Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Fetching contributors…

Cannot retrieve contributors at this time

1532 lines (1145 sloc) 54.761 kB
#===============================================================================
# Copyright (C) 2011-2012 by Andrew Moffat
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.
#===============================================================================
__version__ = "1.02"
__project_url__ = "https://github.com/amoffat/sh"
import platform
if "windows" in platform.system().lower():
raise ImportError("sh 1.0 is currently only supported on linux and osx. \
please install pbs 0.109 (http://pypi.python.org/pypi/pbs) for windows support.")
import sys
IS_PY3 = sys.version_info[0] == 3
import traceback
import os
import re
from glob import glob as original_glob
import shlex
from types import ModuleType
from functools import partial
import inspect
import time as _time
if IS_PY3:
from io import StringIO
from io import BytesIO as cStringIO
from queue import Queue, Empty
else:
from StringIO import StringIO
from cStringIO import OutputType as cStringIO
from Queue import Queue, Empty
IS_OSX = platform.system() == "Darwin"
THIS_DIR = os.path.dirname(os.path.realpath(__file__))
import errno
import warnings
from threading import Thread, Event
import pty
import termios
import signal
import select
import atexit
import gc
import threading
import tty
import pickle
import fcntl
import struct
import resource
from collections import deque
import logging
# this is ugly, but we've added a module-level logging kill switch. the reason
# for it (vs letting the user disable/enable logging through the logging
# module's facilities) is because to enable/disable logging using logging
# facilities, modules need to have their loggers (retrieved with
# logging.getLogger()) named using dot notation. so for example:
#
# log = logging.getLogger("sh.process")
#
# we don't do that though, because we cram a lot of info into the logger name
# for example, a logger name may be
# "<Process 1373 ['/usr/bin/python3.2', '/tmp/tmp2c18zp']>"
# because of this, a user can't disable our loggers (because we lack dot
# notation), and I won't add dot notation because I can't include all the
# data i need in my logger name. so this is really a shortcoming of the
# logging module.
logging_enabled = False
if IS_PY3:
raw_input = input
unicode = str
basestring = str
class ErrorReturnCode(Exception):
truncate_cap = 750
def __init__(self, full_cmd, stdout, stderr):
self.full_cmd = full_cmd
self.stdout = stdout
self.stderr = stderr
if self.stdout is None: tstdout = "<redirected>"
else:
tstdout = self.stdout[:self.truncate_cap]
out_delta = len(self.stdout) - len(tstdout)
if out_delta:
tstdout += ("... (%d more, please see e.stdout)" % out_delta).encode()
if self.stderr is None: tstderr = "<redirected>"
else:
tstderr = self.stderr[:self.truncate_cap]
err_delta = len(self.stderr) - len(tstderr)
if err_delta:
tstderr += ("... (%d more, please see e.stderr)" % err_delta).encode()
msg = "\n\n RAN: %r\n\n STDOUT:\n%s\n\n STDERR:\n%s" %\
(full_cmd, tstdout.decode(), tstderr.decode())
super(ErrorReturnCode, self).__init__(msg)
class CommandNotFound(Exception): pass
rc_exc_regex = re.compile("ErrorReturnCode_(\d+)")
rc_exc_cache = {}
def get_rc_exc(rc):
rc = int(rc)
try: return rc_exc_cache[rc]
except KeyError: pass
name = "ErrorReturnCode_%d" % rc
exc = type(name, (ErrorReturnCode,), {})
rc_exc_cache[rc] = exc
return exc
def which(program):
def is_exe(fpath):
return os.path.exists(fpath) and os.access(fpath, os.X_OK)
fpath, fname = os.path.split(program)
if fpath:
if is_exe(program): return program
else:
if "PATH" not in os.environ: return None
for path in os.environ["PATH"].split(os.pathsep):
exe_file = os.path.join(path, program)
if is_exe(exe_file):
return exe_file
return None
def resolve_program(program):
path = which(program)
if not path:
# our actual command might have a dash in it, but we can't call
# that from python (we have to use underscores), so we'll check
# if a dash version of our underscore command exists and use that
# if it does
if "_" in program: path = which(program.replace("_", "-"))
if not path: return None
return path
# we add this thin wrapper to glob.glob because of a specific edge case where
# glob does not expand to anything. for example, if you try to do
# glob.glob("*.py") and there are no *.py files in the directory, glob.glob
# returns an empty list. this empty list gets passed to the command, and
# then the command fails with a misleading error message. this thin wrapper
# ensures that if there is no expansion, we pass in the original argument,
# so that when the command fails, the error message is clearer
def glob(arg):
return original_glob(arg) or arg
class RunningCommand(object):
def __init__(self, cmd, call_args, stdin, stdout, stderr):
self.call_args = call_args
self.cmd = cmd
self.ran = " ".join(cmd)
self.process = None
self.should_wait = True
spawn_process = True
# with contexts shouldn't run at all yet, they prepend
# to every command in the context
if call_args["with"]:
spawn_process = False
Command._prepend_stack.append(cmd)
if callable(call_args["out"]) or callable(call_args["err"]):
self.should_wait = False
if call_args["piped"] or call_args["iter"] or call_args["iter_noblock"]:
self.should_wait = False
# we're running in the background, return self and let us lazily
# evaluate
if call_args["bg"]: self.should_wait = False
# redirection
if call_args["err_to_out"]: stderr = STDOUT
# set up which stream should write to the pipe
# TODO, make pipe None by default and limit the size of the Queue
# in oproc.OProc
pipe = STDOUT
if call_args["iter"] == "out" or call_args["iter"] is True: pipe = STDOUT
elif call_args["iter"] == "err": pipe = STDERR
if call_args["iter_noblock"] == "out" or call_args["iter_noblock"] is True: pipe = STDOUT
elif call_args["iter_noblock"] == "err": pipe = STDERR
if spawn_process:
self.process = OProc(cmd, stdin, stdout, stderr,
self.call_args, pipe=pipe)
if self.should_wait:
self.wait()
def wait(self):
self._handle_exit_code(self.process.wait())
return self
# here we determine if we had an exception, or an error code that we weren't
# expecting to see. if we did, we create and raise an exception
def _handle_exit_code(self, code):
if code not in self.call_args["ok_code"] and code >= 0: raise get_rc_exc(code)(
" ".join(self.cmd),
self.process.stdout,
self.process.stderr
)
@property
def stdout(self):
self.wait()
return self.process.stdout
@property
def stderr(self):
self.wait()
return self.process.stderr
@property
def exit_code(self):
self.wait()
return self.process.exit_code
@property
def pid(self):
return self.process.pid
def __len__(self):
return len(str(self))
def __enter__(self):
