utils.py

# vim: tabstop=4 shiftwidth=4 softtabstop=4

# Copyright 2010 United States Government as represented by the
# Administrator of the National Aeronautics and Space Administration.
# Copyright 2011 Justin Santa Barbara
# All Rights Reserved.
#
#    Licensed under the Apache License, Version 2.0 (the "License"); you may
#    not use this file except in compliance with the License. You may obtain
#    a copy of the License at
#
#         http://www.apache.org/licenses/LICENSE-2.0
#
#    Unless required by applicable law or agreed to in writing, software
#    distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
#    WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
#    License for the specific language governing permissions and limitations
#    under the License.

"""Utilities and helper functions."""

import contextlib
import datetime
import functools
import inspect
import json
import lockfile
import os
import pyclbr
import random
import re
import shlex
import socket
import struct
import sys
import time
import types
import uuid
import warnings
from xml.sax import saxutils

from eventlet import event
from eventlet import greenthread
from eventlet import semaphore
from eventlet.green import subprocess
import netaddr

from nova import exception
from nova import flags
from nova import log as logging


LOG = logging.getLogger("nova.utils")
ISO_TIME_FORMAT = "%Y-%m-%dT%H:%M:%SZ"
PERFECT_TIME_FORMAT = "%Y-%m-%dT%H:%M:%S.%f"
FLAGS = flags.FLAGS


def import_class(import_str):
    """Returns a class from a string including module and class."""
    mod_str, _sep, class_str = import_str.rpartition('.')
    try:
        __import__(mod_str)
        return getattr(sys.modules[mod_str], class_str)
    except (ImportError, ValueError, AttributeError), exc:
        LOG.debug(_('Inner Exception: %s'), exc)
        raise exception.ClassNotFound(class_name=class_str, exception=exc)


def import_object(import_str):
    """Returns an object including a module or module and class."""
    try:
        __import__(import_str)
        return sys.modules[import_str]
    except ImportError:
        cls = import_class(import_str)
        return cls()


def find_config(config_path):
    """Find a configuration file using the given hint.

    :param config_path: Full or relative path to the config.
    :returns: Full path of the config, if it exists.
    :raises: `nova.exception.ConfigNotFound`

    """
    possible_locations = [
        config_path,
        os.path.join(FLAGS.state_path, "etc", "nova", config_path),
        os.path.join(FLAGS.state_path, "etc", config_path),
        os.path.join(FLAGS.state_path, config_path),
        "/etc/nova/%s" % config_path,
    ]

    for path in possible_locations:
        if os.path.exists(path):
            return os.path.abspath(path)

    raise exception.ConfigNotFound(path=os.path.abspath(config_path))


def vpn_ping(address, port, timeout=0.05, session_id=None):
    """Sends a vpn negotiation packet and returns the server session.

    Returns False on a failure. Basic packet structure is below.

    Client packet (14 bytes)::
     0 1      8 9  13
    +-+--------+-----+
    |x| cli_id |?????|
    +-+--------+-----+
    x = packet identifier 0x38
    cli_id = 64 bit identifier
    ? = unknown, probably flags/padding

    Server packet (26 bytes)::
     0 1      8 9  13 14    21 2225
    +-+--------+-----+--------+----+
    |x| srv_id |?????| cli_id |????|
    +-+--------+-----+--------+----+
    x = packet identifier 0x40
    cli_id = 64 bit identifier
    ? = unknown, probably flags/padding
    bit 9 was 1 and the rest were 0 in testing

    """
    if session_id is None:
        session_id = random.randint(0, 0xffffffffffffffff)
    sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
    data = struct.pack('!BQxxxxxx', 0x38, session_id)
    sock.sendto(data, (address, port))
    sock.settimeout(timeout)
    try:
        received = sock.recv(2048)
    except socket.timeout:
        return False
    finally:
        sock.close()
    fmt = '!BQxxxxxQxxxx'
    if len(received) != struct.calcsize(fmt):
        print struct.calcsize(fmt)
        return False
    (identifier, server_sess, client_sess) = struct.unpack(fmt, received)
    if identifier == 0x40 and client_sess == session_id:
        return server_sess


def fetchfile(url, target):
    LOG.debug(_('Fetching %s') % url)
    execute('curl', '--fail', url, '-o', target)


def execute(*cmd, **kwargs):
    """
    Helper method to execute command with optional retry.

    :cmd                Passed to subprocess.Popen.
    :process_input      Send to opened process.
    :check_exit_code    Single bool, int, or list of allowed exit codes.
                        Defaults to [0].  Raise exception.ProcessExecutionError
                        unless program exits with one of these code.
    :delay_on_retry     True | False. Defaults to True. If set to True, wait a
                        short amount of time before retrying.
    :attempts           How many times to retry cmd.
    :run_as_root        True | False. Defaults to False. If set to True,
                        the command is prefixed by the command specified
                        in the root_helper FLAG.

