Fix parallel test suite

IPython/parallel/apps/baseapp.py

@@ -1,275 +1,275 b''
 # encoding: utf-8
 """
 The Base Application class for IPython.parallel apps
 
 Authors:
 
 * Brian Granger
 * Min RK
 
 """
 
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2008-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 import os
 import logging
 import re
 import sys
 
 from subprocess import Popen, PIPE
 
 from IPython.config.application import catch_config_error, LevelFormatter
 from IPython.core import release
 from IPython.core.crashhandler import CrashHandler
 from IPython.core.application import (
     BaseIPythonApplication,
     base_aliases as base_ip_aliases,
     base_flags as base_ip_flags
 )
 from IPython.utils.path import expand_path
 from IPython.utils.py3compat import unicode_type
 
 from IPython.utils.traitlets import Unicode, Bool, Instance, Dict
 
 #-----------------------------------------------------------------------------
 # Module errors
 #-----------------------------------------------------------------------------
 
 class PIDFileError(Exception):
     pass
 
 
 #-----------------------------------------------------------------------------
 # Crash handler for this application
 #-----------------------------------------------------------------------------
 
 class ParallelCrashHandler(CrashHandler):
     """sys.excepthook for IPython itself, leaves a detailed report on disk."""
 
     def __init__(self, app):
         contact_name = release.authors['Min'][0]
         contact_email = release.author_email
         bug_tracker = 'https://github.com/ipython/ipython/issues'
         super(ParallelCrashHandler,self).__init__(
             app, contact_name, contact_email, bug_tracker
         )
 
 
 #-----------------------------------------------------------------------------
 # Main application
 #-----------------------------------------------------------------------------
 base_aliases = {}
 base_aliases.update(base_ip_aliases)
 base_aliases.update({
     'work-dir' : 'BaseParallelApplication.work_dir',
     'log-to-file' : 'BaseParallelApplication.log_to_file',
     'clean-logs' : 'BaseParallelApplication.clean_logs',
     'log-url' : 'BaseParallelApplication.log_url',
     'cluster-id' : 'BaseParallelApplication.cluster_id',
 })
 
 base_flags = {
     'log-to-file' : (
         {'BaseParallelApplication' : {'log_to_file' : True}},
         "send log output to a file"
     )
 }
 base_flags.update(base_ip_flags)
 
 class BaseParallelApplication(BaseIPythonApplication):
     """The base Application for IPython.parallel apps
     
     Principle extensions to BaseIPyythonApplication:
     
     * work_dir
     * remote logging via pyzmq
     * IOLoop instance
     """
 
     crash_handler_class = ParallelCrashHandler
     
     def _log_level_default(self):
         # temporarily override default_log_level to INFO
         return logging.INFO
     
     def _log_format_default(self):
         """override default log format to include time"""
         return u"%(asctime)s.%(msecs).03d [%(name)s]%(highlevel)s %(message)s"
 
     work_dir = Unicode(os.getcwdu(), config=True,
         help='Set the working dir for the process.'
     )
     def _work_dir_changed(self, name, old, new):
         self.work_dir = unicode_type(expand_path(new))
 
     log_to_file = Bool(config=True,
         help="whether to log to a file")
 
     clean_logs = Bool(False, config=True,
         help="whether to cleanup old logfiles before starting")
 
     log_url = Unicode('', config=True,
         help="The ZMQ URL of the iplogger to aggregate logging.")
 
     cluster_id = Unicode('', config=True,
         help="""String id to add to runtime files, to prevent name collisions when
         using multiple clusters with a single profile simultaneously.
         
         When set, files will be named like: 'ipcontroller-<cluster_id>-engine.json'
         
         Since this is text inserted into filenames, typical recommendations apply:
         Simple character strings are ideal, and spaces are not recommended (but should
         generally work).
         """
     )
     def _cluster_id_changed(self, name, old, new):
         self.name = self.__class__.name
         if new:
             self.name += '-%s'%new
     
     def _config_files_default(self):
         return ['ipcontroller_config.py', 'ipengine_config.py', 'ipcluster_config.py']
     
     loop = Instance('zmq.eventloop.ioloop.IOLoop')
     def _loop_default(self):
         from zmq.eventloop.ioloop import IOLoop
         return IOLoop.instance()
 
     aliases = Dict(base_aliases)
     flags = Dict(base_flags)
     
     @catch_config_error
     def initialize(self, argv=None):
         """initialize the app"""
         super(BaseParallelApplication, self).initialize(argv)
         self.to_work_dir()
         self.reinit_logging()
         
     def to_work_dir(self):
         wd = self.work_dir
         if unicode_type(wd) != os.getcwdu():
             os.chdir(wd)
             self.log.info("Changing to working dir: %s" % wd)
         # This is the working dir by now.
         sys.path.insert(0, '')
 
     def reinit_logging(self):
         # Remove old log files
         log_dir = self.profile_dir.log_dir
         if self.clean_logs:
             for f in os.listdir(log_dir):
                 if re.match(r'%s-\d+\.(log|err|out)' % self.name, f):
                     try:
                         os.remove(os.path.join(log_dir, f))
                     except (OSError, IOError):
                         # probably just conflict from sibling process
                         # already removing it
                         pass
         if self.log_to_file:
             # Start logging to the new log file
             log_filename = self.name + u'-' + str(os.getpid()) + u'.log'
             logfile = os.path.join(log_dir, log_filename)
             open_log_file = open(logfile, 'w')
         else:
             open_log_file = None
         if open_log_file is not None:
             while self.log.handlers:
                 self.log.removeHandler(self.log.handlers[0])
             self._log_handler = logging.StreamHandler(open_log_file)
             self.log.addHandler(self._log_handler)
         else:
             self._log_handler = self.log.handlers[0]
         # Add timestamps to log format:
         self._log_formatter = LevelFormatter(self.log_format,
                                                 datefmt=self.log_datefmt)
         self._log_handler.setFormatter(self._log_formatter)
         # do not propagate log messages to root logger
         # ipcluster app will sometimes print duplicate messages during shutdown
         # if this is 1 (default):
         self.log.propagate = False
 
     def write_pid_file(self, overwrite=False):
         """Create a .pid file in the pid_dir with my pid.
 
         This must be called after pre_construct, which sets `self.pid_dir`.
         This raises :exc:`PIDFileError` if the pid file exists already.
         """
         pid_file = os.path.join(self.profile_dir.pid_dir, self.name + u'.pid')
         if os.path.isfile(pid_file):
             pid = self.get_pid_from_file()
             if not overwrite:
                 raise PIDFileError(
                     'The pid file [%s] already exists. \nThis could mean that this '
                     'server is already running with [pid=%s].' % (pid_file, pid)
                 )
         with open(pid_file, 'w') as f:
             self.log.info("Creating pid file: %s" % pid_file)
             f.write(repr(os.getpid())+'\n')
 
     def remove_pid_file(self):
         """Remove the pid file.
 
         This should be called at shutdown by registering a callback with
         :func:`reactor.addSystemEventTrigger`. This needs to return
         ``None``.
         """
         pid_file = os.path.join(self.profile_dir.pid_dir, self.name + u'.pid')
         if os.path.isfile(pid_file):
             try:
                 self.log.info("Removing pid file: %s" % pid_file)
                 os.remove(pid_file)
             except:
                 self.log.warn("Error removing the pid file: %s" % pid_file)
 
     def get_pid_from_file(self):
         """Get the pid from the pid file.
 
         If the  pid file doesn't exist a :exc:`PIDFileError` is raised.
         """
         pid_file = os.path.join(self.profile_dir.pid_dir, self.name + u'.pid')
         if os.path.isfile(pid_file):
             with open(pid_file, 'r') as f:
                 s = f.read().strip()
                 try:
                     pid = int(s)
                 except:
                     raise PIDFileError("invalid pid file: %s (contents: %r)"%(pid_file, s))
                 return pid
         else:
             raise PIDFileError('pid file not found: %s' % pid_file)
     
     def check_pid(self, pid):
         if os.name == 'nt':
             try:
                 import ctypes
                 # returns 0 if no such process (of ours) exists
                 # positive int otherwise
                 p = ctypes.windll.kernel32.OpenProcess(1,0,pid)
             except Exception:
                 self.log.warn(
                     "Could not determine whether pid %i is running via `OpenProcess`. "
                     " Making the likely assumption that it is."%pid
                 )
                 return True
             return bool(p)
         else:
             try:
                 p = Popen(['ps','x'], stdout=PIPE, stderr=PIPE)
                 output,_ = p.communicate()
             except OSError:
                 self.log.warn(
                     "Could not determine whether pid %i is running via `ps x`. "
                     " Making the likely assumption that it is."%pid
                 )
                 return True
-            pids = map(int, re.findall(r'^\W*\d+', output, re.MULTILINE))
+            pids = list(map(int, re.findall(r'^\W*\d+', output, re.MULTILINE)))
             return pid in pids

IPython/parallel/client/asyncresult.py

@@ -1,707 +1,707 b''
 """AsyncResult objects for the client
 
 Authors:
 
 * MinRK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 from __future__ import print_function
 
 import sys
 import time
 from datetime import datetime
 
 from zmq import MessageTracker
 
 from IPython.core.display import clear_output, display, display_pretty
 from IPython.external.decorator import decorator
 from IPython.parallel import error
 from IPython.utils.py3compat import string_types
 
 #-----------------------------------------------------------------------------
 # Functions
 #-----------------------------------------------------------------------------
 
 def _raw_text(s):
     display_pretty(s, raw=True)
 
 #-----------------------------------------------------------------------------
 # Classes
 #-----------------------------------------------------------------------------
 
 # global empty tracker that's always done:
 finished_tracker = MessageTracker()
 
 @decorator
 def check_ready(f, self, *args, **kwargs):
     """Call spin() to sync state prior to calling the method."""
     self.wait(0)
     if not self._ready:
         raise error.TimeoutError("result not ready")
     return f(self, *args, **kwargs)
 
 class AsyncResult(object):
     """Class for representing results of non-blocking calls.
 
     Provides the same interface as :py:class:`multiprocessing.pool.AsyncResult`.
     """
 
     msg_ids = None
     _targets = None
     _tracker = None
     _single_result = False
 
     def __init__(self, client, msg_ids, fname='unknown', targets=None, tracker=None):
         if isinstance(msg_ids, string_types):
             # always a list
             msg_ids = [msg_ids]
             self._single_result = True
         else:
             self._single_result = False
         if tracker is None:
             # default to always done
             tracker = finished_tracker
         self._client = client
         self.msg_ids = msg_ids
         self._fname=fname
         self._targets = targets
         self._tracker = tracker
         
         self._ready = False
         self._outputs_ready = False
         self._success = None
-        self._metadata = [ self._client.metadata.get(id) for id in self.msg_ids ]
+        self._metadata = [self._client.metadata[id] for id in self.msg_ids]
 
     def __repr__(self):
         if self._ready:
             return "<%s: finished>"%(self.__class__.__name__)
         else:
             return "<%s: %s>"%(self.__class__.__name__,self._fname)
 
 
     def _reconstruct_result(self, res):
         """Reconstruct our result from actual result list (always a list)
 
         Override me in subclasses for turning a list of results
         into the expected form.
         """
         if self._single_result:
             return res[0]
         else:
             return res
 
     def get(self, timeout=-1):
         """Return the result when it arrives.
 
         If `timeout` is not ``None`` and the result does not arrive within
         `timeout` seconds then ``TimeoutError`` is raised. If the
         remote call raised an exception then that exception will be reraised
         by get() inside a `RemoteError`.
         """
         if not self.ready():
             self.wait(timeout)
 
         if self._ready:
             if self._success:
                 return self._result
             else:
                 raise self._exception
         else:
             raise error.TimeoutError("Result not ready.")
 
     def _check_ready(self):
         if not self.ready():
             raise error.TimeoutError("Result not ready.")
     
     def ready(self):
         """Return whether the call has completed."""
         if not self._ready:
             self.wait(0)
         elif not self._outputs_ready:
             self._wait_for_outputs(0)
         
         return self._ready
 
     def wait(self, timeout=-1):
         """Wait until the result is available or until `timeout` seconds pass.
 
         This method always returns None.
         """
         if self._ready:
             self._wait_for_outputs(timeout)
             return
         self._ready = self._client.wait(self.msg_ids, timeout)
         if self._ready:
             try:
-                results = map(self._client.results.get, self.msg_ids)
+                results = list(map(self._client.results.get, self.msg_ids))
                 self._result = results
                 if self._single_result:
                     r = results[0]
                     if isinstance(r, Exception):
                         raise r
                 else:
                     results = error.collect_exceptions(results, self._fname)
                 self._result = self._reconstruct_result(results)
             except Exception as e:
                 self._exception = e
                 self._success = False
             else:
                 self._success = True
             finally:
                 if timeout is None or timeout < 0:
                     # cutoff infinite wait at 10s
                     timeout = 10
                 self._wait_for_outputs(timeout)
 
 
     def successful(self):
         """Return whether the call completed without raising an exception.
 
         Will raise ``AssertionError`` if the result is not ready.
         """
         assert self.ready()
         return self._success
 
     #----------------------------------------------------------------
     # Extra methods not in mp.pool.AsyncResult
     #----------------------------------------------------------------
 
     def get_dict(self, timeout=-1):
         """Get the results as a dict, keyed by engine_id.
 
         timeout behavior is described in `get()`.
         """
 
         results = self.get(timeout)
         if self._single_result:
             results = [results]
         engine_ids = [ md['engine_id'] for md in self._metadata ]
         
         
         rdict = {}
         for engine_id, result in zip(engine_ids, results):
             if engine_id in rdict:
                 raise ValueError("Cannot build dict, %i jobs ran on engine #%i" % (
                     engine_ids.count(engine_id), engine_id)
                 )
             else:
                 rdict[engine_id] = result
 
         return rdict
 
     @property
     def result(self):
         """result property wrapper for `get(timeout=-1)`."""
         return self.get()
 
     # abbreviated alias:
     r = result
 
     @property
     def metadata(self):
         """property for accessing execution metadata."""
         if self._single_result:
             return self._metadata[0]
         else:
             return self._metadata
 
     @property
     def result_dict(self):
         """result property as a dict."""
         return self.get_dict()
 
     def __dict__(self):
         return self.get_dict(0)
 
     def abort(self):
         """abort my tasks."""
         assert not self.ready(), "Can't abort, I am already done!"
         return self._client.abort(self.msg_ids, targets=self._targets, block=True)
 
     @property
     def sent(self):
         """check whether my messages have been sent."""
         return self._tracker.done
 
     def wait_for_send(self, timeout=-1):
         """wait for pyzmq send to complete.
 