# we don't actually do anything here because anything that should
# have been done would have been done in the Command.__call__ call.
# essentially all that has to happen is the comand be pushed on
# the prepend stack.
pass
def __iter__(self):
return self
def next(self):
# we do this because if get blocks, we can't catch a KeyboardInterrupt
# so the slight timeout allows for that.
while True:
try: chunk = self.process._pipe_queue.get(False, .001)
except Empty:
if self.call_args["iter_noblock"]: return errno.EWOULDBLOCK
else:
if chunk is None:
self.wait()
raise StopIteration()
try: return chunk.decode(self.call_args["encoding"])
except UnicodeDecodeError: return chunk
# python 3
__next__ = next
def __exit__(self, typ, value, traceback):
if self.call_args["with"] and Command._prepend_stack:
Command._prepend_stack.pop()
def __str__(self):
if IS_PY3: return self.__unicode__()
else: return unicode(self).encode(self.call_args["encoding"])
def __unicode__(self):
if self.process:
if self.stdout: return self.stdout.decode(self.call_args["encoding"])
return ""
def __eq__(self, other):
return unicode(self) == unicode(other)
def __contains__(self, item):
return item in str(self)
def __getattr__(self, p):
# let these three attributes pass through to the OProc object
if p in ("signal", "terminate", "kill"):
if self.process: return getattr(self.process, p)
else: raise AttributeError
return getattr(unicode(self), p)
def __repr__(self):
try: return str(self)
except UnicodeDecodeError:
if self.process:
if self.stdout: return repr(self.stdout)
return repr("")
def __long__(self):
return long(str(self).strip())
def __float__(self):
return float(str(self).strip())
def __int__(self):
return int(str(self).strip())
class Command(object):
_prepend_stack = []