    :raises exception.Error on receiving unknown arguments
    :raises exception.ProcessExecutionError

    :returns a tuple, (stdout, stderr) from the spawned process, or None if
             the command fails.
    """

    process_input = kwargs.pop('process_input', None)
    check_exit_code = kwargs.pop('check_exit_code', [0])
    ignore_exit_code = False
    if isinstance(check_exit_code, bool):
        ignore_exit_code = not check_exit_code
        check_exit_code = [0]
    elif isinstance(check_exit_code, int):
        check_exit_code = [check_exit_code]
    delay_on_retry = kwargs.pop('delay_on_retry', True)
    attempts = kwargs.pop('attempts', 1)
    run_as_root = kwargs.pop('run_as_root', False)
    shell = kwargs.pop('shell', False)

    if len(kwargs):
        raise exception.Error(_('Got unknown keyword args '
                                'to utils.execute: %r') % kwargs)

    if run_as_root:
        cmd = shlex.split(FLAGS.root_helper) + list(cmd)
    cmd = map(str, cmd)

    while attempts > 0:
        attempts -= 1
        try:
            LOG.debug(_('Running cmd (subprocess): %s'), ' '.join(cmd))
            _PIPE = subprocess.PIPE  # pylint: disable=E1101
            obj = subprocess.Popen(cmd,
                                   stdin=_PIPE,
                                   stdout=_PIPE,
                                   stderr=_PIPE,
                                   close_fds=True,
                                   shell=shell)
            result = None
            if process_input is not None:
                result = obj.communicate(process_input)
            else:
                result = obj.communicate()
            obj.stdin.close()  # pylint: disable=E1101
            _returncode = obj.returncode  # pylint: disable=E1101
            if _returncode:
                LOG.debug(_('Result was %s') % _returncode)
                if not ignore_exit_code \
                    and _returncode not in check_exit_code:
                    (stdout, stderr) = result
                    raise exception.ProcessExecutionError(
                            exit_code=_returncode,
                            stdout=stdout,
                            stderr=stderr,
                            cmd=' '.join(cmd))
            return result
        except exception.ProcessExecutionError:
            if not attempts:
                raise
            else:
                LOG.debug(_('%r failed. Retrying.'), cmd)
                if delay_on_retry:
                    greenthread.sleep(random.randint(20, 200) / 100.0)
        finally:
            # NOTE(termie): this appears to be necessary to let the subprocess
            #               call clean something up in between calls, without
            #               it two execute calls in a row hangs the second one
            greenthread.sleep(0)


def trycmd(*args, **kwargs):
    """
    A wrapper around execute() to more easily handle warnings and errors.

    Returns an (out, err) tuple of strings containing the output of
    the command's stdout and stderr.  If 'err' is not empty then the
    command can be considered to have failed.

    :discard_warnings   True | False. Defaults to False. If set to True,
                        then for succeeding commands, stderr is cleared

    """
    discard_warnings = kwargs.pop('discard_warnings', False)

    try:
        out, err = execute(*args, **kwargs)
        failed = False
    except exception.ProcessExecutionError, exn:
        out, err = '', str(exn)
        LOG.debug(err)
        failed = True

    if not failed and discard_warnings and err:
        # Handle commands that output to stderr but otherwise succeed
        LOG.debug(err)
        err = ''

    return out, err


def ssh_execute(ssh, cmd, process_input=None,
                addl_env=None, check_exit_code=True):
    LOG.debug(_('Running cmd (SSH): %s'), ' '.join(cmd))
    if addl_env:
        raise exception.Error(_('Environment not supported over SSH'))

    if process_input:
        # This is (probably) fixable if we need it...
        raise exception.Error(_('process_input not supported over SSH'))

    stdin_stream, stdout_stream, stderr_stream = ssh.exec_command(cmd)
    channel = stdout_stream.channel

    #stdin.write('process_input would go here')
    #stdin.flush()

    # NOTE(justinsb): This seems suspicious...
    # ...other SSH clients have buffering issues with this approach
    stdout = stdout_stream.read()
    stderr = stderr_stream.read()
    stdin_stream.close()

    exit_status = channel.recv_exit_status()

    # exit_status == -1 if no exit code was returned
    if exit_status != -1:
        LOG.debug(_('Result was %s') % exit_status)
        if check_exit_code and exit_status != 0:
            raise exception.ProcessExecutionError(exit_code=exit_status,
                                                  stdout=stdout,
                                                  stderr=stderr,
                                                  cmd=' '.join(cmd))

    return (stdout, stderr)


def abspath(s):
    return os.path.join(os.path.dirname(__file__), s)


def novadir():
    import nova
    return os.path.abspath(nova.__file__).split('nova/__init__.py')[0]


def default_flagfile(filename='nova.conf', args=None):
    if args is None:
        args = sys.argv
    for arg in args:
        if arg.find('flagfile') != -1:
            break
    else:
        if not os.path.isabs(filename):
            # turn relative filename into an absolute path
            script_dir = os.path.dirname(inspect.stack()[-1][1])
            filename = os.path.abspath(os.path.join(script_dir, filename))
        if not os.path.exists(filename):
            filename = "./nova.conf"
            if not os.path.exists(filename):
                filename = '/etc/nova/nova.conf'
        if os.path.exists(filename):
            flagfile = '--flagfile=%s' % filename
            args.insert(1, flagfile)


def debug(arg):
    LOG.debug(_('debug in callback: %s'), arg)
    return arg


def runthis(prompt, *cmd, **kwargs):
    LOG.debug(_('Running %s'), (' '.join(cmd)))
    rv, err = execute(*cmd, **kwargs)


def generate_uid(topic, size=8):
    characters = '01234567890abcdefghijklmnopqrstuvwxyz'
    choices = [random.choice(characters) for x in xrange(size)]
    return '%s-%s' % (topic, ''.join(choices))