         This is necessary when sending arrays that you intend to edit in-place.
         `timeout` is in seconds, and will raise TimeoutError if it is reached
         before the send completes.
         """
         return self._tracker.wait(timeout)
 
     #-------------------------------------
     # dict-access
     #-------------------------------------
 
     def __getitem__(self, key):
         """getitem returns result value(s) if keyed by int/slice, or metadata if key is str.
         """
         if isinstance(key, int):
             self._check_ready()
             return error.collect_exceptions([self._result[key]], self._fname)[0]
         elif isinstance(key, slice):
             self._check_ready()
             return error.collect_exceptions(self._result[key], self._fname)
         elif isinstance(key, string_types):
             # metadata proxy *does not* require that results are done
             self.wait(0)
             values = [ md[key] for md in self._metadata ]
             if self._single_result:
                 return values[0]
             else:
                 return values
         else:
             raise TypeError("Invalid key type %r, must be 'int','slice', or 'str'"%type(key))
 
     def __getattr__(self, key):
         """getattr maps to getitem for convenient attr access to metadata."""
         try:
             return self.__getitem__(key)
         except (error.TimeoutError, KeyError):
             raise AttributeError("%r object has no attribute %r"%(
                     self.__class__.__name__, key))
 
     # asynchronous iterator:
     def __iter__(self):
         if self._single_result:
             raise TypeError("AsyncResults with a single result are not iterable.")
         try:
             rlist = self.get(0)
         except error.TimeoutError:
             # wait for each result individually
             for msg_id in self.msg_ids:
                 ar = AsyncResult(self._client, msg_id, self._fname)
                 yield ar.get()
         else:
             # already done
             for r in rlist:
                 yield r
     
     def __len__(self):
         return len(self.msg_ids)
     
     #-------------------------------------
     # Sugar methods and attributes
     #-------------------------------------
     
     def timedelta(self, start, end, start_key=min, end_key=max):
         """compute the difference between two sets of timestamps
         
         The default behavior is to use the earliest of the first
         and the latest of the second list, but this can be changed
         by passing a different
         
         Parameters
         ----------
         
         start : one or more datetime objects (e.g. ar.submitted)
         end : one or more datetime objects (e.g. ar.received)
         start_key : callable
             Function to call on `start` to extract the relevant
             entry [defalt: min]
         end_key : callable
             Function to call on `end` to extract the relevant
             entry [default: max]
         
         Returns
         -------
         
         dt : float
             The time elapsed (in seconds) between the two selected timestamps.
         """
         if not isinstance(start, datetime):
             # handle single_result AsyncResults, where ar.stamp is single object,
             # not a list
             start = start_key(start)
         if not isinstance(end, datetime):
             # handle single_result AsyncResults, where ar.stamp is single object,
             # not a list
             end = end_key(end)
         return (end - start).total_seconds()
         
     @property
     def progress(self):
         """the number of tasks which have been completed at this point.
         
         Fractional progress would be given by 1.0 * ar.progress / len(ar)
         """
         self.wait(0)
         return len(self) - len(set(self.msg_ids).intersection(self._client.outstanding))
     
     @property
     def elapsed(self):
         """elapsed time since initial submission"""
         if self.ready():
             return self.wall_time
         
         now = submitted = datetime.now()
         for msg_id in self.msg_ids:
             if msg_id in self._client.metadata:
                 stamp = self._client.metadata[msg_id]['submitted']
                 if stamp and stamp < submitted:
                     submitted = stamp
         return (now-submitted).total_seconds()
     
     @property
     @check_ready
     def serial_time(self):
         """serial computation time of a parallel calculation
         
         Computed as the sum of (completed-started) of each task
         """
         t = 0
         for md in self._metadata:
             t += (md['completed'] - md['started']).total_seconds()
         return t
     
     @property
     @check_ready
     def wall_time(self):
         """actual computation time of a parallel calculation
         
         Computed as the time between the latest `received` stamp
         and the earliest `submitted`.
         
         Only reliable if Client was spinning/waiting when the task finished, because
         the `received` timestamp is created when a result is pulled off of the zmq queue,
         which happens as a result of `client.spin()`.
         
         For similar comparison of other timestamp pairs, check out AsyncResult.timedelta.
         
         """
         return self.timedelta(self.submitted, self.received)
     
     def wait_interactive(self, interval=1., timeout=-1):
         """interactive wait, printing progress at regular intervals"""
         if timeout is None:
             timeout = -1
         N = len(self)
         tic = time.time()
         while not self.ready() and (timeout < 0 or time.time() - tic <= timeout):
             self.wait(interval)
             clear_output(wait=True)
             print("%4i/%i tasks finished after %4i s" % (self.progress, N, self.elapsed), end="")
             sys.stdout.flush()
         print()
         print("done")
     
     def _republish_displaypub(self, content, eid):
         """republish individual displaypub content dicts"""
         try:
             ip = get_ipython()
         except NameError:
             # displaypub is meaningless outside IPython
             return
         md = content['metadata'] or {}
         md['engine'] = eid
         ip.display_pub.publish(content['source'], content['data'], md)
     
     def _display_stream(self, text, prefix='', file=None):
         if not text:
             # nothing to display
             return
         if file is None:
             file = sys.stdout
         end = '' if text.endswith('\n') else '\n'
         
         multiline = text.count('\n') > int(text.endswith('\n'))
         if prefix and multiline and not text.startswith('\n'):
             prefix = prefix + '\n'
         print("%s%s" % (prefix, text), file=file, end=end)
         
     
     def _display_single_result(self):
         self._display_stream(self.stdout)
         self._display_stream(self.stderr, file=sys.stderr)
         
         try:
             get_ipython()
         except NameError:
             # displaypub is meaningless outside IPython
             return
         
         for output in self.outputs:
             self._republish_displaypub(output, self.engine_id)
         
         if self.pyout is not None:
             display(self.get())
     
     def _wait_for_outputs(self, timeout=-1):
         """wait for the 'status=idle' message that indicates we have all outputs
         """
         if self._outputs_ready or not self._success:
             # don't wait on errors
             return
         
         # cast None to -1 for infinite timeout
         if timeout is None:
             timeout = -1
         
         tic = time.time()
         while True:
             self._client._flush_iopub(self._client._iopub_socket)
             self._outputs_ready = all(md['outputs_ready']
                                       for md in self._metadata)
             if self._outputs_ready or \
                (timeout >= 0 and time.time() > tic + timeout):
                 break
             time.sleep(0.01)
     
     @check_ready
     def display_outputs(self, groupby="type"):
         """republish the outputs of the computation
         
         Parameters
         ----------
         
         groupby : str [default: type]
             if 'type':
                 Group outputs by type (show all stdout, then all stderr, etc.):
                 
                 [stdout:1] foo
                 [stdout:2] foo
                 [stderr:1] bar
                 [stderr:2] bar
             if 'engine':
                 Display outputs for each engine before moving on to the next:
                 
                 [stdout:1] foo
                 [stderr:1] bar
                 [stdout:2] foo
                 [stderr:2] bar
                 
             if 'order':
                 Like 'type', but further collate individual displaypub
                 outputs.  This is meant for cases of each command producing
                 several plots, and you would like to see all of the first
                 plots together, then all of the second plots, and so on.
         """
         if self._single_result:
             self._display_single_result()
             return
         
         stdouts = self.stdout
         stderrs = self.stderr
         pyouts  = self.pyout
         output_lists = self.outputs
         results = self.get()
         
         targets = self.engine_id
         
         if groupby == "engine":
             for eid,stdout,stderr,outputs,r,pyout in zip(
                     targets, stdouts, stderrs, output_lists, results, pyouts
                 ):
                 self._display_stream(stdout, '[stdout:%i] ' % eid)
                 self._display_stream(stderr, '[stderr:%i] ' % eid, file=sys.stderr)
                 
                 try:
                     get_ipython()
                 except NameError:
                     # displaypub is meaningless outside IPython
                     return 
                 
                 if outputs or pyout is not None:
                     _raw_text('[output:%i]' % eid)
                 
                 for output in outputs:
                     self._republish_displaypub(output, eid)
                 
                 if pyout is not None:
                     display(r)
         
         elif groupby in ('type', 'order'):
             # republish stdout:
             for eid,stdout in zip(targets, stdouts):
                 self._display_stream(stdout, '[stdout:%i] ' % eid)
         
             # republish stderr:
             for eid,stderr in zip(targets, stderrs):
                 self._display_stream(stderr, '[stderr:%i] ' % eid, file=sys.stderr)
         
             try:
                 get_ipython()
             except NameError:
                 # displaypub is meaningless outside IPython
                 return
             
             if groupby == 'order':
                 output_dict = dict((eid, outputs) for eid,outputs in zip(targets, output_lists))
                 N = max(len(outputs) for outputs in output_lists)
                 for i in range(N):
                     for eid in targets:
                         outputs = output_dict[eid]
                         if len(outputs) >= N:
                             _raw_text('[output:%i]' % eid)
                             self._republish_displaypub(outputs[i], eid)
             else:
                 # republish displaypub output
                 for eid,outputs in zip(targets, output_lists):
                     if outputs:
                         _raw_text('[output:%i]' % eid)
                     for output in outputs:
                         self._republish_displaypub(output, eid)
         
             # finally, add pyout:
             for eid,r,pyout in zip(targets, results, pyouts):
                 if pyout is not None:
                     display(r)
         
         else:
             raise ValueError("groupby must be one of 'type', 'engine', 'collate', not %r" % groupby)
         
         
 
 
 class AsyncMapResult(AsyncResult):
     """Class for representing results of non-blocking gathers.
 
     This will properly reconstruct the gather.
     
     This class is iterable at any time, and will wait on results as they come.
     
     If ordered=False, then the first results to arrive will come first, otherwise
     results will be yielded in the order they were submitted.
     
     """
 
     def __init__(self, client, msg_ids, mapObject, fname='', ordered=True):
         AsyncResult.__init__(self, client, msg_ids, fname=fname)
         self._mapObject = mapObject
         self._single_result = False
         self.ordered = ordered
 
     def _reconstruct_result(self, res):
         """Perform the gather on the actual results."""
         return self._mapObject.joinPartitions(res)
 
     # asynchronous iterator:
     def __iter__(self):
         it = self._ordered_iter if self.ordered else self._unordered_iter
         for r in it():
             yield r
 
     # asynchronous ordered iterator:
     def _ordered_iter(self):
         """iterator for results *as they arrive*, preserving submission order."""
         try:
             rlist = self.get(0)
         except error.TimeoutError:
             # wait for each result individually
             for msg_id in self.msg_ids:
                 ar = AsyncResult(self._client, msg_id, self._fname)
                 rlist = ar.get()
                 try:
                     for r in rlist:
                         yield r
                 except TypeError:
                     # flattened, not a list
                     # this could get broken by flattened data that returns iterables
                     # but most calls to map do not expose the `flatten` argument
                     yield rlist
         else:
             # already done
             for r in rlist:
                 yield r
 
     # asynchronous unordered iterator:
     def _unordered_iter(self):
         """iterator for results *as they arrive*, on FCFS basis, ignoring submission order."""
         try:
             rlist = self.get(0)
         except error.TimeoutError:
             pending = set(self.msg_ids)
             while pending:
                 try:
                     self._client.wait(pending, 1e-3)
                 except error.TimeoutError:
                     # ignore timeout error, because that only means
                     # *some* jobs are outstanding
                     pass
                 # update ready set with those no longer outstanding:
                 ready = pending.difference(self._client.outstanding)
                 # update pending to exclude those that are finished
                 pending = pending.difference(ready)
                 while ready:
                     msg_id = ready.pop()
                     ar = AsyncResult(self._client, msg_id, self._fname)
                     rlist = ar.get()
                     try:
                         for r in rlist:
                             yield r
                     except TypeError:
                         # flattened, not a list
                         # this could get broken by flattened data that returns iterables
                         # but most calls to map do not expose the `flatten` argument
                         yield rlist
         else:
             # already done
             for r in rlist:
                 yield r
 
 
 class AsyncHubResult(AsyncResult):
     """Class to wrap pending results that must be requested from the Hub.
 
     Note that waiting/polling on these objects requires polling the Hubover the network,
     so use `AsyncHubResult.wait()` sparingly.
     """
 
     def _wait_for_outputs(self, timeout=-1):
         """no-op, because HubResults are never incomplete"""
         self._outputs_ready = True
     
     def wait(self, timeout=-1):
         """wait for result to complete."""
         start = time.time()
         if self._ready:
             return
-        local_ids = filter(lambda msg_id: msg_id in self._client.outstanding, self.msg_ids)
+        local_ids = [m for m in self.msg_ids if m in self._client.outstanding]
         local_ready = self._client.wait(local_ids, timeout)
         if local_ready:
-            remote_ids = filter(lambda msg_id: msg_id not in self._client.results, self.msg_ids)
+            remote_ids = [m for m in self.msg_ids if m not in self._client.results]
             if not remote_ids:
                 self._ready = True
             else:
                 rdict = self._client.result_status(remote_ids, status_only=False)
                 pending = rdict['pending']
                 while pending and (timeout < 0 or time.time() < start+timeout):
                     rdict = self._client.result_status(remote_ids, status_only=False)
                     pending = rdict['pending']
                     if pending:
                         time.sleep(0.1)
                 if not pending:
                     self._ready = True
         if self._ready:
             try:
-                results = map(self._client.results.get, self.msg_ids)
+                results = list(map(self._client.results.get, self.msg_ids))
                 self._result = results
                 if self._single_result:
                     r = results[0]
                     if isinstance(r, Exception):
                         raise r
                 else:
                     results = error.collect_exceptions(results, self._fname)
                 self._result = self._reconstruct_result(results)
             except Exception as e:
                 self._exception = e
                 self._success = False
             else:
                 self._success = True
             finally:
-                self._metadata = map(self._client.metadata.get, self.msg_ids)
+                self._metadata = [self._client.metadata[mid] for mid in self.msg_ids]
 
 __all__ = ['AsyncResult', 'AsyncMapResult', 'AsyncHubResult']

IPython/parallel/client/client.py

@@ -1,1854 +1,1855 b''
 """A semi-synchronous Client for the ZMQ cluster
 
 Authors:
 