_call_args = {
# currently unsupported
#"fg": False, # run command in foreground
"bg": False, # run command in background
"with": False, # prepend the command to every command after it
"in": None,
"out": None, # redirect STDOUT
"err": None, # redirect STDERR
"err_to_out": None, # redirect STDERR to STDOUT
# stdin buffer size
# 1 for line, 0 for unbuffered, any other number for that amount
"in_bufsize": 0,
# stdout buffer size, same values as above
"out_bufsize": 1,
"err_bufsize": 1,
# this is how big the output buffers will be for stdout and stderr.
# this is essentially how much output they will store from the process.
# we use a deque, so if it overflows past this amount, the first items
# get pushed off as each new item gets added.
#
# NOTICE
# this is not a *BYTE* size, this is a *CHUNK* size...meaning, that if
# you're buffering out/err at 1024 bytes, the internal buffer size will
# be "internal_bufsize" CHUNKS of 1024 bytes
"internal_bufsize": 3 * 1024**2,
"env": None,
"piped": None,
"iter": None,
"iter_noblock": None,
"ok_code": 0,
"cwd": None,
# this is for programs that expect their input to be from a terminal.
# ssh is one of those programs
"tty_in": False,
"tty_out": True,
"encoding": "utf8",
# how long the process should run before it is auto-killed
"timeout": 0,
}
# these are arguments that cannot be called together, because they wouldn't
# make any sense
_incompatible_call_args = (
#("fg", "bg", "Command can't be run in the foreground and background"),
("err", "err_to_out", "Stderr is already being redirected"),
("piped", "iter", "You cannot iterate when this command is being piped"),
)
@classmethod
def _create(cls, program):
path = resolve_program(program)
if not path: raise CommandNotFound(program)
return cls(path)
def __init__(self, path):
self._path = path
self._partial = False
self._partial_baked_args = []
self._partial_call_args = {}
def __getattribute__(self, name):
# convenience
getattr = partial(object.__getattribute__, self)
if name.startswith("_"): return getattr(name)
if name == "bake": return getattr("bake")
return getattr("bake")(name)
@staticmethod
def _extract_call_args(kwargs, to_override={}):
kwargs = kwargs.copy()
call_args = {}
for parg, default in Command._call_args.items():
key = "_" + parg
if key in kwargs:
call_args[parg] = kwargs[key]
del kwargs[key]
elif parg in to_override:
call_args[parg] = to_override[parg]
# test for incompatible call args
s1 = set(call_args.keys())
for args in Command._incompatible_call_args:
args = list(args)
error = args.pop()
if s1.issuperset(args):
raise TypeError("Invalid special arguments %r: %s" % (args, error))
return call_args, kwargs
def _format_arg(self, arg):
if IS_PY3: arg = str(arg)
else: arg = unicode(arg).encode("utf8")
return arg
def _compile_args(self, args, kwargs):
processed_args = []
# aggregate positional args
for arg in args:
if isinstance(arg, (list, tuple)):
if not arg:
warnings.warn("Empty list passed as an argument to %r. \
If you're using glob.glob(), please use sh.glob() instead." % self.path, stacklevel=3)
for sub_arg in arg: processed_args.append(self._format_arg(sub_arg))
else: processed_args.append(self._format_arg(arg))
# aggregate the keyword arguments
for k,v in kwargs.items():
# we're passing a short arg as a kwarg, example:
# cut(d="\t")
if len(k) == 1:
processed_args.append("-"+k)
if v is not True: processed_args.append(self._format_arg(v))
# we're doing a long arg
else:
k = k.replace("_", "-")
if v is True: processed_args.append("--"+k)
else: processed_args.append("--%s=%s" % (k, self._format_arg(v)))
return processed_args
def bake(self, *args, **kwargs):
fn = Command(self._path)
fn._partial = True
call_args, kwargs = self._extract_call_args(kwargs)
pruned_call_args = call_args
for k,v in Command._call_args.items():
try:
if pruned_call_args[k] == v:
del pruned_call_args[k]
except KeyError: continue
fn._partial_call_args.update(self._partial_call_args)
fn._partial_call_args.update(pruned_call_args)
fn._partial_baked_args.extend(self._partial_baked_args)
fn._partial_baked_args.extend(self._compile_args(args, kwargs))
return fn
def __str__(self):
if IS_PY3: return self.__unicode__()
else: return unicode(self).encode("utf8")
def __eq__(self, other):
try: return str(self) == str(other)
except: return False
def __repr__(self):
return str(self)
def __unicode__(self):
baked_args = " ".join(self._partial_baked_args)
if baked_args: baked_args = " " + baked_args
return self._path + baked_args
def __enter__(self):
self(_with=True)
def __exit__(self, typ, value, traceback):
Command._prepend_stack.pop()
def __call__(self, *args, **kwargs):
kwargs = kwargs.copy()
args = list(args)
cmd = []
# aggregate any 'with' contexts
for prepend in self._prepend_stack: cmd.extend(prepend)
cmd.append(self._path)
# here we extract the special kwargs and override any
# special kwargs from the possibly baked command
tmp_call_args, kwargs = self._extract_call_args(kwargs, self._partial_call_args)
call_args = Command._call_args.copy()
call_args.update(tmp_call_args)
if not isinstance(call_args["ok_code"], (tuple, list)):
call_args["ok_code"] = [call_args["ok_code"]]
# check if we're piping via composition
stdin = call_args["in"]
if args:
first_arg = args.pop(0)
if isinstance(first_arg, RunningCommand):
# it makes sense that if the input pipe of a command is running
# in the background, then this command should run in the
# background as well
if first_arg.call_args["bg"]: call_args["bg"] = True
stdin = first_arg.process._pipe_queue
else:
args.insert(0, first_arg)
processed_args = self._compile_args(args, kwargs)
# makes sure our arguments are broken up correctly
split_args = self._partial_baked_args + processed_args
final_args = split_args
cmd.extend(final_args)
# stdout redirection
stdout = call_args["out"]
if stdout \
and not callable(stdout) \
and not hasattr(stdout, "write") \
and not isinstance(stdout, (cStringIO, StringIO)):
stdout = open(str(stdout), "wb")
# stderr redirection
stderr = call_args["err"]
if stderr and not callable(stderr) and not hasattr(stderr, "write") \
and not isinstance(stderr, (cStringIO, StringIO)):
stderr = open(str(err), "wb")
return RunningCommand(cmd, call_args, stdin, stdout, stderr)
# used in redirecting
STDOUT = -1
STDERR = -2
# Process open = Popen
# Open Process = OProc
class OProc(object):
_procs_to_cleanup = []
_registered_cleanup = False
_default_window_size = (24, 80)
def __init__(self, cmd, stdin, stdout, stderr, call_args,
persist=False, pipe=STDOUT):
self.call_args = call_args
self._single_tty = self.call_args["tty_in"] and self.call_args["tty_out"]
# this logic is a little convoluted, but basically this top-level
# if/else is for consolidating input and output TTYs into a single
# TTY. this is the only way some secure programs like ssh will
# output correctly (is if stdout and stdin are both the same TTY)
if self._single_tty:
self._stdin_fd, self._slave_stdin_fd = pty.openpty()
self._stdout_fd = self._stdin_fd
self._slave_stdout_fd = self._slave_stdin_fd
self._stderr_fd = self._stdin_fd
self._slave_stderr_fd = self._slave_stdin_fd
# do not consolidate stdin and stdout
else:
if self.call_args["tty_in"]:
self._slave_stdin_fd, self._stdin_fd = pty.openpty()
else:
self._slave_stdin_fd, self._stdin_fd = os.pipe()
# tty_out is usually the default
if self.call_args["tty_out"]:
self._stdout_fd, self._slave_stdout_fd = pty.openpty()
else:
self._stdout_fd, self._slave_stdout_fd = os.pipe()