# Default symbols to use for passwords. Avoids visually confusing characters.
# ~6 bits per symbol
DEFAULT_PASSWORD_SYMBOLS = ('23456789',  # Removed: 0,1
                            'ABCDEFGHJKLMNPQRSTUVWXYZ',   # Removed: I, O
                            'abcdefghijkmnopqrstuvwxyz')  # Removed: l


# ~5 bits per symbol
EASIER_PASSWORD_SYMBOLS = ('23456789',  # Removed: 0, 1
                           'ABCDEFGHJKLMNPQRSTUVWXYZ')  # Removed: I, O


def current_audit_period(unit=None):
    if not unit:
        unit = FLAGS.instance_usage_audit_period
    rightnow = utcnow()
    if unit not in ('month', 'day', 'year', 'hour'):
        raise ValueError('Time period must be hour, day, month or year')
    n = 1  # we are currently only using multiples of 1 unit (mdragon)
    if unit == 'month':
        year = rightnow.year - (n // 12)
        n = n % 12
        if n >= rightnow.month:
            year -= 1
            month = 12 + (rightnow.month - n)
        else:
            month = rightnow.month - n
        begin = datetime.datetime(day=1, month=month, year=year)
        end = datetime.datetime(day=1,
                                month=rightnow.month,
                                year=rightnow.year)

    elif unit == 'year':
        begin = datetime.datetime(day=1, month=1, year=rightnow.year - n)
        end = datetime.datetime(day=1, month=1, year=rightnow.year)

    elif unit == 'day':
        b = rightnow - datetime.timedelta(days=n)
        begin = datetime.datetime(day=b.day, month=b.month, year=b.year)
        end = datetime.datetime(day=rightnow.day,
                               month=rightnow.month,
                               year=rightnow.year)
    elif unit == 'hour':
        end = rightnow.replace(minute=0, second=0, microsecond=0)
        begin = end - datetime.timedelta(hours=n)

    return (begin, end)


def usage_from_instance(instance_ref, network_info=None, **kw):
    image_ref_url = "%s/images/%s" % (generate_glance_url(),
            instance_ref['image_ref'])

    usage_info = dict(
          tenant_id=instance_ref['project_id'],
          user_id=instance_ref['user_id'],
          instance_id=instance_ref['uuid'],
          instance_type=instance_ref['instance_type']['name'],
          instance_type_id=instance_ref['instance_type_id'],
          memory_mb=instance_ref['memory_mb'],
          disk_gb=instance_ref['root_gb'] + instance_ref['ephemeral_gb'],
          display_name=instance_ref['display_name'],
          created_at=str(instance_ref['created_at']),
          launched_at=str(instance_ref['launched_at']) \
                      if instance_ref['launched_at'] else '',
          image_ref_url=image_ref_url,
          state=instance_ref['vm_state'],
          state_description=instance_ref['task_state'] \
                             if instance_ref['task_state'] else '')

    # NOTE(jkoelker) This nastyness can go away once compute uses the
    #                network model
    if network_info is not None:
        fixed_ips = []
        for network, info in network_info:
            fixed_ips.extend([ip['ip'] for ip in info['ips']])
        usage_info['fixed_ips'] = fixed_ips

    usage_info.update(kw)
    return usage_info


def generate_password(length=20, symbolgroups=DEFAULT_PASSWORD_SYMBOLS):
    """Generate a random password from the supplied symbol groups.

    At least one symbol from each group will be included. Unpredictable
    results if length is less than the number of symbol groups.

    Believed to be reasonably secure (with a reasonable password length!)

    """
    r = random.SystemRandom()

    # NOTE(jerdfelt): Some password policies require at least one character
    # from each group of symbols, so start off with one random character
    # from each symbol group
    password = [r.choice(s) for s in symbolgroups]
    # If length < len(symbolgroups), the leading characters will only
    # be from the first length groups. Try our best to not be predictable
    # by shuffling and then truncating.
    r.shuffle(password)
    password = password[:length]
    length -= len(password)

    # then fill with random characters from all symbol groups
    symbols = ''.join(symbolgroups)
    password.extend([r.choice(symbols) for _i in xrange(length)])