 * MinRK
 """
 from __future__ import print_function
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 import os
 import json
 import sys
 from threading import Thread, Event
 import time
 import warnings
 from datetime import datetime
 from getpass import getpass
 from pprint import pprint
 
 pjoin = os.path.join
 
 import zmq
 # from zmq.eventloop import ioloop, zmqstream
 
 from IPython.config.configurable import MultipleInstanceError
 from IPython.core.application import BaseIPythonApplication
 from IPython.core.profiledir import ProfileDir, ProfileDirError
 
 from IPython.utils.capture import RichOutput
 from IPython.utils.coloransi import TermColors
 from IPython.utils.jsonutil import rekey
 from IPython.utils.localinterfaces import localhost, is_local_ip
 from IPython.utils.path import get_ipython_dir
 from IPython.utils.py3compat import cast_bytes, string_types, xrange, iteritems
 from IPython.utils.traitlets import (HasTraits, Integer, Instance, Unicode,
                                     Dict, List, Bool, Set, Any)
 from IPython.external.decorator import decorator
 from IPython.external.ssh import tunnel
 
 from IPython.parallel import Reference
 from IPython.parallel import error
 from IPython.parallel import util
 
 from IPython.kernel.zmq.session import Session, Message
 from IPython.kernel.zmq import serialize
 
 from .asyncresult import AsyncResult, AsyncHubResult
 from .view import DirectView, LoadBalancedView
 
 #--------------------------------------------------------------------------
 # Decorators for Client methods
 #--------------------------------------------------------------------------
 
 @decorator
 def spin_first(f, self, *args, **kwargs):
     """Call spin() to sync state prior to calling the method."""
     self.spin()
     return f(self, *args, **kwargs)
 
 
 #--------------------------------------------------------------------------
 # Classes
 #--------------------------------------------------------------------------
 
 
 class ExecuteReply(RichOutput):
     """wrapper for finished Execute results"""
     def __init__(self, msg_id, content, metadata):
         self.msg_id = msg_id
         self._content = content
         self.execution_count = content['execution_count']
         self.metadata = metadata
     
     # RichOutput overrides
     
     @property
     def source(self):
         pyout = self.metadata['pyout']
         if pyout:
             return pyout.get('source', '')
     
     @property
     def data(self):
         pyout = self.metadata['pyout']
         if pyout:
             return pyout.get('data', {})
     
     @property
     def _metadata(self):
         pyout = self.metadata['pyout']
         if pyout:
             return pyout.get('metadata', {})
     
     def display(self):
         from IPython.display import publish_display_data
         publish_display_data(self.source, self.data, self.metadata)
     
     def _repr_mime_(self, mime):
         if mime not in self.data:
             return
         data = self.data[mime]
         if mime in self._metadata:
             return data, self._metadata[mime]
         else:
             return data
     
     def __getitem__(self, key):
         return self.metadata[key]
     
     def __getattr__(self, key):
         if key not in self.metadata:
             raise AttributeError(key)
         return self.metadata[key]
     
     def __repr__(self):
         pyout = self.metadata['pyout'] or {'data':{}}
         text_out = pyout['data'].get('text/plain', '')
         if len(text_out) > 32:
             text_out = text_out[:29] + '...'
         
         return "<ExecuteReply[%i]: %s>" % (self.execution_count, text_out)
     
     def _repr_pretty_(self, p, cycle):
         pyout = self.metadata['pyout'] or {'data':{}}
         text_out = pyout['data'].get('text/plain', '')
         
         if not text_out:
             return
         
         try:
             ip = get_ipython()
         except NameError:
             colors = "NoColor"
         else:
             colors = ip.colors
         
         if colors == "NoColor":
             out = normal = ""
         else:
             out = TermColors.Red
             normal = TermColors.Normal
         
         if '\n' in text_out and not text_out.startswith('\n'):
             # add newline for multiline reprs
             text_out = '\n' + text_out
         
         p.text(
             out + u'Out[%i:%i]: ' % (
                 self.metadata['engine_id'], self.execution_count
             ) + normal + text_out
         )
 
 
 class Metadata(dict):
     """Subclass of dict for initializing metadata values.
 
     Attribute access works on keys.
 
     These objects have a strict set of keys - errors will raise if you try
     to add new keys.
     """
     def __init__(self, *args, **kwargs):
         dict.__init__(self)
         md = {'msg_id' : None,
               'submitted' : None,
               'started' : None,
               'completed' : None,
               'received' : None,
               'engine_uuid' : None,
               'engine_id' : None,
               'follow' : None,
               'after' : None,
               'status' : None,
 
               'pyin' : None,
               'pyout' : None,
               'pyerr' : None,
               'stdout' : '',
               'stderr' : '',
               'outputs' : [],
               'data': {},
               'outputs_ready' : False,
             }
         self.update(md)
         self.update(dict(*args, **kwargs))
 
     def __getattr__(self, key):
         """getattr aliased to getitem"""
         if key in self:
             return self[key]
         else:
             raise AttributeError(key)
 
     def __setattr__(self, key, value):
         """setattr aliased to setitem, with strict"""
         if key in self:
             self[key] = value
         else:
             raise AttributeError(key)
 
     def __setitem__(self, key, value):
         """strict static key enforcement"""
         if key in self:
             dict.__setitem__(self, key, value)
         else:
             raise KeyError(key)
 
 
 class Client(HasTraits):
     """A semi-synchronous client to the IPython ZMQ cluster
 
     Parameters
     ----------
 
     url_file : str/unicode; path to ipcontroller-client.json
         This JSON file should contain all the information needed to connect to a cluster,
         and is likely the only argument needed.
         Connection information for the Hub's registration.  If a json connector
         file is given, then likely no further configuration is necessary.
         [Default: use profile]
     profile : bytes
         The name of the Cluster profile to be used to find connector information.
         If run from an IPython application, the default profile will be the same
         as the running application, otherwise it will be 'default'.
     cluster_id : str
         String id to added to runtime files, to prevent name collisions when using
         multiple clusters with a single profile simultaneously.
         When set, will look for files named like: 'ipcontroller-<cluster_id>-client.json'
         Since this is text inserted into filenames, typical recommendations apply:
         Simple character strings are ideal, and spaces are not recommended (but
         should generally work)
     context : zmq.Context
         Pass an existing zmq.Context instance, otherwise the client will create its own.
     debug : bool
         flag for lots of message printing for debug purposes
     timeout : int/float
         time (in seconds) to wait for connection replies from the Hub
         [Default: 10]
 
     #-------------- session related args ----------------
 
     config : Config object
         If specified, this will be relayed to the Session for configuration
     username : str
         set username for the session object
 
     #-------------- ssh related args ----------------
     # These are args for configuring the ssh tunnel to be used
     # credentials are used to forward connections over ssh to the Controller
     # Note that the ip given in `addr` needs to be relative to sshserver
     # The most basic case is to leave addr as pointing to localhost (127.0.0.1),
     # and set sshserver as the same machine the Controller is on. However,
     # the only requirement is that sshserver is able to see the Controller
     # (i.e. is within the same trusted network).
 
     sshserver : str
         A string of the form passed to ssh, i.e. 'server.tld' or '[e-mail]:port'
         If keyfile or password is specified, and this is not, it will default to
         the ip given in addr.
     sshkey : str; path to ssh private key file
         This specifies a key to be used in ssh login, default None.
         Regular default ssh keys will be used without specifying this argument.
     password : str
         Your ssh password to sshserver. Note that if this is left None,
         you will be prompted for it if passwordless key based login is unavailable.
     paramiko : bool
         flag for whether to use paramiko instead of shell ssh for tunneling.
         [default: True on win32, False else]
 
 
     Attributes
     ----------
 
     ids : list of int engine IDs
         requesting the ids attribute always synchronizes
         the registration state. To request ids without synchronization,
         use semi-private _ids attributes.
 
     history : list of msg_ids
         a list of msg_ids, keeping track of all the execution
         messages you have submitted in order.
 
     outstanding : set of msg_ids
         a set of msg_ids that have been submitted, but whose
         results have not yet been received.
 
     results : dict
         a dict of all our results, keyed by msg_id
 
     block : bool
         determines default behavior when block not specified
         in execution methods
 
     Methods
     -------
 
     spin
         flushes incoming results and registration state changes
         control methods spin, and requesting `ids` also ensures up to date
 
     wait
         wait on one or more msg_ids
 
     execution methods
         apply
         legacy: execute, run
 
     data movement
         push, pull, scatter, gather
 
     query methods
         queue_status, get_result, purge, result_status
 
     control methods
         abort, shutdown
 
     """
 
 
     block = Bool(False)
     outstanding = Set()
     results = Instance('collections.defaultdict', (dict,))
     metadata = Instance('collections.defaultdict', (Metadata,))
     history = List()
     debug = Bool(False)
     _spin_thread = Any()
     _stop_spinning = Any()
 
     profile=Unicode()
     def _profile_default(self):
         if BaseIPythonApplication.initialized():
             # an IPython app *might* be running, try to get its profile
             try:
                 return BaseIPythonApplication.instance().profile
             except (AttributeError, MultipleInstanceError):
                 # could be a *different* subclass of config.Application,
                 # which would raise one of these two errors.
                 return u'default'
         else:
             return u'default'
 
 
     _outstanding_dict = Instance('collections.defaultdict', (set,))
     _ids = List()
     _connected=Bool(False)
     _ssh=Bool(False)
     _context = Instance('zmq.Context')
     _config = Dict()
     _engines=Instance(util.ReverseDict, (), {})
     # _hub_socket=Instance('zmq.Socket')
     _query_socket=Instance('zmq.Socket')
     _control_socket=Instance('zmq.Socket')
     _iopub_socket=Instance('zmq.Socket')
     _notification_socket=Instance('zmq.Socket')
     _mux_socket=Instance('zmq.Socket')
     _task_socket=Instance('zmq.Socket')
     _task_scheme=Unicode()
     _closed = False
     _ignored_control_replies=Integer(0)
     _ignored_hub_replies=Integer(0)
 
     def __new__(self, *args, **kw):
         # don't raise on positional args
         return HasTraits.__new__(self, **kw)
 
     def __init__(self, url_file=None, profile=None, profile_dir=None, ipython_dir=None,
             context=None, debug=False,
             sshserver=None, sshkey=None, password=None, paramiko=None,
             timeout=10, cluster_id=None, **extra_args
             ):
         if profile:
             super(Client, self).__init__(debug=debug, profile=profile)
         else:
             super(Client, self).__init__(debug=debug)
         if context is None:
             context = zmq.Context.instance()
         self._context = context
         self._stop_spinning = Event()
         
         if 'url_or_file' in extra_args:
             url_file = extra_args['url_or_file']
             warnings.warn("url_or_file arg no longer supported, use url_file", DeprecationWarning)
         
         if url_file and util.is_url(url_file):
             raise ValueError("single urls cannot be specified, url-files must be used.")
 
         self._setup_profile_dir(self.profile, profile_dir, ipython_dir)
         
         if self._cd is not None:
             if url_file is None:
                 if not cluster_id:
                     client_json = 'ipcontroller-client.json'
                 else:
                     client_json = 'ipcontroller-%s-client.json' % cluster_id
                 url_file = pjoin(self._cd.security_dir, client_json)
         if url_file is None:
             raise ValueError(
                 "I can't find enough information to connect to a hub!"
                 " Please specify at least one of url_file or profile."
             )
         
         with open(url_file) as f:
             cfg = json.load(f)
         
         self._task_scheme = cfg['task_scheme']
 
         # sync defaults from args, json:
         if sshserver:
             cfg['ssh'] = sshserver
 
         location = cfg.setdefault('location', None)
         
         proto,addr = cfg['interface'].split('://')
         addr = util.disambiguate_ip_address(addr, location)
         cfg['interface'] = "%s://%s" % (proto, addr)
         
         # turn interface,port into full urls:
         for key in ('control', 'task', 'mux', 'iopub', 'notification', 'registration'):
             cfg[key] = cfg['interface'] + ':%i' % cfg[key]
         
         url = cfg['registration']
         
         if location is not None and addr == localhost():
             # location specified, and connection is expected to be local
             if not is_local_ip(location) and not sshserver:
                 # load ssh from JSON *only* if the controller is not on
                 # this machine
                 sshserver=cfg['ssh']
             if not is_local_ip(location) and not sshserver:
                 # warn if no ssh specified, but SSH is probably needed
                 # This is only a warning, because the most likely cause
                 # is a local Controller on a laptop whose IP is dynamic
                 warnings.warn("""
             Controller appears to be listening on localhost, but not on this machine.
             If this is true, you should specify Client(...,sshserver='you@%s')
             or instruct your controller to listen on an external IP."""%location,
                 RuntimeWarning)
         elif not sshserver:
             # otherwise sync with cfg
             sshserver = cfg['ssh']
 
         self._config = cfg
 
         self._ssh = bool(sshserver or sshkey or password)
         if self._ssh and sshserver is None:
             # default to ssh via localhost
             sshserver = addr
         if self._ssh and password is None:
             if tunnel.try_passwordless_ssh(sshserver, sshkey, paramiko):
                 password=False
             else:
                 password = getpass("SSH Password for %s: "%sshserver)
         ssh_kwargs = dict(keyfile=sshkey, password=password, paramiko=paramiko)
 
         # configure and construct the session
         try:
             extra_args['packer'] = cfg['pack']
             extra_args['unpacker'] = cfg['unpack']
             extra_args['key'] = cast_bytes(cfg['key'])
             extra_args['signature_scheme'] = cfg['signature_scheme']
         except KeyError as exc:
             msg = '\n'.join([
                 "Connection file is invalid (missing '{}'), possibly from an old version of IPython.",
                 "If you are reusing connection files, remove them and start ipcontroller again."
             ])
             raise ValueError(msg.format(exc.message))
         
         self.session = Session(**extra_args)
 
         self._query_socket = self._context.socket(zmq.DEALER)
 
         if self._ssh:
             tunnel.tunnel_connection(self._query_socket, cfg['registration'], sshserver, **ssh_kwargs)
         else:
             self._query_socket.connect(cfg['registration'])
 
         self.session.debug = self.debug
 
         self._notification_handlers = {'registration_notification' : self._register_engine,
                                     'unregistration_notification' : self._unregister_engine,
                                     'shutdown_notification' : lambda msg: self.close(),
                                     }
         self._queue_handlers = {'execute_reply' : self._handle_execute_reply,
                                 'apply_reply' : self._handle_apply_reply}
         
         try:
             self._connect(sshserver, ssh_kwargs, timeout)
         except:
             self.close(linger=0)
             raise
         