# unless STDERR is going to STDOUT, it ALWAYS needs to be a pipe,
# and never a PTY. the reason for this is not totally clear to me,
# but it has to do with the fact that if STDERR isn't set as the
# CTTY (because STDOUT is), the STDERR buffer won't always flush
# by the time the process exits, and the data will be lost.
# i've only seen this on OSX.
if stderr is not STDOUT:
self._stderr_fd, self._slave_stderr_fd = os.pipe()
self.pid = os.fork()
# child
if self.pid == 0:
# this piece of ugliness is due to a bug where we can lose output
# if we do os.close(self._slave_stdout_fd) in the parent after
# the child starts writing.
# see http://bugs.python.org/issue15898
if IS_OSX and IS_PY3: _time.sleep(0.01)
os.setsid()
if self.call_args["tty_out"]:
# set raw mode, so there isn't any weird translation of newlines
# to \r\n and other oddities. we're not outputting to a terminal
# anyways
#
# we HAVE to do this here, and not in the parent thread, because
# we have to guarantee that this is set before the child process
# is run, and we can't do it twice.
tty.setraw(self._stdout_fd)
os.close(self._stdin_fd)
if not self._single_tty:
os.close(self._stdout_fd)
if stderr is not STDOUT: os.close(self._stderr_fd)
if self.call_args["cwd"]: os.chdir(self.call_args["cwd"])
os.dup2(self._slave_stdin_fd, 0)
os.dup2(self._slave_stdout_fd, 1)
# we're not directing stderr to stdout? then set self._slave_stderr_fd to
# fd 2, the common stderr fd
if stderr is STDOUT: os.dup2(self._slave_stdout_fd, 2)
else: os.dup2(self._slave_stderr_fd, 2)
# don't inherit file descriptors
max_fd = resource.getrlimit(resource.RLIMIT_NOFILE)[0]
os.closerange(3, max_fd)
# set our controlling terminal
if self.call_args["tty_out"]:
tmp_fd = os.open(os.ttyname(1), os.O_RDWR)
os.close(tmp_fd)
if self.call_args["tty_out"]:
self.setwinsize(1)
# actually execute the process
if self.call_args["env"] is None: os.execv(cmd[0], cmd)
else: os.execve(cmd[0], cmd, self.call_args["env"])
os._exit(255)
# parent
else:
if not OProc._registered_cleanup:
atexit.register(OProc._cleanup_procs)
OProc._registered_cleanup = True
self.started = _time.time()
self.cmd = cmd
self.exit_code = None
self._done_callbacks = []
self.stdin = stdin or Queue()
self._pipe_queue = Queue()
# this is used to prevent a race condition when we're waiting for
# a process to end, and the OProc's internal threads are also checking
# for the processes's end
self._wait_lock = threading.Lock()
# these are for aggregating the stdout and stderr. we use a deque
# because we don't want to overflow
self._stdout = deque(maxlen=self.call_args["internal_bufsize"])
self._stderr = deque(maxlen=self.call_args["internal_bufsize"])
if self.call_args["tty_in"]: self.setwinsize(self._stdin_fd)
self.log = logging.getLogger("process %r" % self)
os.close(self._slave_stdin_fd)
if not self._single_tty:
os.close(self._slave_stdout_fd)
if stderr is not STDOUT: os.close(self._slave_stderr_fd)
if logging_enabled: self.log.debug("started process")
if not persist: OProc._procs_to_cleanup.append(self)
if self.call_args["tty_in"]:
attr = termios.tcgetattr(self._stdin_fd)
attr[3] &= ~termios.ECHO
termios.tcsetattr(self._stdin_fd, termios.TCSANOW, attr)
# this represents the connection from a Queue object (or whatever
# we're using to feed STDIN) to the process's STDIN fd
self._stdin_stream = StreamWriter("stdin", self, self._stdin_fd,
self.stdin, self.call_args["in_bufsize"])
stdout_pipe = self._pipe_queue if pipe is STDOUT else None
# this represents the connection from a process's STDOUT fd to
# wherever it has to go, sometimes a pipe Queue (that we will use
# to pipe data to other processes), and also an internal deque
# that we use to aggregate all the output
self._stdout_stream = StreamReader("stdout", self, self._stdout_fd, stdout,
self._stdout, self.call_args["out_bufsize"], stdout_pipe)
if stderr is STDOUT or self._single_tty: self._stderr_stream = None
else:
stderr_pipe = self._pipe_queue if pipe is STDERR else None
self._stderr_stream = StreamReader("stderr", self, self._stderr_fd, stderr,
self._stderr, self.call_args["err_bufsize"], stderr_pipe)
# start the main io threads
self._input_thread = self._start_thread(self.input_thread, self._stdin_stream)
self._output_thread = self._start_thread(self.output_thread, self._stdout_stream, self._stderr_stream)
def __repr__(self):
return "<Process %d %r>" % (self.pid, self.cmd)
# also borrowed from pexpect.py
@staticmethod
def setwinsize(fd):
rows, cols = OProc._default_window_size
TIOCSWINSZ = getattr(termios, 'TIOCSWINSZ', -2146929561)
if TIOCSWINSZ == 2148037735: # L is not required in Python >= 2.2.