    # finally shuffle to ensure first x characters aren't from a
    # predictable group
    r.shuffle(password)

    return ''.join(password)


def last_octet(address):
    return int(address.split('.')[-1])


def get_my_linklocal(interface):
    try:
        if_str = execute('ip', '-f', 'inet6', '-o', 'addr', 'show', interface)
        condition = '\s+inet6\s+([0-9a-f:]+)/\d+\s+scope\s+link'
        links = [re.search(condition, x) for x in if_str[0].split('\n')]
        address = [w.group(1) for w in links if w is not None]
        if address[0] is not None:
            return address[0]
        else:
            raise exception.Error(_('Link Local address is not found.:%s')
                                  % if_str)
    except Exception as ex:
        raise exception.Error(_("Couldn't get Link Local IP of %(interface)s"
                                " :%(ex)s") % locals())


def utcnow():
    """Overridable version of utils.utcnow."""
    if utcnow.override_time:
        return utcnow.override_time
    return datetime.datetime.utcnow()


utcnow.override_time = None


def is_older_than(before, seconds):
    """Return True if before is older than seconds."""
    return utcnow() - before > datetime.timedelta(seconds=seconds)


def utcnow_ts():
    """Timestamp version of our utcnow function."""
    return time.mktime(utcnow().timetuple())


def set_time_override(override_time=datetime.datetime.utcnow()):
    """Override utils.utcnow to return a constant time."""
    utcnow.override_time = override_time


def advance_time_delta(timedelta):
    """Advance overriden time using a datetime.timedelta."""
    assert(not utcnow.override_time is None)
    utcnow.override_time += timedelta


def advance_time_seconds(seconds):
    """Advance overriden time by seconds."""
    advance_time_delta(datetime.timedelta(0, seconds))


def clear_time_override():
    """Remove the overridden time."""
    utcnow.override_time = None


def strtime(at=None, fmt=PERFECT_TIME_FORMAT):
    """Returns formatted utcnow."""
    if not at:
        at = utcnow()
    return at.strftime(fmt)


def parse_strtime(timestr, fmt=PERFECT_TIME_FORMAT):
    """Turn a formatted time back into a datetime."""
    return datetime.datetime.strptime(timestr, fmt)


def isotime(at=None):
    """Returns iso formatted utcnow."""
    return strtime(at, ISO_TIME_FORMAT)


def parse_isotime(timestr):
    """Turn an iso formatted time back into a datetime."""
    return parse_strtime(timestr, ISO_TIME_FORMAT)


def parse_mailmap(mailmap='.mailmap'):
    mapping = {}
    if os.path.exists(mailmap):
        fp = open(mailmap, 'r')
        for l in fp:
            l = l.strip()
            if not l.startswith('#') and ' ' in l:
                canonical_email, alias = l.split(' ')
                mapping[alias.lower()] = canonical_email.lower()
    return mapping


def str_dict_replace(s, mapping):
    for s1, s2 in mapping.iteritems():
        s = s.replace(s1, s2)
    return s


class LazyPluggable(object):
    """A pluggable backend loaded lazily based on some value."""

    def __init__(self, pivot, **backends):
        self.__backends = backends
        self.__pivot = pivot
        self.__backend = None

    def __get_backend(self):
        if not self.__backend:
            backend_name = self.__pivot.value
            if backend_name not in self.__backends:
                raise exception.Error(_('Invalid backend: %s') % backend_name)

            backend = self.__backends[backend_name]
            if isinstance(backend, tuple):
                name = backend[0]
                fromlist = backend[1]
            else:
                name = backend
                fromlist = backend

            self.__backend = __import__(name, None, None, fromlist)
            LOG.debug(_('backend %s'), self.__backend)
        return self.__backend

    def __getattr__(self, key):
        backend = self.__get_backend()
        return getattr(backend, key)


class LoopingCallDone(Exception):
    """Exception to break out and stop a LoopingCall.

    The poll-function passed to LoopingCall can raise this exception to
    break out of the loop normally. This is somewhat analogous to
    StopIteration.

    An optional return-value can be included as the argument to the exception;
    this return-value will be returned by LoopingCall.wait()

    """

    def __init__(self, retvalue=True):
        """:param retvalue: Value that LoopingCall.wait() should return."""
        self.retvalue = retvalue


class LoopingCall(object):
    def __init__(self, f=None, *args, **kw):
        self.args = args
        self.kw = kw
        self.f = f
        self._running = False

    def start(self, interval, now=True):
        self._running = True
        done = event.Event()

        def _inner():
            if not now:
                greenthread.sleep(interval)
            try:
                while self._running:
                    self.f(*self.args, **self.kw)
                    if not self._running:
                        break
                    greenthread.sleep(interval)
            except LoopingCallDone, e:
                self.stop()
                done.send(e.retvalue)
            except Exception:
                logging.exception('in looping call')
                done.send_exception(*sys.exc_info())
                return
            else:
                done.send(True)

        self.done = done

        greenthread.spawn(_inner)
        return self.done

    def stop(self):
        self._running = False

    def wait(self):
        return self.done.wait()


def xhtml_escape(value):
    """Escapes a string so it is valid within XML or XHTML.

    Code is directly from the utf8 function in
    http://github.com/facebook/tornado/blob/master/tornado/escape.py

    """
    return saxutils.escape(value, {'"': '&quot;'})


def utf8(value):
    """Try to turn a string into utf-8 if possible.