         # last step: setup magics, if we are in IPython:
         
         try:
             ip = get_ipython()
         except NameError:
             return
         else:
             if 'px' not in ip.magics_manager.magics:
                 # in IPython but we are the first Client.
                 # activate a default view for parallel magics.
                 self.activate()
 
     def __del__(self):
         """cleanup sockets, but _not_ context."""
         self.close()
 
     def _setup_profile_dir(self, profile, profile_dir, ipython_dir):
         if ipython_dir is None:
             ipython_dir = get_ipython_dir()
         if profile_dir is not None:
             try:
                 self._cd = ProfileDir.find_profile_dir(profile_dir)
                 return
             except ProfileDirError:
                 pass
         elif profile is not None:
             try:
                 self._cd = ProfileDir.find_profile_dir_by_name(
                     ipython_dir, profile)
                 return
             except ProfileDirError:
                 pass
         self._cd = None
 
     def _update_engines(self, engines):
         """Update our engines dict and _ids from a dict of the form: {id:uuid}."""
         for k,v in iteritems(engines):
             eid = int(k)
             if eid not in self._engines:
                 self._ids.append(eid)
             self._engines[eid] = v
         self._ids = sorted(self._ids)
-        if sorted(self._engines.keys()) != range(len(self._engines)) and \
+        if sorted(self._engines.keys()) != list(range(len(self._engines))) and \
                         self._task_scheme == 'pure' and self._task_socket:
             self._stop_scheduling_tasks()
 
     def _stop_scheduling_tasks(self):
         """Stop scheduling tasks because an engine has been unregistered
         from a pure ZMQ scheduler.
         """
         self._task_socket.close()
         self._task_socket = None
         msg = "An engine has been unregistered, and we are using pure " +\
               "ZMQ task scheduling.  Task farming will be disabled."
         if self.outstanding:
             msg += " If you were running tasks when this happened, " +\
                    "some `outstanding` msg_ids may never resolve."
         warnings.warn(msg, RuntimeWarning)
 
     def _build_targets(self, targets):
         """Turn valid target IDs or 'all' into two lists:
         (int_ids, uuids).
         """
         if not self._ids:
             # flush notification socket if no engines yet, just in case
             if not self.ids:
                 raise error.NoEnginesRegistered("Can't build targets without any engines")
 
         if targets is None:
             targets = self._ids
         elif isinstance(targets, string_types):
             if targets.lower() == 'all':
                 targets = self._ids
             else:
                 raise TypeError("%r not valid str target, must be 'all'"%(targets))
         elif isinstance(targets, int):
             if targets < 0:
                 targets = self.ids[targets]
             if targets not in self._ids:
                 raise IndexError("No such engine: %i"%targets)
             targets = [targets]
 
         if isinstance(targets, slice):
-            indices = range(len(self._ids))[targets]
+            indices = list(range(len(self._ids))[targets])
             ids = self.ids
             targets = [ ids[i] for i in indices ]
 
         if not isinstance(targets, (tuple, list, xrange)):
             raise TypeError("targets by int/slice/collection of ints only, not %s"%(type(targets)))
 
         return [cast_bytes(self._engines[t]) for t in targets], list(targets)
 
     def _connect(self, sshserver, ssh_kwargs, timeout):
         """setup all our socket connections to the cluster. This is called from
         __init__."""
 
         # Maybe allow reconnecting?
         if self._connected:
             return
         self._connected=True
 
         def connect_socket(s, url):
             if self._ssh:
                 return tunnel.tunnel_connection(s, url, sshserver, **ssh_kwargs)
             else:
                 return s.connect(url)
 
         self.session.send(self._query_socket, 'connection_request')
         # use Poller because zmq.select has wrong units in pyzmq 2.1.7
         poller = zmq.Poller()
         poller.register(self._query_socket, zmq.POLLIN)
         # poll expects milliseconds, timeout is seconds
         evts = poller.poll(timeout*1000)
         if not evts:
             raise error.TimeoutError("Hub connection request timed out")
         idents,msg = self.session.recv(self._query_socket,mode=0)
         if self.debug:
             pprint(msg)
         content = msg['content']
         # self._config['registration'] = dict(content)
         cfg = self._config
         if content['status'] == 'ok':
             self._mux_socket = self._context.socket(zmq.DEALER)
             connect_socket(self._mux_socket, cfg['mux'])
 
             self._task_socket = self._context.socket(zmq.DEALER)
             connect_socket(self._task_socket, cfg['task'])
 
             self._notification_socket = self._context.socket(zmq.SUB)
             self._notification_socket.setsockopt(zmq.SUBSCRIBE, b'')
             connect_socket(self._notification_socket, cfg['notification'])
 
             self._control_socket = self._context.socket(zmq.DEALER)
             connect_socket(self._control_socket, cfg['control'])
 
             self._iopub_socket = self._context.socket(zmq.SUB)
             self._iopub_socket.setsockopt(zmq.SUBSCRIBE, b'')
             connect_socket(self._iopub_socket, cfg['iopub'])
 
             self._update_engines(dict(content['engines']))
         else:
             self._connected = False
             raise Exception("Failed to connect!")
 
     #--------------------------------------------------------------------------
     # handlers and callbacks for incoming messages
     #--------------------------------------------------------------------------
 
     def _unwrap_exception(self, content):
         """unwrap exception, and remap engine_id to int."""
         e = error.unwrap_exception(content)
         # print e.traceback
         if e.engine_info:
             e_uuid = e.engine_info['engine_uuid']
             eid = self._engines[e_uuid]
             e.engine_info['engine_id'] = eid
         return e
 
     def _extract_metadata(self, msg):
         header = msg['header']
         parent = msg['parent_header']
         msg_meta = msg['metadata']
         content = msg['content']
         md = {'msg_id' : parent['msg_id'],
               'received' : datetime.now(),
               'engine_uuid' : msg_meta.get('engine', None),
               'follow' : msg_meta.get('follow', []),
               'after' : msg_meta.get('after', []),
               'status' : content['status'],
             }
 
         if md['engine_uuid'] is not None:
             md['engine_id'] = self._engines.get(md['engine_uuid'], None)
 
         if 'date' in parent:
             md['submitted'] = parent['date']
         if 'started' in msg_meta:
             md['started'] = msg_meta['started']
         if 'date' in header:
             md['completed'] = header['date']
         return md
 
     def _register_engine(self, msg):
         """Register a new engine, and update our connection info."""
         content = msg['content']
         eid = content['id']
         d = {eid : content['uuid']}
         self._update_engines(d)
 
     def _unregister_engine(self, msg):
         """Unregister an engine that has died."""
         content = msg['content']
         eid = int(content['id'])
         if eid in self._ids:
             self._ids.remove(eid)
             uuid = self._engines.pop(eid)
 
             self._handle_stranded_msgs(eid, uuid)
 
         if self._task_socket and self._task_scheme == 'pure':
             self._stop_scheduling_tasks()
 
     def _handle_stranded_msgs(self, eid, uuid):
         """Handle messages known to be on an engine when the engine unregisters.
 
         It is possible that this will fire prematurely - that is, an engine will
         go down after completing a result, and the client will be notified
         of the unregistration and later receive the successful result.
         """
 
         outstanding = self._outstanding_dict[uuid]
 
         for msg_id in list(outstanding):
             if msg_id in self.results:
                 # we already
                 continue
             try:
                 raise error.EngineError("Engine %r died while running task %r"%(eid, msg_id))
             except:
                 content = error.wrap_exception()
             # build a fake message:
             msg = self.session.msg('apply_reply', content=content)
             msg['parent_header']['msg_id'] = msg_id
             msg['metadata']['engine'] = uuid
             self._handle_apply_reply(msg)
 
     def _handle_execute_reply(self, msg):
         """Save the reply to an execute_request into our results.
 
         execute messages are never actually used. apply is used instead.
         """
 
         parent = msg['parent_header']
         msg_id = parent['msg_id']
         if msg_id not in self.outstanding:
             if msg_id in self.history:
                 print(("got stale result: %s"%msg_id))
             else:
                 print(("got unknown result: %s"%msg_id))
         else:
             self.outstanding.remove(msg_id)
 
         content = msg['content']
         header = msg['header']
 
         # construct metadata:
         md = self.metadata[msg_id]
         md.update(self._extract_metadata(msg))
         # is this redundant?
         self.metadata[msg_id] = md
         
         e_outstanding = self._outstanding_dict[md['engine_uuid']]
         if msg_id in e_outstanding:
             e_outstanding.remove(msg_id)
 
         # construct result:
         if content['status'] == 'ok':
             self.results[msg_id] = ExecuteReply(msg_id, content, md)
         elif content['status'] == 'aborted':
             self.results[msg_id] = error.TaskAborted(msg_id)
         elif content['status'] == 'resubmitted':
             # TODO: handle resubmission
             pass
         else:
             self.results[msg_id] = self._unwrap_exception(content)
 
     def _handle_apply_reply(self, msg):
         """Save the reply to an apply_request into our results."""
         parent = msg['parent_header']
         msg_id = parent['msg_id']
         if msg_id not in self.outstanding:
             if msg_id in self.history:
                 print(("got stale result: %s"%msg_id))
                 print(self.results[msg_id])
                 print(msg)
             else:
                 print(("got unknown result: %s"%msg_id))
         else:
             self.outstanding.remove(msg_id)
         content = msg['content']
         header = msg['header']
 
         # construct metadata:
         md = self.metadata[msg_id]
         md.update(self._extract_metadata(msg))
         # is this redundant?
         self.metadata[msg_id] = md
 
         e_outstanding = self._outstanding_dict[md['engine_uuid']]
         if msg_id in e_outstanding:
             e_outstanding.remove(msg_id)
 
         # construct result:
         if content['status'] == 'ok':
             self.results[msg_id] = serialize.unserialize_object(msg['buffers'])[0]
         elif content['status'] == 'aborted':
             self.results[msg_id] = error.TaskAborted(msg_id)
         elif content['status'] == 'resubmitted':
             # TODO: handle resubmission
             pass
         else:
             self.results[msg_id] = self._unwrap_exception(content)
 
     def _flush_notifications(self):
         """Flush notifications of engine registrations waiting
         in ZMQ queue."""
         idents,msg = self.session.recv(self._notification_socket, mode=zmq.NOBLOCK)
         while msg is not None:
             if self.debug:
                 pprint(msg)
             msg_type = msg['header']['msg_type']
             handler = self._notification_handlers.get(msg_type, None)
             if handler is None:
                 raise Exception("Unhandled message type: %s" % msg_type)
             else:
                 handler(msg)
             idents,msg = self.session.recv(self._notification_socket, mode=zmq.NOBLOCK)
 
     def _flush_results(self, sock):
         """Flush task or queue results waiting in ZMQ queue."""
         idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
         while msg is not None:
             if self.debug:
                 pprint(msg)
             msg_type = msg['header']['msg_type']
             handler = self._queue_handlers.get(msg_type, None)
             if handler is None:
                 raise Exception("Unhandled message type: %s" % msg_type)
             else:
                 handler(msg)
             idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
 
     def _flush_control(self, sock):
         """Flush replies from the control channel waiting
         in the ZMQ queue.
 
         Currently: ignore them."""
         if self._ignored_control_replies <= 0:
             return
         idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
         while msg is not None:
             self._ignored_control_replies -= 1
             if self.debug:
                 pprint(msg)
             idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
 
     def _flush_ignored_control(self):
         """flush ignored control replies"""
         while self._ignored_control_replies > 0:
             self.session.recv(self._control_socket)
             self._ignored_control_replies -= 1
 
     def _flush_ignored_hub_replies(self):
         ident,msg = self.session.recv(self._query_socket, mode=zmq.NOBLOCK)
         while msg is not None:
             ident,msg = self.session.recv(self._query_socket, mode=zmq.NOBLOCK)
 
     def _flush_iopub(self, sock):
         """Flush replies from the iopub channel waiting
         in the ZMQ queue.
         """
         idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
         while msg is not None:
             if self.debug:
                 pprint(msg)
             parent = msg['parent_header']
             # ignore IOPub messages with no parent.
             # Caused by print statements or warnings from before the first execution.
             if not parent:
                 idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
                 continue
             msg_id = parent['msg_id']
             content = msg['content']
             header = msg['header']
             msg_type = msg['header']['msg_type']
 
             # init metadata:
             md = self.metadata[msg_id]
 
             if msg_type == 'stream':
                 name = content['name']
                 s = md[name] or ''
                 md[name] = s + content['data']
             elif msg_type == 'pyerr':
                 md.update({'pyerr' : self._unwrap_exception(content)})
             elif msg_type == 'pyin':
                 md.update({'pyin' : content['code']})
             elif msg_type == 'display_data':
                 md['outputs'].append(content)
             elif msg_type == 'pyout':
                 md['pyout'] = content
             elif msg_type == 'data_message':
                 data, remainder = serialize.unserialize_object(msg['buffers'])
                 md['data'].update(data)
             elif msg_type == 'status':
                 # idle message comes after all outputs
                 if content['execution_state'] == 'idle':
                     md['outputs_ready'] = True
             else:
                 # unhandled msg_type (status, etc.)
                 pass
 
             # reduntant?
             self.metadata[msg_id] = md
 
             idents,msg = self.session.recv(sock, mode=zmq.NOBLOCK)
 
     #--------------------------------------------------------------------------
     # len, getitem
     #--------------------------------------------------------------------------
 
     def __len__(self):
         """len(client) returns # of engines."""
         return len(self.ids)
 
     def __getitem__(self, key):
         """index access returns DirectView multiplexer objects
 
         Must be int, slice, or list/tuple/xrange of ints"""
         if not isinstance(key, (int, slice, tuple, list, xrange)):
             raise TypeError("key by int/slice/iterable of ints only, not %s"%(type(key)))
         else:
             return self.direct_view(key)
 
     #--------------------------------------------------------------------------
     # Begin public methods
     #--------------------------------------------------------------------------
 
     @property
     def ids(self):
         """Always up-to-date ids property."""
         self._flush_notifications()
         # always copy:
         return list(self._ids)
 
     def activate(self, targets='all', suffix=''):
         """Create a DirectView and register it with IPython magics
         
         Defines the magics `%px, %autopx, %pxresult, %%px`
         
         Parameters
         ----------
         
         targets: int, list of ints, or 'all'
             The engines on which the view's magics will run
         suffix: str [default: '']
             The suffix, if any, for the magics.  This allows you to have
             multiple views associated with parallel magics at the same time.
             
             e.g. ``rc.activate(targets=0, suffix='0')`` will give you
             the magics ``%px0``, ``%pxresult0``, etc. for running magics just
             on engine 0.
         """
         view = self.direct_view(targets)
         view.block = True
         view.activate(suffix)
         return view
 
     def close(self, linger=None):
         """Close my zmq Sockets
         
         If `linger`, set the zmq LINGER socket option,
         which allows discarding of messages.
         """
         if self._closed:
             return
         self.stop_spin_thread()
         snames = [ trait for trait in self.trait_names() if trait.endswith("socket") ]
         for name in snames:
             socket = getattr(self, name)
             if socket is not None and not socket.closed:
                 if linger is not None:
                     socket.close(linger=linger)
                 else:
                     socket.close()
         self._closed = True
 
     def _spin_every(self, interval=1):
         """target func for use in spin_thread"""
         while True:
             if self._stop_spinning.is_set():
                 return
             time.sleep(interval)
             self.spin()
 
     def spin_thread(self, interval=1):
         """call Client.spin() in a background thread on some regular interval
         
         This helps ensure that messages don't pile up too much in the zmq queue
         while you are working on other things, or just leaving an idle terminal.
         
