: L{ChannelFile}
|
||||
|
@ -756,14 +753,14 @@ class Channel (object):
|
|||
The first time C{fileno} is called on a channel, a pipe is created to
|
||||
simulate real OS-level file descriptor (FD) behavior. Because of this,
|
||||
two OS-level FDs are created, which will use up FDs faster than normal.
|
||||
You won't notice this effect unless you open hundreds or thousands of
|
||||
channels simultaneously, but it's still notable.
|
||||
You won't notice this effect unless you open hundreds or thousands of
|
||||
channels simultaneously, but it's still notable.
|
||||
|
||||
@return: an OS-level file descriptor
|
||||
@rtype: int
|
||||
|
||||
@warning: This method causes channel reads to be slightly less
|
||||
efficient.
|
||||
efficient.
|
||||
"""
|
||||
self.lock.acquire()
|
||||
try:
|
||||
|
@ -1095,13 +1092,12 @@ class Channel (object):
|
|||
class ChannelFile (BufferedFile):
|
||||
"""
|
||||
A file-like wrapper around L{Channel}. A ChannelFile is created by calling
|
||||
L{Channel.makefile} and doesn't have the non-portable side effect of
|
||||
L{Channel.fileno}.
|
||||
L{Channel.makefile}.
|
||||
|
||||
@bug: To correctly emulate the file object created from a socket's
|
||||
C{makefile} method, a L{Channel} and its C{ChannelFile} should be able to
|
||||
be closed or garbage-collected independently. Currently, closing the
|
||||
C{ChannelFile} does nothing but flush the buffer.
|
||||
C{makefile} method, a L{Channel} and its C{ChannelFile} should be able
|
||||
to be closed or garbage-collected independently. Currently, closing
|
||||
the C{ChannelFile} does nothing but flush the buffer.
|
||||
"""
|
||||
|
||||
def __init__(self, channel, mode = 'r', bufsize = -1):
|
||||
|
|
Loading…
Reference in New Issue