    Code is directly from the utf8 function in
    http://github.com/facebook/tornado/blob/master/tornado/escape.py

    """
    if isinstance(value, unicode):
        return value.encode('utf-8')
    assert isinstance(value, str)
    return value


def to_primitive(value, convert_instances=False, level=0):
    """Convert a complex object into primitives.

    Handy for JSON serialization. We can optionally handle instances,
    but since this is a recursive function, we could have cyclical
    data structures.

    To handle cyclical data structures we could track the actual objects
    visited in a set, but not all objects are hashable. Instead we just
    track the depth of the object inspections and don't go too deep.

    Therefore, convert_instances=True is lossy ... be aware.

    """
    nasty = [inspect.ismodule, inspect.isclass, inspect.ismethod,
             inspect.isfunction, inspect.isgeneratorfunction,
             inspect.isgenerator, inspect.istraceback, inspect.isframe,
             inspect.iscode, inspect.isbuiltin, inspect.isroutine,
             inspect.isabstract]
    for test in nasty:
        if test(value):
            return unicode(value)

    # FIXME(vish): Workaround for LP bug 852095. Without this workaround,
    #              tests that raise an exception in a mocked method that
    #              has a @wrap_exception with a notifier will fail. If
    #              we up the dependency to 0.5.4 (when it is released) we
    #              can remove this workaround.
    if getattr(value, '__module__', None) == 'mox':
        return 'mock'

    if level > 3:
        return '?'

    # The try block may not be necessary after the class check above,
    # but just in case ...
    try:
        if isinstance(value, (list, tuple)):
            o = []
            for v in value:
                o.append(to_primitive(v, convert_instances=convert_instances,
                                      level=level))
            return o
        elif isinstance(value, dict):
            o = {}
            for k, v in value.iteritems():
                o[k] = to_primitive(v, convert_instances=convert_instances,
                                    level=level)
            return o
        elif isinstance(value, datetime.datetime):
            return str(value)
        elif hasattr(value, 'iteritems'):
            return to_primitive(dict(value.iteritems()),
                                convert_instances=convert_instances,
                                level=level)
        elif hasattr(value, '__iter__'):
            return to_primitive(list(value), level)
        elif convert_instances and hasattr(value, '__dict__'):
            # Likely an instance of something. Watch for cycles.
            # Ignore class member vars.
            return to_primitive(value.__dict__,
                                convert_instances=convert_instances,
                                level=level + 1)
        else:
            return value
    except TypeError, e:
        # Class objects are tricky since they may define something like
        # __iter__ defined but it isn't callable as list().
        return unicode(value)


def dumps(value):
    try:
        return json.dumps(value)
    except TypeError:
        pass
    return json.dumps(to_primitive(value))


def loads(s):
    return json.loads(s)


try:
    import anyjson
except ImportError:
    pass
else:
    anyjson._modules.append(("nova.utils", "dumps", TypeError,
                                           "loads", ValueError))
    anyjson.force_implementation("nova.utils")


_semaphores = {}


class _NoopContextManager(object):
    def __enter__(self):
        pass

    def __exit__(self, exc_type, exc_val, exc_tb):
        pass


def synchronized(name, external=False):
    """Synchronization decorator.

    Decorating a method like so:
    @synchronized('mylock')
    def foo(self, *args):
       ...

    ensures that only one thread will execute the bar method at a time.

    Different methods can share the same lock:
    @synchronized('mylock')
    def foo(self, *args):
       ...

    @synchronized('mylock')
    def bar(self, *args):
       ...

    This way only one of either foo or bar can be executing at a time.

    The external keyword argument denotes whether this lock should work across
    multiple processes. This means that if two different workers both run a
    a method decorated with @synchronized('mylock', external=True), only one
    of them will execute at a time.

    """

    def wrap(f):
        @functools.wraps(f)
        def inner(*args, **kwargs):
            # NOTE(soren): If we ever go natively threaded, this will be racy.
            #              See http://stackoverflow.com/questions/5390569/dyn\
            #              amically-allocating-and-destroying-mutexes
            if name not in _semaphores:
                _semaphores[name] = semaphore.Semaphore()
            sem = _semaphores[name]
            LOG.debug(_('Attempting to grab semaphore "%(lock)s" for method '
                        '"%(method)s"...' % {'lock': name,
                                             'method': f.__name__}))
            with sem:
                LOG.debug(_('Got semaphore "%(lock)s" for method '
                            '"%(method)s"...' % {'lock': name,
                                                 'method': f.__name__}))
                if external:
                    LOG.debug(_('Attempting to grab file lock "%(lock)s" for '
                                'method "%(method)s"...' %
                                {'lock': name, 'method': f.__name__}))
                    lock_file_path = os.path.join(FLAGS.lock_path,
                                                  'nova-%s.lock' % name)
                    lock = lockfile.FileLock(lock_file_path)
                else:
                    lock = _NoopContextManager()

                with lock:
                    if external:
                        LOG.debug(_('Got file lock "%(lock)s" for '
                                    'method "%(method)s"...' %
                                    {'lock': name, 'method': f.__name__}))
                    retval = f(*args, **kwargs)

            # If no-one else is waiting for it, delete it.
            # See note about possible raciness above.
            if not sem.balance < 1:
                del _semaphores[name]

            return retval
        return inner
    return wrap


def get_from_path(items, path):
    """Returns a list of items matching the specified path.