         It also helps limit potential padding of the `received` timestamp
         on AsyncResult objects, used for timings.
         
         Parameters
         ----------
         
         interval : float, optional
             The interval on which to spin the client in the background thread
             (simply passed to time.sleep).
         
         Notes
         -----
         
         For precision timing, you may want to use this method to put a bound
         on the jitter (in seconds) in `received` timestamps used
         in AsyncResult.wall_time.
         
         """
         if self._spin_thread is not None:
             self.stop_spin_thread()
         self._stop_spinning.clear()
         self._spin_thread = Thread(target=self._spin_every, args=(interval,))
         self._spin_thread.daemon = True
         self._spin_thread.start()
     
     def stop_spin_thread(self):
         """stop background spin_thread, if any"""
         if self._spin_thread is not None:
             self._stop_spinning.set()
             self._spin_thread.join()
             self._spin_thread = None
 
     def spin(self):
         """Flush any registration notifications and execution results
         waiting in the ZMQ queue.
         """
         if self._notification_socket:
             self._flush_notifications()
         if self._iopub_socket:
             self._flush_iopub(self._iopub_socket)
         if self._mux_socket:
             self._flush_results(self._mux_socket)
         if self._task_socket:
             self._flush_results(self._task_socket)
         if self._control_socket:
             self._flush_control(self._control_socket)
         if self._query_socket:
             self._flush_ignored_hub_replies()
 
     def wait(self, jobs=None, timeout=-1):
         """waits on one or more `jobs`, for up to `timeout` seconds.
 
         Parameters
         ----------
 
         jobs : int, str, or list of ints and/or strs, or one or more AsyncResult objects
                 ints are indices to self.history
                 strs are msg_ids
                 default: wait on all outstanding messages
         timeout : float
                 a time in seconds, after which to give up.
                 default is -1, which means no timeout
 
         Returns
         -------
 
         True : when all msg_ids are done
         False : timeout reached, some msg_ids still outstanding
         """
         tic = time.time()
         if jobs is None:
             theids = self.outstanding
         else:
             if isinstance(jobs, string_types + (int, AsyncResult)):
                 jobs = [jobs]
             theids = set()
             for job in jobs:
                 if isinstance(job, int):
                     # index access
                     job = self.history[job]
                 elif isinstance(job, AsyncResult):
-                    map(theids.add, job.msg_ids)
+                    theids.update(job.msg_ids)
                     continue
                 theids.add(job)
         if not theids.intersection(self.outstanding):
             return True
         self.spin()
         while theids.intersection(self.outstanding):
             if timeout >= 0 and ( time.time()-tic ) > timeout:
                 break
             time.sleep(1e-3)
             self.spin()
         return len(theids.intersection(self.outstanding)) == 0
 
     #--------------------------------------------------------------------------
     # Control methods
     #--------------------------------------------------------------------------
 
     @spin_first
     def clear(self, targets=None, block=None):
         """Clear the namespace in target(s)."""
         block = self.block if block is None else block
         targets = self._build_targets(targets)[0]
         for t in targets:
             self.session.send(self._control_socket, 'clear_request', content={}, ident=t)
         error = False
         if block:
             self._flush_ignored_control()
             for i in range(len(targets)):
                 idents,msg = self.session.recv(self._control_socket,0)
                 if self.debug:
                     pprint(msg)
                 if msg['content']['status'] != 'ok':
                     error = self._unwrap_exception(msg['content'])
         else:
             self._ignored_control_replies += len(targets)
         if error:
             raise error
 
 
     @spin_first
     def abort(self, jobs=None, targets=None, block=None):
         """Abort specific jobs from the execution queues of target(s).
 
         This is a mechanism to prevent jobs that have already been submitted
         from executing.
 
         Parameters
         ----------
 
         jobs : msg_id, list of msg_ids, or AsyncResult
             The jobs to be aborted
             
             If unspecified/None: abort all outstanding jobs.
 
         """
         block = self.block if block is None else block
         jobs = jobs if jobs is not None else list(self.outstanding)
         targets = self._build_targets(targets)[0]
         
         msg_ids = []
         if isinstance(jobs, string_types + (AsyncResult,)):
             jobs = [jobs]
-        bad_ids = filter(lambda obj: not isinstance(obj, string_types + (AsyncResult,)), jobs)
+        bad_ids = [obj for obj in jobs if not isinstance(obj, string_types + (AsyncResult,))]
         if bad_ids:
             raise TypeError("Invalid msg_id type %r, expected str or AsyncResult"%bad_ids[0])
         for j in jobs:
             if isinstance(j, AsyncResult):
                 msg_ids.extend(j.msg_ids)
             else:
                 msg_ids.append(j)
         content = dict(msg_ids=msg_ids)
         for t in targets:
             self.session.send(self._control_socket, 'abort_request',
                     content=content, ident=t)
         error = False
         if block:
             self._flush_ignored_control()
             for i in range(len(targets)):
                 idents,msg = self.session.recv(self._control_socket,0)
                 if self.debug:
                     pprint(msg)
                 if msg['content']['status'] != 'ok':
                     error = self._unwrap_exception(msg['content'])
         else:
             self._ignored_control_replies += len(targets)
         if error:
             raise error
 
     @spin_first
     def shutdown(self, targets='all', restart=False, hub=False, block=None):
         """Terminates one or more engine processes, optionally including the hub.
         
         Parameters
         ----------
         
         targets: list of ints or 'all' [default: all]
             Which engines to shutdown.
         hub: bool [default: False]
             Whether to include the Hub.  hub=True implies targets='all'.
         block: bool [default: self.block]
             Whether to wait for clean shutdown replies or not.
         restart: bool [default: False]
             NOT IMPLEMENTED
             whether to restart engines after shutting them down.
         """
         from IPython.parallel.error import NoEnginesRegistered
         if restart:
             raise NotImplementedError("Engine restart is not yet implemented")
         
         block = self.block if block is None else block
         if hub:
             targets = 'all'
         try:
             targets = self._build_targets(targets)[0]
         except NoEnginesRegistered:
             targets = []
         for t in targets:
             self.session.send(self._control_socket, 'shutdown_request',
                         content={'restart':restart},ident=t)
         error = False
         if block or hub:
             self._flush_ignored_control()
             for i in range(len(targets)):
                 idents,msg = self.session.recv(self._control_socket, 0)
                 if self.debug:
                     pprint(msg)
                 if msg['content']['status'] != 'ok':
                     error = self._unwrap_exception(msg['content'])
         else:
             self._ignored_control_replies += len(targets)
 
         if hub:
             time.sleep(0.25)
             self.session.send(self._query_socket, 'shutdown_request')
             idents,msg = self.session.recv(self._query_socket, 0)
             if self.debug:
                 pprint(msg)
             if msg['content']['status'] != 'ok':
                 error = self._unwrap_exception(msg['content'])
 
         if error:
             raise error
 
     #--------------------------------------------------------------------------
     # Execution related methods
     #--------------------------------------------------------------------------
 
     def _maybe_raise(self, result):
         """wrapper for maybe raising an exception if apply failed."""
         if isinstance(result, error.RemoteError):
             raise result
 
         return result
 
     def send_apply_request(self, socket, f, args=None, kwargs=None, metadata=None, track=False,
                             ident=None):
         """construct and send an apply message via a socket.
 
         This is the principal method with which all engine execution is performed by views.
         """
 
         if self._closed:
             raise RuntimeError("Client cannot be used after its sockets have been closed")
         
         # defaults:
         args = args if args is not None else []
         kwargs = kwargs if kwargs is not None else {}
         metadata = metadata if metadata is not None else {}
 
         # validate arguments
         if not callable(f) and not isinstance(f, Reference):
             raise TypeError("f must be callable, not %s"%type(f))
         if not isinstance(args, (tuple, list)):
             raise TypeError("args must be tuple or list, not %s"%type(args))
         if not isinstance(kwargs, dict):
             raise TypeError("kwargs must be dict, not %s"%type(kwargs))
         if not isinstance(metadata, dict):
             raise TypeError("metadata must be dict, not %s"%type(metadata))
 
         bufs = serialize.pack_apply_message(f, args, kwargs,
             buffer_threshold=self.session.buffer_threshold,
             item_threshold=self.session.item_threshold,
         )
 
         msg = self.session.send(socket, "apply_request", buffers=bufs, ident=ident,
                             metadata=metadata, track=track)
 
         msg_id = msg['header']['msg_id']
         self.outstanding.add(msg_id)
         if ident:
             # possibly routed to a specific engine
             if isinstance(ident, list):
                 ident = ident[-1]
             if ident in self._engines.values():
                 # save for later, in case of engine death
                 self._outstanding_dict[ident].add(msg_id)
         self.history.append(msg_id)
         self.metadata[msg_id]['submitted'] = datetime.now()
 
         return msg
 
     def send_execute_request(self, socket, code, silent=True, metadata=None, ident=None):
         """construct and send an execute request via a socket.
 
         """
 
         if self._closed:
             raise RuntimeError("Client cannot be used after its sockets have been closed")
         
         # defaults:
         metadata = metadata if metadata is not None else {}
 
         # validate arguments
         if not isinstance(code, string_types):
             raise TypeError("code must be text, not %s" % type(code))
         if not isinstance(metadata, dict):
             raise TypeError("metadata must be dict, not %s" % type(metadata))
         
         content = dict(code=code, silent=bool(silent), user_variables=[], user_expressions={})
 
 
         msg = self.session.send(socket, "execute_request", content=content, ident=ident,
                             metadata=metadata)
 
         msg_id = msg['header']['msg_id']
         self.outstanding.add(msg_id)
         if ident:
             # possibly routed to a specific engine
             if isinstance(ident, list):
                 ident = ident[-1]
             if ident in self._engines.values():
                 # save for later, in case of engine death
                 self._outstanding_dict[ident].add(msg_id)
         self.history.append(msg_id)
         self.metadata[msg_id]['submitted'] = datetime.now()
 
         return msg
 
     #--------------------------------------------------------------------------
     # construct a View object
     #--------------------------------------------------------------------------
 
     def load_balanced_view(self, targets=None):
         """construct a DirectView object.
 
         If no arguments are specified, create a LoadBalancedView
         using all engines.
 
         Parameters
         ----------
 
         targets: list,slice,int,etc. [default: use all engines]
             The subset of engines across which to load-balance
         """
         if targets == 'all':
             targets = None
         if targets is not None:
             targets = self._build_targets(targets)[1]
         return LoadBalancedView(client=self, socket=self._task_socket, targets=targets)
 
     def direct_view(self, targets='all'):
         """construct a DirectView object.
 
         If no targets are specified, create a DirectView using all engines.
         
         rc.direct_view('all') is distinguished from rc[:] in that 'all' will
         evaluate the target engines at each execution, whereas rc[:] will connect to
         all *current* engines, and that list will not change.
         
         That is, 'all' will always use all engines, whereas rc[:] will not use
         engines added after the DirectView is constructed.
 
         Parameters
         ----------
 
         targets: list,slice,int,etc. [default: use all engines]
             The engines to use for the View
         """
         single = isinstance(targets, int)
         # allow 'all' to be lazily evaluated at each execution
         if targets != 'all':
             targets = self._build_targets(targets)[1]
         if single:
             targets = targets[0]
         return DirectView(client=self, socket=self._mux_socket, targets=targets)
 
     #--------------------------------------------------------------------------
     # Query methods
     #--------------------------------------------------------------------------
 
     @spin_first
     def get_result(self, indices_or_msg_ids=None, block=None):
         """Retrieve a result by msg_id or history index, wrapped in an AsyncResult object.
 
         If the client already has the results, no request to the Hub will be made.
 
         This is a convenient way to construct AsyncResult objects, which are wrappers
         that include metadata about execution, and allow for awaiting results that
         were not submitted by this Client.
 
         It can also be a convenient way to retrieve the metadata associated with
         blocking execution, since it always retrieves
 
         Examples
         --------
         ::
 
             In [10]: r = client.apply()
 
         Parameters
         ----------
 
         indices_or_msg_ids : integer history index, str msg_id, or list of either
             The indices or msg_ids of indices to be retrieved
 
         block : bool
             Whether to wait for the result to be done
 
         Returns
         -------
 
         AsyncResult
             A single AsyncResult object will always be returned.
 
         AsyncHubResult
             A subclass of AsyncResult that retrieves results from the Hub
 
         """
         block = self.block if block is None else block
         if indices_or_msg_ids is None:
             indices_or_msg_ids = -1
         
         single_result = False
         if not isinstance(indices_or_msg_ids, (list,tuple)):
             indices_or_msg_ids = [indices_or_msg_ids]
             single_result = True
 
         theids = []
         for id in indices_or_msg_ids:
             if isinstance(id, int):
                 id = self.history[id]
             if not isinstance(id, string_types):
                 raise TypeError("indices must be str or int, not %r"%id)
             theids.append(id)
 
-        local_ids = filter(lambda msg_id: msg_id in self.outstanding or msg_id in self.results, theids)
-        remote_ids = filter(lambda msg_id: msg_id not in local_ids, theids)
+        local_ids = [msg_id for msg_id in theids if (msg_id in self.outstanding or msg_id in self.results)]
+        remote_ids = [msg_id for msg_id in theids if msg_id not in local_ids]
         
         # given single msg_id initially, get_result shot get the result itself,
         # not a length-one list
         if single_result:
             theids = theids[0]
 
         if remote_ids:
             ar = AsyncHubResult(self, msg_ids=theids)
         else:
             ar = AsyncResult(self, msg_ids=theids)
 
         if block:
             ar.wait()
 
         return ar
 
     @spin_first
     def resubmit(self, indices_or_msg_ids=None, metadata=None, block=None):
         """Resubmit one or more tasks.
 
         in-flight tasks may not be resubmitted.
 