TIOCSWINSZ = -2146929561 # Same bits, but with sign.
s = struct.pack('HHHH', rows, cols, 0, 0)
fcntl.ioctl(fd, TIOCSWINSZ, s)
@staticmethod
def _start_thread(fn, *args):
thrd = threading.Thread(target=fn, args=args)
thrd.daemon = True
thrd.start()
return thrd
def in_bufsize(self, buf):
self._stdin_stream.stream_bufferer.change_buffering(buf)
def out_bufsize(self, buf):
self._stdout_stream.stream_bufferer.change_buffering(buf)
def err_bufsize(self, buf):
if self._stderr_stream:
self._stderr_stream.stream_bufferer.change_buffering(buf)
def input_thread(self, stdin):
done = False
while not done and self.alive:
if logging_enabled: self.log.debug("%r ready for more input", stdin)
done = stdin.write()
stdin.close()
def output_thread(self, stdout, stderr):
readers = []
errors = []
if stdout is not None:
readers.append(stdout)
errors.append(stdout)
if stderr is not None:
readers.append(stderr)
errors.append(stderr)
while readers:
outputs, inputs, err = select.select(readers, [], errors, 0.1)
# stdout and stderr
for stream in outputs:
if logging_enabled: self.log.debug("%r ready to be read from", stream)
done = stream.read()
if done: readers.remove(stream)
for stream in err:
pass
# test if the process has been running too long
if self.call_args["timeout"]:
now = _time.time()
if now - self.started > self.call_args["timeout"]:
if logging_enabled: self.log.debug("we've been running too long")
self.kill()
# this is here because stdout may be the controlling TTY, and
# we can't close it until the process has ended, otherwise the
# child will get SIGHUP. typically, if we've broken out of
# the above loop, and we're here, the process is just about to
# end, so it's probably ok to aggressively poll self.alive
#
# the other option to this would be to do the CTTY close from
# the method that does the actual os.waitpid() call, but the
# problem with that is that the above loop might still be
# running, and closing the fd will cause some operation to
# fail. this is less complex than wrapping all the ops
# in the above loop with out-of-band fd-close exceptions
while self.alive: _time.sleep(0.001)
if stdout: stdout.close()
if stderr: stderr.close()
@property
def stdout(self):
return "".encode(self.call_args["encoding"]).join(self._stdout)
@property
def stderr(self):
return "".encode(self.call_args["encoding"]).join(self._stderr)
def signal(self, sig):
if logging_enabled: self.log.debug("sending signal %d", sig)
try: os.kill(self.pid, sig)
except OSError: pass
def kill(self):
if logging_enabled: self.log.debug("killing")
self.signal(signal.SIGKILL)
def terminate(self):
if logging_enabled: self.log.debug("terminating")
self.signal(signal.SIGTERM)
@staticmethod
def _cleanup_procs():
for proc in OProc._procs_to_cleanup:
proc.kill()
proc.wait()
def _handle_exit_code(self, exit_code):
# if we exited from a signal, let our exit code reflect that
if os.WIFSIGNALED(exit_code): return -os.WTERMSIG(exit_code)
# otherwise just give us a normal exit code
elif os.WIFEXITED(exit_code): return os.WEXITSTATUS(exit_code)
else: raise RuntimeError("Unknown child exit status!")
@property
def alive(self):
if self.exit_code is not None: return False
# what we're doing here essentially is making sure that the main thread
# (or another thread), isn't calling .wait() on the process. because
# .wait() calls os.waitpid(self.pid, 0), we can't do an os.waitpid
# here...because if we did, and the process exited while in this
# thread, the main thread's os.waitpid(self.pid, 0) would raise OSError
# (because the process ended in another thread).
#
# so essentially what we're doing is, using this lock, checking if
# we're calling .wait(), and if we are, let .wait() get the exit code
# and handle the status, otherwise let us do it.
acquired = self._wait_lock.acquire(False)
if not acquired:
if self.exit_code is not None: return False
return True
try:
# WNOHANG is just that...we're calling waitpid without hanging...
# essentially polling the process
pid, exit_code = os.waitpid(self.pid, os.WNOHANG)
if pid == self.pid:
self.exit_code = self._handle_exit_code(exit_code)
return False
# no child process
except OSError: return False
else: return True
finally: self._wait_lock.release()
def wait(self):
if logging_enabled: self.log.debug("acquiring wait lock to wait for completion")
with self._wait_lock:
if logging_enabled: self.log.debug("got wait lock")
if self.exit_code is None:
if logging_enabled: self.log.debug("exit code not set, waiting on pid")
pid, exit_code = os.waitpid(self.pid, 0)
self.exit_code = self._handle_exit_code(exit_code)
else:
if logging_enabled: self.log.debug("exit code already set (%d), no need to wait", self.exit_code)
self._input_thread.join()
self._output_thread.join()
for cb in self._done_callbacks: cb()
return self.exit_code
class DoneReadingStdin(Exception): pass
class NoStdinData(Exception): pass
# this guy is for reading from some input (the stream) and writing to our
# opened process's stdin fd. the stream can be a Queue, a callable, something
# with the "read" method, a string, or an iterable
class StreamWriter(object):
def __init__(self, name, process, stream, stdin, bufsize):
self.name = name
self.process = process
self.stream = stream
self.stdin = stdin
self.log = logging.getLogger(repr(self))
self.stream_bufferer = StreamBufferer(self.process.call_args["encoding"],
bufsize)
# determine buffering for reading from the input we set for stdin
if bufsize == 1: self.bufsize = 1024
elif bufsize == 0: self.bufsize = 1
else: self.bufsize = bufsize
if isinstance(stdin, Queue):
log_msg = "queue"
self.get_chunk = self.get_queue_chunk
elif callable(stdin):
log_msg = "callable"
self.get_chunk = self.get_callable_chunk
# also handles stringio
elif hasattr(stdin, "read"):
log_msg = "file descriptor"
self.get_chunk = self.get_file_chunk
elif isinstance(stdin, basestring):
log_msg = "string"
if bufsize == 1:
# TODO, make the split() be a generator
self.stdin = iter((c+"\n" for c in stdin.split("\n")))
else:
self.stdin = iter(stdin[i:i+self.bufsize] for i in range(0, len(stdin), self.bufsize))
self.get_chunk = self.get_iter_chunk
else:
log_msg = "general iterable"
self.stdin = iter(stdin)
self.get_chunk = self.get_iter_chunk
if logging_enabled: self.log.debug("parsed stdin as a %s", log_msg)
def __repr__(self):
return "<StreamWriter %s for %r>" % (self.name, self.process)
def fileno(self):
return self.stream
def get_queue_chunk(self):
try: chunk = self.stdin.get(True, 0.01)
except Empty: raise NoStdinData
if chunk is None: raise DoneReadingStdin
return chunk
def get_callable_chunk(self):
try: return self.stdin()
except: raise DoneReadingStdin
def get_iter_chunk(self):
try:
if IS_PY3: return self.stdin.__next__()
else: return self.stdin.next()
except StopIteration: raise DoneReadingStdin
def get_file_chunk(self):
if self.stream_bufferer.type == 1: chunk = self.stdin.readline()
else: chunk = self.stdin.read(self.bufsize)
if not chunk: raise DoneReadingStdin
else: return chunk
# the return value answers the questions "are we done writing forever?"
def write(self):
# get_chunk may sometimes return bytes, and sometimes returns trings
# because of the nature of the different types of STDIN objects we
# support
try: chunk = self.get_chunk()
except DoneReadingStdin:
if logging_enabled: self.log.debug("done reading")
if self.process.call_args["tty_in"]:
# EOF time
try: char = termios.tcgetattr(self.stream)[6][termios.VEOF]
except: char = chr(4).encode()
os.write(self.stream, char)
return True
except NoStdinData:
if logging_enabled: self.log.debug("received no data")
return False
# if we're not bytes, make us bytes
if IS_PY3 and hasattr(chunk, "encode"):
chunk = chunk.encode(self.process.call_args["encoding"])
for chunk in self.stream_bufferer.process(chunk):
if logging_enabled: self.log.debug("got chunk size %d: %r", len(chunk), chunk[:30])
if logging_enabled: self.log.debug("writing chunk to process")
try:
os.write(self.stream, chunk)
except OSError:
if logging_enabled: self.log.debug("OSError writing stdin chunk")
return True
def close(self):
if logging_enabled: self.log.debug("closing, but flushing first")
chunk = self.stream_bufferer.flush()
if logging_enabled: self.log.debug("got chunk size %d to flush: %r", len(chunk), chunk[:30])
try:
if chunk: os.write(self.stream, chunk)
if not self.process.call_args["tty_in"]:
if logging_enabled: self.log.debug("we used a TTY, so closing the stream")
os.close(self.stream)
except OSError: pass
class StreamReader(object):
def __init__(self, name, process, stream, handler, buffer, bufsize, pipe_queue=None):
self.name = name
self.process = process
self.stream = stream
self.buffer = buffer
self.pipe_queue = pipe_queue
self.log = logging.getLogger(repr(self))
self.stream_bufferer = StreamBufferer(self.process.call_args["encoding"],
bufsize)
# determine buffering
if bufsize == 1: self.bufsize = 1024
elif bufsize == 0: self.bufsize = 1
else: self.bufsize = bufsize
# here we're determining the handler type by doing some basic checks
# on the handler object
self.handler = handler
if callable(handler): self.handler_type = "fn"
elif isinstance(handler, StringIO): self.handler_type = "stringio"
elif isinstance(handler, cStringIO):
self.handler_type = "cstringio"
elif hasattr(handler, "write"): self.handler_type = "fd"
else: self.handler_type = None
self.should_quit = False