    Takes an XPath-like expression e.g. prop1/prop2/prop3, and for each item
    in items, looks up items[prop1][prop2][prop3].  Like XPath, if any of the
    intermediate results are lists it will treat each list item individually.
    A 'None' in items or any child expressions will be ignored, this function
    will not throw because of None (anywhere) in items.  The returned list
    will contain no None values.

    """
    if path is None:
        raise exception.Error('Invalid mini_xpath')

    (first_token, sep, remainder) = path.partition('/')

    if first_token == '':
        raise exception.Error('Invalid mini_xpath')

    results = []

    if items is None:
        return results

    if not isinstance(items, list):
        # Wrap single objects in a list
        items = [items]

    for item in items:
        if item is None:
            continue
        get_method = getattr(item, 'get', None)
        if get_method is None:
            continue
        child = get_method(first_token)
        if child is None:
            continue
        if isinstance(child, list):
            # Flatten intermediate lists
            for x in child:
                results.append(x)
        else:
            results.append(child)

    if not sep:
        # No more tokens
        return results
    else:
        return get_from_path(results, remainder)


def flatten_dict(dict_, flattened=None):
    """Recursively flatten a nested dictionary."""
    flattened = flattened or {}
    for key, value in dict_.iteritems():
        if hasattr(value, 'iteritems'):
            flatten_dict(value, flattened)
        else:
            flattened[key] = value
    return flattened


def partition_dict(dict_, keys):
    """Return two dicts, one with `keys` the other with everything else."""
    intersection = {}
    difference = {}
    for key, value in dict_.iteritems():
        if key in keys:
            intersection[key] = value
        else:
            difference[key] = value
    return intersection, difference


def map_dict_keys(dict_, key_map):
    """Return a dict in which the dictionaries keys are mapped to new keys."""
    mapped = {}
    for key, value in dict_.iteritems():
        mapped_key = key_map[key] if key in key_map else key
        mapped[mapped_key] = value
    return mapped


def subset_dict(dict_, keys):
    """Return a dict that only contains a subset of keys."""
    subset = partition_dict(dict_, keys)[0]
    return subset


def check_isinstance(obj, cls):
    """Checks that obj is of type cls, and lets PyLint infer types."""
    if isinstance(obj, cls):
        return obj
    raise Exception(_('Expected object of type: %s') % (str(cls)))
    # TODO(justinsb): Can we make this better??
    return cls()  # Ugly PyLint hack


def parse_server_string(server_str):
    """
    Parses the given server_string and returns a list of host and port.
    If it's not a combination of host part and port, the port element
    is a null string. If the input is invalid expression, return a null
    list.
    """
    try:
        # First of all, exclude pure IPv6 address (w/o port).
        if netaddr.valid_ipv6(server_str):
            return (server_str, '')

        # Next, check if this is IPv6 address with a port number combination.
        if server_str.find("]:") != -1:
            (address, port) = server_str.replace('[', '', 1).split(']:')
            return (address, port)

        # Third, check if this is a combination of an address and a port
        if server_str.find(':') == -1:
            return (server_str, '')

        # This must be a combination of an address and a port
        (address, port) = server_str.split(':')
        return (address, port)

    except Exception:
        LOG.debug(_('Invalid server_string: %s' % server_str))
        return ('', '')


def gen_uuid():
    return uuid.uuid4()


def is_uuid_like(val):
    """For our purposes, a UUID is a string in canonical form:

        aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa
    """
    try:
        uuid.UUID(val)
        return True
    except (TypeError, ValueError, AttributeError):
        return False


def bool_from_str(val):
    """Convert a string representation of a bool into a bool value"""

    if not val:
        return False
    try:
        return True if int(val) else False
    except ValueError:
        return val.lower() == 'true'


def is_valid_ipv4(address):
    """valid the address strictly as per format xxx.xxx.xxx.xxx.
    where xxx is a value between 0 and 255.
    """
    parts = address.split(".")
    if len(parts) != 4:
        return False
    for item in parts:
        try:
            if not 0 <= int(item) <= 255:
                return False
        except ValueError:
            return False
    return True


def is_valid_cidr(address):
    """Check if the provided ipv4 or ipv6 address is a valid
    CIDR address or not"""
    try:
        # Validate the correct CIDR Address
        netaddr.IPNetwork(address)
    except netaddr.core.AddrFormatError:
        return False

    # Prior validation partially verify /xx part
    # Verify it here
    ip_segment = address.split('/')

    if (len(ip_segment) <= 1 or
        ip_segment[1] == ''):
        return False

    return True


def monkey_patch():
    """  If the Flags.monkey_patch set as True,
    this function patches a decorator
    for all functions in specified modules.
    You can set decorators for each modules
    using FLAGS.monkey_patch_modules.
    The format is "Module path:Decorator function".
    Example: 'nova.api.ec2.cloud:nova.notifier.api.notify_decorator'