         Parameters
         ----------
 
         indices_or_msg_ids : integer history index, str msg_id, or list of either
             The indices or msg_ids of indices to be retrieved
 
         block : bool
             Whether to wait for the result to be done
 
         Returns
         -------
 
         AsyncHubResult
             A subclass of AsyncResult that retrieves results from the Hub
 
         """
         block = self.block if block is None else block
         if indices_or_msg_ids is None:
             indices_or_msg_ids = -1
 
         if not isinstance(indices_or_msg_ids, (list,tuple)):
             indices_or_msg_ids = [indices_or_msg_ids]
 
         theids = []
         for id in indices_or_msg_ids:
             if isinstance(id, int):
                 id = self.history[id]
             if not isinstance(id, string_types):
                 raise TypeError("indices must be str or int, not %r"%id)
             theids.append(id)
 
         content = dict(msg_ids = theids)
 
         self.session.send(self._query_socket, 'resubmit_request', content)
 
         zmq.select([self._query_socket], [], [])
         idents,msg = self.session.recv(self._query_socket, zmq.NOBLOCK)
         if self.debug:
             pprint(msg)
         content = msg['content']
         if content['status'] != 'ok':
             raise self._unwrap_exception(content)
         mapping = content['resubmitted']
         new_ids = [ mapping[msg_id] for msg_id in theids ]
 
         ar = AsyncHubResult(self, msg_ids=new_ids)
 
         if block:
             ar.wait()
 
         return ar
 
     @spin_first
     def result_status(self, msg_ids, status_only=True):
         """Check on the status of the result(s) of the apply request with `msg_ids`.
 
         If status_only is False, then the actual results will be retrieved, else
         only the status of the results will be checked.
 
         Parameters
         ----------
 
         msg_ids : list of msg_ids
             if int:
                 Passed as index to self.history for convenience.
         status_only : bool (default: True)
             if False:
                 Retrieve the actual results of completed tasks.
 
         Returns
         -------
 
         results : dict
             There will always be the keys 'pending' and 'completed', which will
             be lists of msg_ids that are incomplete or complete. If `status_only`
             is False, then completed results will be keyed by their `msg_id`.
         """
         if not isinstance(msg_ids, (list,tuple)):
             msg_ids = [msg_ids]
 
         theids = []
         for msg_id in msg_ids:
             if isinstance(msg_id, int):
                 msg_id = self.history[msg_id]
             if not isinstance(msg_id, string_types):
                 raise TypeError("msg_ids must be str, not %r"%msg_id)
             theids.append(msg_id)
 
         completed = []
         local_results = {}
 
         # comment this block out to temporarily disable local shortcut:
         for msg_id in theids:
             if msg_id in self.results:
                 completed.append(msg_id)
                 local_results[msg_id] = self.results[msg_id]
                 theids.remove(msg_id)
 
         if theids: # some not locally cached
             content = dict(msg_ids=theids, status_only=status_only)
             msg = self.session.send(self._query_socket, "result_request", content=content)
             zmq.select([self._query_socket], [], [])
             idents,msg = self.session.recv(self._query_socket, zmq.NOBLOCK)
             if self.debug:
                 pprint(msg)
             content = msg['content']
             if content['status'] != 'ok':
                 raise self._unwrap_exception(content)
             buffers = msg['buffers']
         else:
             content = dict(completed=[],pending=[])
 
         content['completed'].extend(completed)
 
         if status_only:
             return content
 
         failures = []
         # load cached results into result:
         content.update(local_results)
 
         # update cache with results:
         for msg_id in sorted(theids):
             if msg_id in content['completed']:
                 rec = content[msg_id]
                 parent = rec['header']
                 header = rec['result_header']
                 rcontent = rec['result_content']
                 iodict = rec['io']
                 if isinstance(rcontent, str):
                     rcontent = self.session.unpack(rcontent)
 
                 md = self.metadata[msg_id]
                 md_msg = dict(
                     content=rcontent,
                     parent_header=parent,
                     header=header,
                     metadata=rec['result_metadata'],
                 )
                 md.update(self._extract_metadata(md_msg))
                 if rec.get('received'):
                     md['received'] = rec['received']
                 md.update(iodict)
                 
                 if rcontent['status'] == 'ok':
                     if header['msg_type'] == 'apply_reply':
                         res,buffers = serialize.unserialize_object(buffers)
                     elif header['msg_type'] == 'execute_reply':
                         res = ExecuteReply(msg_id, rcontent, md)
                     else:
                         raise KeyError("unhandled msg type: %r" % header['msg_type'])
                 else:
                     res = self._unwrap_exception(rcontent)
                     failures.append(res)
 
                 self.results[msg_id] = res
                 content[msg_id] = res
 
         if len(theids) == 1 and failures:
             raise failures[0]
 
         error.collect_exceptions(failures, "result_status")
         return content
 
     @spin_first
     def queue_status(self, targets='all', verbose=False):
         """Fetch the status of engine queues.
 
         Parameters
         ----------
 
         targets : int/str/list of ints/strs
                 the engines whose states are to be queried.
                 default : all
         verbose : bool
                 Whether to return lengths only, or lists of ids for each element
         """
         if targets == 'all':
             # allow 'all' to be evaluated on the engine
             engine_ids = None
         else:
             engine_ids = self._build_targets(targets)[1]
         content = dict(targets=engine_ids, verbose=verbose)
         self.session.send(self._query_socket, "queue_request", content=content)
         idents,msg = self.session.recv(self._query_socket, 0)
         if self.debug:
             pprint(msg)
         content = msg['content']
         status = content.pop('status')
         if status != 'ok':
             raise self._unwrap_exception(content)
         content = rekey(content)
         if isinstance(targets, int):
             return content[targets]
         else:
             return content
 
     def _build_msgids_from_target(self, targets=None):
         """Build a list of msg_ids from the list of engine targets"""
         if not targets: # needed as _build_targets otherwise uses all engines
             return []
         target_ids = self._build_targets(targets)[0] 
-        return filter(lambda md_id: self.metadata[md_id]["engine_uuid"] in target_ids, self.metadata)
+        return [md_id for md_id in self.metadata if self.metadata[md_id]["engine_uuid"] in target_ids]
     
     def _build_msgids_from_jobs(self, jobs=None):
         """Build a list of msg_ids from "jobs" """
         if not jobs:
             return []
         msg_ids = []
         if isinstance(jobs, string_types + (AsyncResult,)):
             jobs = [jobs]
-        bad_ids = filter(lambda obj: not isinstance(obj, string_types + (AsyncResult)), jobs)
+        bad_ids = [obj for obj in jobs if not isinstance(obj, string_types + (AsyncResult,))]
         if bad_ids:
             raise TypeError("Invalid msg_id type %r, expected str or AsyncResult"%bad_ids[0])
         for j in jobs:
             if isinstance(j, AsyncResult):
                 msg_ids.extend(j.msg_ids)
             else:
                 msg_ids.append(j)
         return msg_ids        
         
     def purge_local_results(self, jobs=[], targets=[]):
         """Clears the client caches of results and frees such memory.
         
         Individual results can be purged by msg_id, or the entire
         history of specific targets can be purged.
 
         Use `purge_local_results('all')` to scrub everything from the Clients's db.
 
         The client must have no outstanding tasks before purging the caches.
         Raises `AssertionError` if there are still outstanding tasks.
 
         After this call all `AsyncResults` are invalid and should be discarded.
 
         If you must "reget" the results, you can still do so by using
         `client.get_result(msg_id)` or `client.get_result(asyncresult)`. This will
         redownload the results from the hub if they are still available
         (i.e `client.purge_hub_results(...)` has not been called.        
 
         Parameters
         ----------
 
         jobs : str or list of str or AsyncResult objects
                 the msg_ids whose results should be purged.
         targets : int/str/list of ints/strs
                 The targets, by int_id, whose entire results are to be purged.
 
                 default : None
         """
         assert not self.outstanding, "Can't purge a client with outstanding tasks!"
         
         if not targets and not jobs:
             raise ValueError("Must specify at least one of `targets` and `jobs`")
                 
         if jobs == 'all':
             self.results.clear()
             self.metadata.clear()
             return
         else:
             msg_ids = []
             msg_ids.extend(self._build_msgids_from_target(targets))
             msg_ids.extend(self._build_msgids_from_jobs(jobs))
-            map(self.results.pop, msg_ids)
-            map(self.metadata.pop, msg_ids)
+            for mid in msg_ids:
+                self.results.pop(mid)
+                self.metadata.pop(mid)
 
 
     @spin_first
     def purge_hub_results(self, jobs=[], targets=[]):
         """Tell the Hub to forget results.
 
         Individual results can be purged by msg_id, or the entire
         history of specific targets can be purged.
 
         Use `purge_results('all')` to scrub everything from the Hub's db.
 
         Parameters
         ----------
 
         jobs : str or list of str or AsyncResult objects
                 the msg_ids whose results should be forgotten.
         targets : int/str/list of ints/strs
                 The targets, by int_id, whose entire history is to be purged.
 
                 default : None
         """
         if not targets and not jobs:
             raise ValueError("Must specify at least one of `targets` and `jobs`")
         if targets:
             targets = self._build_targets(targets)[1]
 
         # construct msg_ids from jobs
         if jobs == 'all':
             msg_ids = jobs
         else:
             msg_ids = self._build_msgids_from_jobs(jobs)
 
         content = dict(engine_ids=targets, msg_ids=msg_ids)
         self.session.send(self._query_socket, "purge_request", content=content)
         idents, msg = self.session.recv(self._query_socket, 0)
         if self.debug:
             pprint(msg)
         content = msg['content']
         if content['status'] != 'ok':
             raise self._unwrap_exception(content)
 
     def purge_results(self,  jobs=[], targets=[]):
         """Clears the cached results from both the hub and the local client
                 
         Individual results can be purged by msg_id, or the entire
         history of specific targets can be purged.
 
         Use `purge_results('all')` to scrub every cached result from both the Hub's and 
         the Client's db.
         
         Equivalent to calling both `purge_hub_results()` and `purge_client_results()` with 
         the same arguments.
 
         Parameters
         ----------
 
         jobs : str or list of str or AsyncResult objects
                 the msg_ids whose results should be forgotten.
         targets : int/str/list of ints/strs
                 The targets, by int_id, whose entire history is to be purged.
 
                 default : None
         """
         self.purge_local_results(jobs=jobs, targets=targets)
         self.purge_hub_results(jobs=jobs, targets=targets)
 
     def purge_everything(self):
         """Clears all content from previous Tasks from both the hub and the local client
         
         In addition to calling `purge_results("all")` it also deletes the history and 
         other bookkeeping lists.        
         """
         self.purge_results("all")
         self.history = []
         self.session.digest_history.clear()
 
     @spin_first
     def hub_history(self):
         """Get the Hub's history
 
         Just like the Client, the Hub has a history, which is a list of msg_ids.
         This will contain the history of all clients, and, depending on configuration,
         may contain history across multiple cluster sessions.
 
         Any msg_id returned here is a valid argument to `get_result`.
 
         Returns
         -------
 
         msg_ids : list of strs
                 list of all msg_ids, ordered by task submission time.
         """
 
         self.session.send(self._query_socket, "history_request", content={})
         idents, msg = self.session.recv(self._query_socket, 0)
 
         if self.debug:
             pprint(msg)
         content = msg['content']
         if content['status'] != 'ok':
             raise self._unwrap_exception(content)
         else:
             return content['history']
 
     @spin_first
     def db_query(self, query, keys=None):
         """Query the Hub's TaskRecord database
 
         This will return a list of task record dicts that match `query`
 
         Parameters
         ----------
 
         query : mongodb query dict
             The search dict. See mongodb query docs for details.
         keys : list of strs [optional]
             The subset of keys to be returned.  The default is to fetch everything but buffers.
             'msg_id' will *always* be included.
         """
         if isinstance(keys, string_types):
             keys = [keys]
         content = dict(query=query, keys=keys)
         self.session.send(self._query_socket, "db_request", content=content)
         idents, msg = self.session.recv(self._query_socket, 0)
         if self.debug:
             pprint(msg)
         content = msg['content']
         if content['status'] != 'ok':
             raise self._unwrap_exception(content)
 
         records = content['records']
 
         buffer_lens = content['buffer_lens']
         result_buffer_lens = content['result_buffer_lens']
         buffers = msg['buffers']
         has_bufs = buffer_lens is not None
         has_rbufs = result_buffer_lens is not None
         for i,rec in enumerate(records):
             # relink buffers
             if has_bufs:
                 blen = buffer_lens[i]
                 rec['buffers'], buffers = buffers[:blen],buffers[blen:]
             if has_rbufs:
                 blen = result_buffer_lens[i]
                 rec['result_buffers'], buffers = buffers[:blen],buffers[blen:]
 
         return records
 
 __all__ = [ 'Client' ]

IPython/parallel/client/view.py

@@ -1,1118 +1,1119 b''
 """Views of remote engines.
 
 Authors:
 
 * Min RK
 """
 from __future__ import print_function
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 import imp
 import sys
 import warnings
 from contextlib import contextmanager
 from types import ModuleType
 
 import zmq
 
 from IPython.testing.skipdoctest import skip_doctest
 from IPython.utils.traitlets import (
     HasTraits, Any, Bool, List, Dict, Set, Instance, CFloat, Integer
 )
 from IPython.external.decorator import decorator
 
 from IPython.parallel import util
 from IPython.parallel.controller.dependency import Dependency, dependent
-from IPython.utils.py3compat import string_types, iteritems
+from IPython.utils.py3compat import string_types, iteritems, PY3
 
 from . import map as Map
 from .asyncresult import AsyncResult, AsyncMapResult
 from .remotefunction import ParallelFunction, parallel, remote, getname
 
 #-----------------------------------------------------------------------------
 # Decorators
 #-----------------------------------------------------------------------------
 
 @decorator
 def save_ids(f, self, *args, **kwargs):
     """Keep our history and outstanding attributes up to date after a method call."""
     n_previous = len(self.client.history)
     try:
         ret = f(self, *args, **kwargs)
     finally:
         nmsgs = len(self.client.history) - n_previous
         msg_ids = self.client.history[-nmsgs:]
         self.history.extend(msg_ids)
-        map(self.outstanding.add, msg_ids)
+        self.outstanding.update(msg_ids)
     return ret
 
 @decorator
 def sync_results(f, self, *args, **kwargs):
     """sync relevant results from self.client to our results attribute."""
     if self._in_sync_results:
         return f(self, *args, **kwargs)
     self._in_sync_results = True
     try:
         ret = f(self, *args, **kwargs)
     finally:
         self._in_sync_results = False
         self._sync_results()
     return ret
 
 @decorator
 def spin_after(f, self, *args, **kwargs):
     """call spin after the method."""
     ret = f(self, *args, **kwargs)
     self.spin()
     return ret
 
 #-----------------------------------------------------------------------------
 # Classes
 #-----------------------------------------------------------------------------
 
 @skip_doctest
 class View(HasTraits):
     """Base View class for more convenint apply(f,*args,**kwargs) syntax via attributes.
 