# here we choose how to call the callback, depending on how many
# arguments it takes. the reason for this is to make it as easy as
# possible for people to use, without limiting them. a new user will
# assume the callback takes 1 argument (the data). as they get more
# advanced, they may want to terminate the process, or pass some stdin
# back, and will realize that they can pass a callback of more args
if self.handler_type == "fn":
implied_arg = 0
if inspect.ismethod(handler):
implied_arg = 1
num_args = len(inspect.getargspec(handler).args)
else:
if inspect.isfunction(handler):
num_args = len(inspect.getargspec(handler).args)
# is an object instance with __call__ method
else:
implied_arg = 1
num_args = len(inspect.getargspec(handler.__call__).args)
self.handler_args = ()
if num_args == implied_arg + 2: self.handler_args = (self.process.stdin,)
elif num_args == implied_arg + 3: self.handler_args = (self.process.stdin, self.process)
def fileno(self):
return self.stream
def __repr__(self):
return "<StreamReader %s for %r>" % (self.name, self.process)
def close(self):
chunk = self.stream_bufferer.flush()
if logging_enabled: self.log.debug("got chunk size %d to flush: %r", len(chunk), chunk[:30])
if chunk: self.write_chunk(chunk)
if self.handler_type == "fd" and hasattr(self.handler, "close"):
self.handler.flush()
if self.pipe_queue: self.pipe_queue.put(None)
try: os.close(self.stream)
except OSError: pass
def write_chunk(self, chunk):
# in PY3, the chunk coming in will be bytes, so keep that in mind
if self.handler_type == "fn" and not self.should_quit:
# try to use the encoding first, if that doesn't work, send
# the bytes
try: to_handler = chunk.decode(self.process.call_args["encoding"])
except UnicodeDecodeError: to_handler = chunk
self.should_quit = self.handler(to_handler, *self.handler_args)
elif self.handler_type == "stringio":
self.handler.write(chunk.decode(self.process.call_args["encoding"]))
elif self.handler_type in ("cstringio", "fd"):
self.handler.write(chunk)
if self.pipe_queue:
if logging_enabled: self.log.debug("putting chunk onto pipe: %r", chunk[:30])
self.pipe_queue.put(chunk)
self.buffer.append(chunk)
def read(self):
# if we're PY3, we're reading bytes, otherwise we're reading
# str
try: chunk = os.read(self.stream, self.bufsize)
except OSError as e:
if logging_enabled: self.log.debug("got errno %d, done reading", e.errno)
return True
if not chunk:
if logging_enabled: self.log.debug("got no chunk, done reading")
return True
if logging_enabled: self.log.debug("got chunk size %d: %r", len(chunk), chunk[:30])
for chunk in self.stream_bufferer.process(chunk):
self.write_chunk(chunk)
# this is used for feeding in chunks of stdout/stderr, and breaking it up into
# chunks that will actually be put into the internal buffers. for example, if
# you have two processes, one being piped to the other, and you want that,
# first process to feed lines of data (instead of the chunks however they
# come in), OProc will use an instance of this class to chop up the data and
# feed it as lines to be sent down the pipe
class StreamBufferer(object):
def __init__(self, encoding="utf8", buffer_type=1):
# 0 for unbuffered, 1 for line, everything else for that amount
self.type = buffer_type
self.buffer = []
self.n_buffer_count = 0
self.encoding = encoding
# this is for if we change buffering types. if we change from line
# buffered to unbuffered, its very possible that our self.buffer list
# has data that was being saved up (while we searched for a newline).
# we need to use that up, so we don't lose it
self._use_up_buffer_first = False
# the buffering lock is used because we might chance the buffering
# types from a different thread. for example, if we have a stdout
# callback, we might use it to change the way stdin buffers. so we
# lock
self._buffering_lock = threading.RLock()
self.log = logging.getLogger("stream_bufferer")
def change_buffering(self, new_type):
# TODO, when we stop supporting 2.6, make this a with context
if logging_enabled: self.log.debug("acquiring buffering lock for changing buffering")
self._buffering_lock.acquire()
if logging_enabled: self.log.debug("got buffering lock for changing buffering")
try:
if new_type == 0: self._use_up_buffer_first = True
self.type = new_type
finally:
self._buffering_lock.release()
if logging_enabled: self.log.debug("released buffering lock for changing buffering")
def process(self, chunk):
# MAKE SURE THAT THE INPUT IS PY3 BYTES
# THE OUTPUT IS ALWAYS PY3 BYTES
# TODO, when we stop supporting 2.6, make this a with context
if logging_enabled: self.log.debug("acquiring buffering lock to process chunk (buffering: %d)", self.type)
self._buffering_lock.acquire()
if logging_enabled: self.log.debug("got buffering lock to process chunk (buffering: %d)", self.type)
try:
# we've encountered binary, permanently switch to N size buffering
# since matching on newline doesn't make sense anymore
if self.type == 1:
try: chunk.decode(self.encoding)
except:
if logging_enabled: self.log.debug("detected binary data, changing buffering")
self.change_buffering(1024)
# unbuffered
if self.type == 0:
if self._use_up_buffer_first:
self._use_up_buffer_first = False
to_write = self.buffer
self.buffer = []
to_write.append(chunk)
return to_write
return [chunk]
# line buffered
elif self.type == 1:
total_to_write = []
chunk = chunk.decode(self.encoding)
while True:
newline = chunk.find("\n")
if newline == -1: break
chunk_to_write = chunk[:newline+1]
if self.buffer:
# this is ugly, but it's designed to take the existing
# bytes buffer, join it together, tack on our latest
# chunk, then convert the whole thing to a string.
# it's necessary, i'm sure. read the whole block to
# see why.
chunk_to_write = "".encode(self.encoding).join(self.buffer) \
+ chunk_to_write.encode(self.encoding)
chunk_to_write = chunk_to_write.decode(self.encoding)
self.buffer = []
self.n_buffer_count = 0
chunk = chunk[newline+1:]
total_to_write.append(chunk_to_write.encode(self.encoding))
if chunk:
self.buffer.append(chunk.encode(self.encoding))
self.n_buffer_count += len(chunk)
return total_to_write
# N size buffered
else:
total_to_write = []
while True:
overage = self.n_buffer_count + len(chunk) - self.type
if overage >= 0:
ret = "".encode(self.encoding).join(self.buffer) + chunk
chunk_to_write = ret[:self.type]
chunk = ret[self.type:]
total_to_write.append(chunk_to_write)
self.buffer = []
self.n_buffer_count = 0
else:
self.buffer.append(chunk)
self.n_buffer_count += len(chunk)
break
return total_to_write
finally:
self._buffering_lock.release()
if logging_enabled: self.log.debug("released buffering lock for processing chunk (buffering: %d)", self.type)
def flush(self):
if logging_enabled: self.log.debug("acquiring buffering lock for flushing buffer")
self._buffering_lock.acquire()
if logging_enabled: self.log.debug("got buffering lock for flushing buffer")
try:
ret = "".encode(self.encoding).join(self.buffer)
self.buffer = []
return ret
finally:
self._buffering_lock.release()
if logging_enabled: self.log.debug("released buffering lock for flushing buffer")
# this allows lookups to names that aren't found in the global scope to be
# searched for as a program name. for example, if "ls" isn't found in this
# module's scope, we consider it a system program and try to find it.
class Environment(dict):
def __init__(self, *args, **kwargs):
dict.__init__(self, *args, **kwargs)
self["Command"] = Command
self["CommandNotFound"] = CommandNotFound
self["ErrorReturnCode"] = ErrorReturnCode
self["ARGV"] = sys.argv[1:]
for i, arg in enumerate(sys.argv):
self["ARG%d" % i] = arg
# this needs to be last
self["env"] = os.environ
def __setitem__(self, k, v):
# are we altering an environment variable?
if "env" in self and k in self["env"]: self["env"][k] = v
# no? just setting a regular name
else: dict.__setitem__(self, k, v)
def __missing__(self, k):
# the only way we'd get to here is if we've tried to
# import * from a repl. so, raise an exception, since
# that's really the only sensible thing to do
if k == "__all__":
raise ImportError("Cannot import * from sh. \
Please import sh or import programs individually.")
# if we end with "_" just go ahead and skip searching
# our namespace for python stuff. this was mainly for the
# command "id", which is a popular program for finding
# if a user exists, but also a python function for getting
# the address of an object. so can call the python
# version by "id" and the program version with "id_"
if not k.endswith("_"):
# check if we're naming a dynamically generated ReturnCode exception
try: return rc_exc_cache[k]
except KeyError:
m = rc_exc_regex.match(k)
if m: return get_rc_exc(int(m.group(1)))
# are we naming a commandline argument?
if k.startswith("ARG"):
return None
# is it a builtin?
try: return getattr(self["__builtins__"], k)
except AttributeError: pass
elif not k.startswith("_"): k = k.rstrip("_")
# how about an environment variable?
try: return os.environ[k]
except KeyError: pass
# is it a custom builtin?
builtin = getattr(self, "b_"+k, None)
if builtin: return builtin
# it must be a command then
return Command._create(k)
# methods that begin with "b_" are custom builtins and will override any
# program that exists in our path. this is useful for things like
# common shell builtins that people are used to, but which aren't actually
# full-fledged system binaries
def b_cd(self, path):
os.chdir(path)
def b_which(self, program):
return which(program)
def run_repl(env):
banner = "\n>> sh v{version}\n>> https://github.com/amoffat/sh\n"
print(banner.format(version=__version__))
while True:
try: line = raw_input("sh> ")
except (ValueError, EOFError): break
try: exec(compile(line, "<dummy>", "single"), env, env)
except SystemExit: break
except: print(traceback.format_exc())
# cleans up our last line
print("")
# this is a thin wrapper around THIS module (we patch sys.modules[__name__]).
# this is in the case that the user does a "from sh import whatever"
# in other words, they only want to import certain programs, not the whole
# system PATH worth of commands. in this case, we just proxy the
# import lookup to our Environment class
class SelfWrapper(ModuleType):
def __init__(self, self_module):
# this is super ugly to have to copy attributes like this,
# but it seems to be the only way to make reload() behave
# nicely. if i make these attributes dynamic lookups in
# __getattr__, reload sometimes chokes in weird ways...
for attr in ["__builtins__", "__doc__", "__name__", "__package__"]:
setattr(self, attr, getattr(self_module, attr))
self.self_module = self_module
self.env = Environment(globals())
def __getattr__(self, name):
return self.env[name]
# we're being run as a stand-alone script
if __name__ == "__main__":
try: arg = sys.argv.pop(1)
except: arg = None
if arg == "test":
import subprocess
def run_test(version):
py_version = "python%s" % version
py_bin = which(py_version)
if py_bin:
print("Testing %s" % py_version.capitalize())
p = subprocess.Popen([py_bin, os.path.join(THIS_DIR, "test.py")]
+ sys.argv[1:])
p.wait()
else:
print("Couldn't find %s, skipping" % py_version.capitalize())
versions = ("2.6", "2.7", "3.1", "3.2")
for version in versions: run_test(version)
else:
globs = globals()
f_globals = {}
for k in ["__builtins__", "__doc__", "__name__", "__package__"]:
f_globals[k] = globs[k]
env = Environment(f_globals)
run_repl(env)
# we're being imported from somewhere
else:
self = sys.modules[__name__]
sys.modules[__name__] = SelfWrapper(self)
Jump to Line
Something went wrong with that request. Please try again.