    Parameters of the decorator is as follows.
    (See nova.notifier.api.notify_decorator)

    name - name of the function
    function - object of the function
    """
    # If FLAGS.monkey_patch is not True, this function do nothing.
    if not FLAGS.monkey_patch:
        return
    # Get list of modules and decorators
    for module_and_decorator in FLAGS.monkey_patch_modules:
        module, decorator_name = module_and_decorator.split(':')
        # import decorator function
        decorator = import_class(decorator_name)
        __import__(module)
        # Retrieve module information using pyclbr
        module_data = pyclbr.readmodule_ex(module)
        for key in module_data.keys():
            # set the decorator for the class methods
            if isinstance(module_data[key], pyclbr.Class):
                clz = import_class("%s.%s" % (module, key))
                for method, func in inspect.getmembers(clz, inspect.ismethod):
                    setattr(clz, method,
                        decorator("%s.%s.%s" % (module, key, method), func))
            # set the decorator for the function
            if isinstance(module_data[key], pyclbr.Function):
                func = import_class("%s.%s" % (module, key))
                setattr(sys.modules[module], key,
                    decorator("%s.%s" % (module, key), func))


def convert_to_list_dict(lst, label):
    """Convert a value or list into a list of dicts"""
    if not lst:
        return None
    if not isinstance(lst, list):
        lst = [lst]
    return [{label: x} for x in lst]


def timefunc(func):
    """Decorator that logs how long a particular function took to execute"""
    @functools.wraps(func)
    def inner(*args, **kwargs):
        start_time = time.time()
        try:
            return func(*args, **kwargs)
        finally:
            total_time = time.time() - start_time
            LOG.debug(_("timefunc: '%(name)s' took %(total_time).2f secs") %
                      dict(name=func.__name__, total_time=total_time))
    return inner


def generate_glance_url():
    """Generate the URL to glance."""
    # TODO(jk0): This will eventually need to take SSL into consideration
    # when supported in glance.
    return "http://%s:%d" % (FLAGS.glance_host, FLAGS.glance_port)


@contextlib.contextmanager
def save_and_reraise_exception():
    """Save current exception, run some code and then re-raise.

    In some cases the exception context can be cleared, resulting in None
    being attempted to be reraised after an exception handler is run. This
    can happen when eventlet switches greenthreads or when running an
    exception handler, code raises and catches and exception. In both
    cases the exception context will be cleared.

    To work around this, we save the exception state, run handler code, and
    then re-raise the original exception. If another exception occurs, the
    saved exception is logged and the new exception is reraised.
    """
    type_, value, traceback = sys.exc_info()
    try:
        yield
    except Exception:
        LOG.exception(_('Original exception being dropped'),
                      exc_info=(type_, value, traceback))
        raise
    raise type_, value, traceback


@contextlib.contextmanager
def logging_error(message):
    """Catches exception, write message to the log, re-raise.
    This is a common refinement of save_and_reraise that writes a specific
    message to the log.
    """
    try:
        yield
    except Exception as error:
        with save_and_reraise_exception():
            LOG.exception(message)


def make_dev_path(dev, partition=None, base='/dev'):
    """Return a path to a particular device.

    >>> make_dev_path('xvdc')
    /dev/xvdc

    >>> make_dev_path('xvdc', 1)
    /dev/xvdc1
    """
    path = os.path.join(base, dev)
    if partition:
        path += str(partition)
    return path


def total_seconds(td):
    """Local total_seconds implementation for compatibility with python 2.6"""
    if hasattr(td, 'total_seconds'):
        return td.total_seconds()
    else:
        return ((td.days * 86400 + td.seconds) * 10 ** 6 +
                td.microseconds) / 10.0 ** 6


def sanitize_hostname(hostname):
    """Return a hostname which conforms to RFC-952 and RFC-1123 specs."""
    if isinstance(hostname, unicode):
        hostname = hostname.encode('latin-1', 'ignore')

    hostname = re.sub('[ _]', '-', hostname)
    hostname = re.sub('[^\w.-]+', '', hostname)
    hostname = hostname.lower()
    hostname = hostname.strip('.-')

    return hostname


def read_cached_file(filename, cache_info, reload_func=None):
    """Read from a file if it has been modified.

    :param cache_info: dictionary to hold opaque cache.
    :param reload_func: optional function to be called with data when
                        file is reloaded due to a modification.

    :returns: data from file

    """
    mtime = os.path.getmtime(filename)
    if not cache_info or mtime != cache_info.get('mtime'):
        with open(filename) as fap:
            cache_info['data'] = fap.read()
        cache_info['mtime'] = mtime
        if reload_func:
            reload_func(cache_info['data'])
    return cache_info['data']


@contextlib.contextmanager
def temporary_mutation(obj, **kwargs):
    """Temporarily set the attr on a particular object to a given value then
    revert when finished.