     Don't use this class, use subclasses.
 
     Methods
     -------
 
     spin
         flushes incoming results and registration state changes
         control methods spin, and requesting `ids` also ensures up to date
 
     wait
         wait on one or more msg_ids
 
     execution methods
         apply
         legacy: execute, run
 
     data movement
         push, pull, scatter, gather
 
     query methods
         get_result, queue_status, purge_results, result_status
 
     control methods
         abort, shutdown
 
     """
     # flags
     block=Bool(False)
     track=Bool(True)
     targets = Any()
 
     history=List()
     outstanding = Set()
     results = Dict()
     client = Instance('IPython.parallel.Client')
 
     _socket = Instance('zmq.Socket')
     _flag_names = List(['targets', 'block', 'track'])
     _in_sync_results = Bool(False)
     _targets = Any()
     _idents = Any()
 
     def __init__(self, client=None, socket=None, **flags):
         super(View, self).__init__(client=client, _socket=socket)
         self.results = client.results
         self.block = client.block
 
         self.set_flags(**flags)
 
         assert not self.__class__ is View, "Don't use base View objects, use subclasses"
 
     def __repr__(self):
         strtargets = str(self.targets)
         if len(strtargets) > 16:
             strtargets = strtargets[:12]+'...]'
         return "<%s %s>"%(self.__class__.__name__, strtargets)
     
     def __len__(self):
         if isinstance(self.targets, list):
             return len(self.targets)
         elif isinstance(self.targets, int):
             return 1
         else:
             return len(self.client)
     
     def set_flags(self, **kwargs):
         """set my attribute flags by keyword.
 
         Views determine behavior with a few attributes (`block`, `track`, etc.).
         These attributes can be set all at once by name with this method.
 
         Parameters
         ----------
 
         block : bool
             whether to wait for results
         track : bool
             whether to create a MessageTracker to allow the user to
             safely edit after arrays and buffers during non-copying
             sends.
         """
         for name, value in iteritems(kwargs):
             if name not in self._flag_names:
                 raise KeyError("Invalid name: %r"%name)
             else:
                 setattr(self, name, value)
 
     @contextmanager
     def temp_flags(self, **kwargs):
         """temporarily set flags, for use in `with` statements.
 
         See set_flags for permanent setting of flags
 
         Examples
         --------
 
         >>> view.track=False
         ...
         >>> with view.temp_flags(track=True):
         ...    ar = view.apply(dostuff, my_big_array)
         ...    ar.tracker.wait() # wait for send to finish
         >>> view.track
         False
 
         """
         # preflight: save flags, and set temporaries
         saved_flags = {}
         for f in self._flag_names:
             saved_flags[f] = getattr(self, f)
         self.set_flags(**kwargs)
         # yield to the with-statement block
         try:
             yield
         finally:
             # postflight: restore saved flags
             self.set_flags(**saved_flags)
 
 
     #----------------------------------------------------------------
     # apply
     #----------------------------------------------------------------
 
     def _sync_results(self):
         """to be called by @sync_results decorator
         
         after submitting any tasks.
         """
         delta = self.outstanding.difference(self.client.outstanding)
         completed = self.outstanding.intersection(delta)
         self.outstanding = self.outstanding.difference(completed)
         
     @sync_results
     @save_ids
     def _really_apply(self, f, args, kwargs, block=None, **options):
         """wrapper for client.send_apply_request"""
         raise NotImplementedError("Implement in subclasses")
 
     def apply(self, f, *args, **kwargs):
         """calls f(*args, **kwargs) on remote engines, returning the result.
 
         This method sets all apply flags via this View's attributes.
 
         if self.block is False:
             returns AsyncResult
         else:
             returns actual result of f(*args, **kwargs)
         """
         return self._really_apply(f, args, kwargs)
 
     def apply_async(self, f, *args, **kwargs):
         """calls f(*args, **kwargs) on remote engines in a nonblocking manner.
 
         returns AsyncResult
         """
         return self._really_apply(f, args, kwargs, block=False)
 
     @spin_after
     def apply_sync(self, f, *args, **kwargs):
         """calls f(*args, **kwargs) on remote engines in a blocking manner,
          returning the result.
 
         returns: actual result of f(*args, **kwargs)
         """
         return self._really_apply(f, args, kwargs, block=True)
 
     #----------------------------------------------------------------
     # wrappers for client and control methods
     #----------------------------------------------------------------
     @sync_results
     def spin(self):
         """spin the client, and sync"""
         self.client.spin()
 
     @sync_results
     def wait(self, jobs=None, timeout=-1):
         """waits on one or more `jobs`, for up to `timeout` seconds.
 
         Parameters
         ----------
 
         jobs : int, str, or list of ints and/or strs, or one or more AsyncResult objects
                 ints are indices to self.history
                 strs are msg_ids
                 default: wait on all outstanding messages
         timeout : float
                 a time in seconds, after which to give up.
                 default is -1, which means no timeout
 
         Returns
         -------
 
         True : when all msg_ids are done
         False : timeout reached, some msg_ids still outstanding
         """
         if jobs is None:
             jobs = self.history
         return self.client.wait(jobs, timeout)
 
     def abort(self, jobs=None, targets=None, block=None):
         """Abort jobs on my engines.
 
         Parameters
         ----------
 
         jobs : None, str, list of strs, optional
             if None: abort all jobs.
             else: abort specific msg_id(s).
         """
         block = block if block is not None else self.block
         targets = targets if targets is not None else self.targets
         jobs = jobs if jobs is not None else list(self.outstanding)
         
         return self.client.abort(jobs=jobs, targets=targets, block=block)
 
     def queue_status(self, targets=None, verbose=False):
         """Fetch the Queue status of my engines"""
         targets = targets if targets is not None else self.targets
         return self.client.queue_status(targets=targets, verbose=verbose)
 
     def purge_results(self, jobs=[], targets=[]):
         """Instruct the controller to forget specific results."""
         if targets is None or targets == 'all':
             targets = self.targets
         return self.client.purge_results(jobs=jobs, targets=targets)
 
     def shutdown(self, targets=None, restart=False, hub=False, block=None):
         """Terminates one or more engine processes, optionally including the hub.
         """
         block = self.block if block is None else block
         if targets is None or targets == 'all':
             targets = self.targets
         return self.client.shutdown(targets=targets, restart=restart, hub=hub, block=block)
 
     @spin_after
     def get_result(self, indices_or_msg_ids=None):
         """return one or more results, specified by history index or msg_id.
 
         See client.get_result for details.
 
         """
 
         if indices_or_msg_ids is None:
             indices_or_msg_ids = -1
         if isinstance(indices_or_msg_ids, int):
             indices_or_msg_ids = self.history[indices_or_msg_ids]
         elif isinstance(indices_or_msg_ids, (list,tuple,set)):
             indices_or_msg_ids = list(indices_or_msg_ids)
             for i,index in enumerate(indices_or_msg_ids):
                 if isinstance(index, int):
                     indices_or_msg_ids[i] = self.history[index]
         return self.client.get_result(indices_or_msg_ids)
 
     #-------------------------------------------------------------------
     # Map
     #-------------------------------------------------------------------
 
     @sync_results
     def map(self, f, *sequences, **kwargs):
         """override in subclasses"""
         raise NotImplementedError
 
     def map_async(self, f, *sequences, **kwargs):
         """Parallel version of builtin `map`, using this view's engines.
 
         This is equivalent to map(...block=False)
 
         See `self.map` for details.
         """
         if 'block' in kwargs:
             raise TypeError("map_async doesn't take a `block` keyword argument.")
         kwargs['block'] = False
         return self.map(f,*sequences,**kwargs)
 
     def map_sync(self, f, *sequences, **kwargs):
         """Parallel version of builtin `map`, using this view's engines.
 
         This is equivalent to map(...block=True)
 
         See `self.map` for details.
         """
         if 'block' in kwargs:
             raise TypeError("map_sync doesn't take a `block` keyword argument.")
         kwargs['block'] = True
         return self.map(f,*sequences,**kwargs)
 
     def imap(self, f, *sequences, **kwargs):
         """Parallel version of `itertools.imap`.
 
         See `self.map` for details.
 
         """
 
         return iter(self.map_async(f,*sequences, **kwargs))
 
     #-------------------------------------------------------------------
     # Decorators
     #-------------------------------------------------------------------
 
     def remote(self, block=None, **flags):
         """Decorator for making a RemoteFunction"""
         block = self.block if block is None else block
         return remote(self, block=block, **flags)
 
     def parallel(self, dist='b', block=None, **flags):
         """Decorator for making a ParallelFunction"""
         block = self.block if block is None else block
         return parallel(self, dist=dist, block=block, **flags)
 
 @skip_doctest
 class DirectView(View):
     """Direct Multiplexer View of one or more engines.
 
     These are created via indexed access to a client:
 
     >>> dv_1 = client[1]
     >>> dv_all = client[:]
     >>> dv_even = client[::2]
     >>> dv_some = client[1:3]
 
     This object provides dictionary access to engine namespaces:
 
     # push a=5:
     >>> dv['a'] = 5
     # pull 'foo':
     >>> db['foo']
 
     """
 
     def __init__(self, client=None, socket=None, targets=None):
         super(DirectView, self).__init__(client=client, socket=socket, targets=targets)
 
     @property
     def importer(self):
         """sync_imports(local=True) as a property.
 
         See sync_imports for details.
 
         """
         return self.sync_imports(True)
 
     @contextmanager
     def sync_imports(self, local=True, quiet=False):
         """Context Manager for performing simultaneous local and remote imports.
 
         'import x as y' will *not* work.  The 'as y' part will simply be ignored.
         
         If `local=True`, then the package will also be imported locally.
         
         If `quiet=True`, no output will be produced when attempting remote 
         imports. 
         
         Note that remote-only (`local=False`) imports have not been implemented.
         
         >>> with view.sync_imports():
         ...    from numpy import recarray
         importing recarray from numpy on engine(s)
 
         """
         from IPython.utils.py3compat import builtin_mod
         local_import = builtin_mod.__import__
         modules = set()
         results = []
         @util.interactive
         def remote_import(name, fromlist, level):
             """the function to be passed to apply, that actually performs the import
             on the engine, and loads up the user namespace.
             """
             import sys
             user_ns = globals()
             mod = __import__(name, fromlist=fromlist, level=level)
             if fromlist:
                 for key in fromlist:
                     user_ns[key] = getattr(mod, key)
             else:
                 user_ns[name] = sys.modules[name]
 
         def view_import(name, globals={}, locals={}, fromlist=[], level=0):
             """the drop-in replacement for __import__, that optionally imports
             locally as well.
             """
             # don't override nested imports
             save_import = builtin_mod.__import__
             builtin_mod.__import__ = local_import
 
             if imp.lock_held():
                 # this is a side-effect import, don't do it remotely, or even
                 # ignore the local effects
                 return local_import(name, globals, locals, fromlist, level)
 
             imp.acquire_lock()
             if local:
                 mod = local_import(name, globals, locals, fromlist, level)
             else:
                 raise NotImplementedError("remote-only imports not yet implemented")
             imp.release_lock()
 
             key = name+':'+','.join(fromlist or [])
             if level <= 0 and key not in modules:
                 modules.add(key)
                 if not quiet:
                     if fromlist:
                         print("importing %s from %s on engine(s)"%(','.join(fromlist), name))
                     else:
                         print("importing %s on engine(s)"%name)
                 results.append(self.apply_async(remote_import, name, fromlist, level))
             # restore override
             builtin_mod.__import__ = save_import
 
             return mod
 
         # override __import__
         builtin_mod.__import__ = view_import
         try:
             # enter the block
             yield
         except ImportError:
             if local:
                 raise
             else:
                 # ignore import errors if not doing local imports
                 pass
         finally:
             # always restore __import__
             builtin_mod.__import__ = local_import
 
         for r in results:
             # raise possible remote ImportErrors here
             r.get()
 
 
     @sync_results
     @save_ids
     def _really_apply(self, f, args=None, kwargs=None, targets=None, block=None, track=None):
         """calls f(*args, **kwargs) on remote engines, returning the result.
 
         This method sets all of `apply`'s flags via this View's attributes.
 
         Parameters
         ----------
 
         f : callable
 
         args : list [default: empty]
 
         kwargs : dict [default: empty]
 
         targets : target list [default: self.targets]
             where to run
         block : bool [default: self.block]
             whether to block
         track : bool [default: self.track]
             whether to ask zmq to track the message, for safe non-copying sends
 
         Returns
         -------
 
         if self.block is False:
             returns AsyncResult
         else:
             returns actual result of f(*args, **kwargs) on the engine(s)
             This will be a list of self.targets is also a list (even length 1), or
             the single result if self.targets is an integer engine id
         """
         args = [] if args is None else args
         kwargs = {} if kwargs is None else kwargs
         block = self.block if block is None else block
         track = self.track if track is None else track
         targets = self.targets if targets is None else targets
         
         _idents, _targets = self.client._build_targets(targets)
         msg_ids = []
         trackers = []
         for ident in _idents:
             msg = self.client.send_apply_request(self._socket, f, args, kwargs, track=track,
                                     ident=ident)
             if track:
                 trackers.append(msg['tracker'])
             msg_ids.append(msg['header']['msg_id'])
         if isinstance(targets, int):
             msg_ids = msg_ids[0]
         tracker = None if track is False else zmq.MessageTracker(*trackers)
         ar = AsyncResult(self.client, msg_ids, fname=getname(f), targets=_targets, tracker=tracker)
         if block:
             try:
                 return ar.get()
             except KeyboardInterrupt:
                 pass
         return ar
 
 
     @sync_results
     def map(self, f, *sequences, **kwargs):
         """view.map(f, *sequences, block=self.block) => list|AsyncMapResult
 
         Parallel version of builtin `map`, using this View's `targets`.
 
         There will be one task per target, so work will be chunked
         if the sequences are longer than `targets`.
 
         Results can be iterated as they are ready, but will become available in chunks.
 