    One use of this is to temporarily set the read_deleted flag on a context
    object:

        with temporary_mutation(context, read_deleted="yes"):
            do_something_that_needed_deleted_objects()
    """
    NOT_PRESENT = object()

    old_values = {}
    for attr, new_value in kwargs.items():
        old_values[attr] = getattr(obj, attr, NOT_PRESENT)
        setattr(obj, attr, new_value)

    try:
        yield
    finally:
        for attr, old_value in old_values.items():
            if old_value is NOT_PRESENT:
                del obj[attr]
            else:
                setattr(obj, attr, old_value)


def warn_deprecated_class(cls, msg):
    """
    Issues a warning to indicate that the given class is deprecated.
    If a message is given, it is appended to the deprecation warning.
    """

    fullname = '%s.%s' % (cls.__module__, cls.__name__)
    if msg:
        fullmsg = _("Class %(fullname)s is deprecated: %(msg)s")
    else:
        fullmsg = _("Class %(fullname)s is deprecated")

    # Issue the warning
    warnings.warn(fullmsg % locals(), DeprecationWarning, stacklevel=3)


def warn_deprecated_function(func, msg):
    """
    Issues a warning to indicate that the given function is
    deprecated.  If a message is given, it is appended to the
    deprecation warning.
    """

    name = func.__name__

    # Find the function's definition
    sourcefile = inspect.getsourcefile(func)

    # Find the line number, if possible
    if inspect.ismethod(func):
        code = func.im_func.func_code
    else:
        code = func.func_code
    lineno = getattr(code, 'co_firstlineno', None)

    if lineno is None:
        location = sourcefile
    else:
        location = "%s:%d" % (sourcefile, lineno)

    # Build up the message
    if msg:
        fullmsg = _("Function %(name)s in %(location)s is deprecated: %(msg)s")
    else:
        fullmsg = _("Function %(name)s in %(location)s is deprecated")

    # Issue the warning
    warnings.warn(fullmsg % locals(), DeprecationWarning, stacklevel=3)


def _stubout(klass, message):
    """
    Scans a class and generates wrapping stubs for __new__() and every
    class and static method.  Returns a dictionary which can be passed
    to type() to generate a wrapping class.
    """

    overrides = {}

    def makestub_class(name, func):
        """
        Create a stub for wrapping class methods.
        """

        def stub(cls, *args, **kwargs):
            warn_deprecated_class(klass, message)
            return func(*args, **kwargs)

        # Overwrite the stub's name
        stub.__name__ = name
        stub.func_name = name

        return classmethod(stub)

    def makestub_static(name, func):
        """
        Create a stub for wrapping static methods.
        """

        def stub(*args, **kwargs):
            warn_deprecated_class(klass, message)
            return func(*args, **kwargs)

        # Overwrite the stub's name
        stub.__name__ = name
        stub.func_name = name

        return staticmethod(stub)

    for name, kind, _klass, _obj in inspect.classify_class_attrs(klass):
        # We're only interested in __new__(), class methods, and
        # static methods...
        if (name != '__new__' and
            kind not in ('class method', 'static method')):
            continue

        # Get the function...
        func = getattr(klass, name)

        # Override it in the class
        if kind == 'class method':
            stub = makestub_class(name, func)
        elif kind == 'static method' or name == '__new__':
            stub = makestub_static(name, func)

        # Save it in the overrides dictionary...
        overrides[name] = stub

    # Apply the overrides
    for name, stub in overrides.items():
        setattr(klass, name, stub)


def deprecated(message=''):
    """
    Marks a function, class, or method as being deprecated.  For
    functions and methods, emits a warning each time the function or
    method is called.  For classes, generates a new subclass which
    will emit a warning each time the class is instantiated, or each
    time any class or static method is called.

    If a message is passed to the decorator, that message will be
    appended to the emitted warning.  This may be used to suggest an
    alternate way of achieving the desired effect, or to explain why
    the function, class, or method is deprecated.
    """

    def decorator(f_or_c):
        # Make sure we can deprecate it...
        if not callable(f_or_c) or isinstance(f_or_c, types.ClassType):
            warnings.warn("Cannot mark object %r as deprecated" % f_or_c,
                          DeprecationWarning, stacklevel=2)
            return f_or_c

        # If we're deprecating a class, create a subclass of it and
        # stub out all the class and static methods
        if inspect.isclass(f_or_c):
            klass = f_or_c
            _stubout(klass, message)
            return klass

        # OK, it's a function; use a traditional wrapper...
        func = f_or_c

        @functools.wraps(func)
        def wrapper(*args, **kwargs):
            warn_deprecated_function(func, message)

            return func(*args, **kwargs)

        return wrapper
    return decorator


def _showwarning(message, category, filename, lineno, file=None, line=None):
    """
    Redirect warnings into logging.
    """

    fmtmsg = warnings.formatwarning(message, category, filename, lineno, line)
    LOG.warning(fmtmsg)


# Install our warnings handler
warnings.showwarning = _showwarning


def service_is_up(service):
    """Check whether a service is up based on last heartbeat."""
    last_heartbeat = service['updated_at'] or service['created_at']
    # Timestamps in DB are UTC.
    elapsed = total_seconds(utcnow() - last_heartbeat)
    return abs(elapsed) <= FLAGS.service_down_time