         Parameters
         ----------
 
         f : callable
             function to be mapped
         *sequences: one or more sequences of matching length
             the sequences to be distributed and passed to `f`
         block : bool
             whether to wait for the result or not [default self.block]
 
         Returns
         -------
 
         if block=False:
             AsyncMapResult
                 An object like AsyncResult, but which reassembles the sequence of results
                 into a single list. AsyncMapResults can be iterated through before all
                 results are complete.
         else:
             list
                 the result of map(f,*sequences)
         """
 
         block = kwargs.pop('block', self.block)
         for k in kwargs.keys():
             if k not in ['block', 'track']:
                 raise TypeError("invalid keyword arg, %r"%k)
 
         assert len(sequences) > 0, "must have some sequences to map onto!"
         pf = ParallelFunction(self, f, block=block, **kwargs)
         return pf.map(*sequences)
 
     @sync_results
     @save_ids
     def execute(self, code, silent=True, targets=None, block=None):
         """Executes `code` on `targets` in blocking or nonblocking manner.
 
         ``execute`` is always `bound` (affects engine namespace)
 
         Parameters
         ----------
 
         code : str
                 the code string to be executed
         block : bool
                 whether or not to wait until done to return
                 default: self.block
         """
         block = self.block if block is None else block
         targets = self.targets if targets is None else targets
 
         _idents, _targets = self.client._build_targets(targets)
         msg_ids = []
         trackers = []
         for ident in _idents:
             msg = self.client.send_execute_request(self._socket, code, silent=silent, ident=ident)
             msg_ids.append(msg['header']['msg_id'])
         if isinstance(targets, int):
             msg_ids = msg_ids[0]
         ar = AsyncResult(self.client, msg_ids, fname='execute', targets=_targets)
         if block:
             try:
                 ar.get()
             except KeyboardInterrupt:
                 pass
         return ar
 
     def run(self, filename, targets=None, block=None):
         """Execute contents of `filename` on my engine(s).
 
         This simply reads the contents of the file and calls `execute`.
 
         Parameters
         ----------
 
         filename : str
                 The path to the file
         targets : int/str/list of ints/strs
                 the engines on which to execute
                 default : all
         block : bool
                 whether or not to wait until done
                 default: self.block
 
         """
         with open(filename, 'r') as f:
             # add newline in case of trailing indented whitespace
             # which will cause SyntaxError
             code = f.read()+'\n'
         return self.execute(code, block=block, targets=targets)
 
     def update(self, ns):
         """update remote namespace with dict `ns`
 
         See `push` for details.
         """
         return self.push(ns, block=self.block, track=self.track)
 
     def push(self, ns, targets=None, block=None, track=None):
         """update remote namespace with dict `ns`
 
         Parameters
         ----------
 
         ns : dict
             dict of keys with which to update engine namespace(s)
         block : bool [default : self.block]
             whether to wait to be notified of engine receipt
 
         """
 
         block = block if block is not None else self.block
         track = track if track is not None else self.track
         targets = targets if targets is not None else self.targets
         # applier = self.apply_sync if block else self.apply_async
         if not isinstance(ns, dict):
             raise TypeError("Must be a dict, not %s"%type(ns))
         return self._really_apply(util._push, kwargs=ns, block=block, track=track, targets=targets)
 
     def get(self, key_s):
         """get object(s) by `key_s` from remote namespace
 
         see `pull` for details.
         """
         # block = block if block is not None else self.block
         return self.pull(key_s, block=True)
 
     def pull(self, names, targets=None, block=None):
         """get object(s) by `name` from remote namespace
 
         will return one object if it is a key.
         can also take a list of keys, in which case it will return a list of objects.
         """
         block = block if block is not None else self.block
         targets = targets if targets is not None else self.targets
         applier = self.apply_sync if block else self.apply_async
         if isinstance(names, string_types):
             pass
         elif isinstance(names, (list,tuple,set)):
             for key in names:
                 if not isinstance(key, string_types):
                     raise TypeError("keys must be str, not type %r"%type(key))
         else:
             raise TypeError("names must be strs, not %r"%names)
         return self._really_apply(util._pull, (names,), block=block, targets=targets)
 
     def scatter(self, key, seq, dist='b', flatten=False, targets=None, block=None, track=None):
         """
         Partition a Python sequence and send the partitions to a set of engines.
         """
         block = block if block is not None else self.block
         track = track if track is not None else self.track
         targets = targets if targets is not None else self.targets
         
         # construct integer ID list:
         targets = self.client._build_targets(targets)[1]
 
         mapObject = Map.dists[dist]()
         nparts = len(targets)
         msg_ids = []
         trackers = []
         for index, engineid in enumerate(targets):
             partition = mapObject.getPartition(seq, index, nparts)
             if flatten and len(partition) == 1:
                 ns = {key: partition[0]}
             else:
                 ns = {key: partition}
             r = self.push(ns, block=False, track=track, targets=engineid)
             msg_ids.extend(r.msg_ids)
             if track:
                 trackers.append(r._tracker)
 
         if track:
             tracker = zmq.MessageTracker(*trackers)
         else:
             tracker = None
 
         r = AsyncResult(self.client, msg_ids, fname='scatter', targets=targets, tracker=tracker)
         if block:
             r.wait()
         else:
             return r
 
     @sync_results
     @save_ids
     def gather(self, key, dist='b', targets=None, block=None):
         """
         Gather a partitioned sequence on a set of engines as a single local seq.
         """
         block = block if block is not None else self.block
         targets = targets if targets is not None else self.targets
         mapObject = Map.dists[dist]()
         msg_ids = []
 
         # construct integer ID list:
         targets = self.client._build_targets(targets)[1]
         
         for index, engineid in enumerate(targets):
             msg_ids.extend(self.pull(key, block=False, targets=engineid).msg_ids)
 
         r = AsyncMapResult(self.client, msg_ids, mapObject, fname='gather')
 
         if block:
             try:
                 return r.get()
             except KeyboardInterrupt:
                 pass
         return r
 
     def __getitem__(self, key):
         return self.get(key)
 
     def __setitem__(self,key, value):
         self.update({key:value})
 
     def clear(self, targets=None, block=None):
         """Clear the remote namespaces on my engines."""
         block = block if block is not None else self.block
         targets = targets if targets is not None else self.targets
         return self.client.clear(targets=targets, block=block)
 
     #----------------------------------------
     # activate for %px, %autopx, etc. magics
     #----------------------------------------
 
     def activate(self, suffix=''):
         """Activate IPython magics associated with this View
         
         Defines the magics `%px, %autopx, %pxresult, %%px, %pxconfig`
         
         Parameters
         ----------
         
         suffix: str [default: '']
             The suffix, if any, for the magics.  This allows you to have
             multiple views associated with parallel magics at the same time.
             
             e.g. ``rc[::2].activate(suffix='_even')`` will give you
             the magics ``%px_even``, ``%pxresult_even``, etc. for running magics 
             on the even engines.
         """
         
         from IPython.parallel.client.magics import ParallelMagics
         
         try:
             # This is injected into __builtins__.
             ip = get_ipython()
         except NameError:
             print("The IPython parallel magics (%px, etc.) only work within IPython.")
             return
         
         M = ParallelMagics(ip, self, suffix)
         ip.magics_manager.register(M)
 
 
 @skip_doctest
 class LoadBalancedView(View):
     """An load-balancing View that only executes via the Task scheduler.
 
     Load-balanced views can be created with the client's `view` method:
 
     >>> v = client.load_balanced_view()
 
     or targets can be specified, to restrict the potential destinations:
 
     >>> v = client.client.load_balanced_view([1,3])
 
     which would restrict loadbalancing to between engines 1 and 3.
 
     """
 
     follow=Any()
     after=Any()
     timeout=CFloat()
     retries = Integer(0)
 
     _task_scheme = Any()
     _flag_names = List(['targets', 'block', 'track', 'follow', 'after', 'timeout', 'retries'])
 
     def __init__(self, client=None, socket=None, **flags):
         super(LoadBalancedView, self).__init__(client=client, socket=socket, **flags)
         self._task_scheme=client._task_scheme
 
     def _validate_dependency(self, dep):
         """validate a dependency.
 
         For use in `set_flags`.
         """
         if dep is None or isinstance(dep, string_types + (AsyncResult, Dependency)):
             return True
         elif isinstance(dep, (list,set, tuple)):
             for d in dep:
                 if not isinstance(d, string_types + (AsyncResult,)):
                     return False
         elif isinstance(dep, dict):
             if set(dep.keys()) != set(Dependency().as_dict().keys()):
                 return False
             if not isinstance(dep['msg_ids'], list):
                 return False
             for d in dep['msg_ids']:
                 if not isinstance(d, string_types):
                     return False
         else:
             return False
 
         return True
 
     def _render_dependency(self, dep):
         """helper for building jsonable dependencies from various input forms."""
         if isinstance(dep, Dependency):
             return dep.as_dict()
         elif isinstance(dep, AsyncResult):
             return dep.msg_ids
         elif dep is None:
             return []
         else:
             # pass to Dependency constructor
             return list(Dependency(dep))
 
     def set_flags(self, **kwargs):
         """set my attribute flags by keyword.
 
         A View is a wrapper for the Client's apply method, but with attributes
         that specify keyword arguments, those attributes can be set by keyword
         argument with this method.
 
         Parameters
         ----------
 
         block : bool
             whether to wait for results
         track : bool
             whether to create a MessageTracker to allow the user to
             safely edit after arrays and buffers during non-copying
             sends.
 
         after : Dependency or collection of msg_ids
             Only for load-balanced execution (targets=None)
             Specify a list of msg_ids as a time-based dependency.
             This job will only be run *after* the dependencies
             have been met.
 
         follow : Dependency or collection of msg_ids
             Only for load-balanced execution (targets=None)
             Specify a list of msg_ids as a location-based dependency.
             This job will only be run on an engine where this dependency
             is met.
 
         timeout : float/int or None
             Only for load-balanced execution (targets=None)
             Specify an amount of time (in seconds) for the scheduler to
             wait for dependencies to be met before failing with a
             DependencyTimeout.
 
         retries : int
             Number of times a task will be retried on failure.
         """
 
         super(LoadBalancedView, self).set_flags(**kwargs)
         for name in ('follow', 'after'):
             if name in kwargs:
                 value = kwargs[name]
                 if self._validate_dependency(value):
                     setattr(self, name, value)
                 else:
                     raise ValueError("Invalid dependency: %r"%value)
         if 'timeout' in kwargs:
             t = kwargs['timeout']
-            if not isinstance(t, (int, long, float, type(None))):
+            if not isinstance(t, (int, float, type(None))):
+                if (not PY3) and (not isinstance(t, long)):
                 raise TypeError("Invalid type for timeout: %r"%type(t))
             if t is not None:
                 if t < 0:
                     raise ValueError("Invalid timeout: %s"%t)
             self.timeout = t
 
     @sync_results
     @save_ids
     def _really_apply(self, f, args=None, kwargs=None, block=None, track=None,
                                         after=None, follow=None, timeout=None,
                                         targets=None, retries=None):
         """calls f(*args, **kwargs) on a remote engine, returning the result.
 
         This method temporarily sets all of `apply`'s flags for a single call.
 
         Parameters
         ----------
 
         f : callable
 
         args : list [default: empty]
 
         kwargs : dict [default: empty]
 
         block : bool [default: self.block]
             whether to block
         track : bool [default: self.track]
             whether to ask zmq to track the message, for safe non-copying sends
 
         !!!!!! TODO: THE REST HERE  !!!!
 
         Returns
         -------
 
         if self.block is False:
             returns AsyncResult
         else:
             returns actual result of f(*args, **kwargs) on the engine(s)
             This will be a list of self.targets is also a list (even length 1), or
             the single result if self.targets is an integer engine id
         """
 
         # validate whether we can run
         if self._socket.closed:
             msg = "Task farming is disabled"
             if self._task_scheme == 'pure':
                 msg += " because the pure ZMQ scheduler cannot handle"
                 msg += " disappearing engines."
             raise RuntimeError(msg)
 
         if self._task_scheme == 'pure':
             # pure zmq scheme doesn't support extra features
             msg = "Pure ZMQ scheduler doesn't support the following flags:"
             "follow, after, retries, targets, timeout"
             if (follow or after or retries or targets or timeout):
                 # hard fail on Scheduler flags
                 raise RuntimeError(msg)
             if isinstance(f, dependent):
                 # soft warn on functional dependencies
                 warnings.warn(msg, RuntimeWarning)
 
         # build args
         args = [] if args is None else args
         kwargs = {} if kwargs is None else kwargs
         block = self.block if block is None else block
         track = self.track if track is None else track
         after = self.after if after is None else after
         retries = self.retries if retries is None else retries
         follow = self.follow if follow is None else follow
         timeout = self.timeout if timeout is None else timeout
         targets = self.targets if targets is None else targets
 
         if not isinstance(retries, int):
             raise TypeError('retries must be int, not %r'%type(retries))
 
         if targets is None:
             idents = []
         else:
             idents = self.client._build_targets(targets)[0]
             # ensure *not* bytes
             idents = [ ident.decode() for ident in idents ]
 
         after = self._render_dependency(after)
         follow = self._render_dependency(follow)
         metadata = dict(after=after, follow=follow, timeout=timeout, targets=idents, retries=retries)
 
         msg = self.client.send_apply_request(self._socket, f, args, kwargs, track=track,
                                 metadata=metadata)
         tracker = None if track is False else msg['tracker']
 
         ar = AsyncResult(self.client, msg['header']['msg_id'], fname=getname(f), targets=None, tracker=tracker)
 
         if block:
             try:
                 return ar.get()
             except KeyboardInterrupt:
                 pass
         return ar
 
     @sync_results
     @save_ids
     def map(self, f, *sequences, **kwargs):
         """view.map(f, *sequences, block=self.block, chunksize=1, ordered=True) => list|AsyncMapResult
 
         Parallel version of builtin `map`, load-balanced by this View.
 
         `block`, and `chunksize` can be specified by keyword only.
 
         Each `chunksize` elements will be a separate task, and will be
         load-balanced. This lets individual elements be available for iteration
         as soon as they arrive.
 
         Parameters
         ----------
 
         f : callable
             function to be mapped
         *sequences: one or more sequences of matching length
             the sequences to be distributed and passed to `f`
         block : bool [default self.block]
             whether to wait for the result or not
         track : bool
             whether to create a MessageTracker to allow the user to
             safely edit after arrays and buffers during non-copying
             sends.
         chunksize : int [default 1]
             how many elements should be in each task.
         ordered : bool [default True]
             Whether the results should be gathered as they arrive, or enforce
             the order of submission.
             
             Only applies when iterating through AsyncMapResult as results arrive.
             Has no effect when block=True.
 
         Returns
         -------
 
         if block=False:
             AsyncMapResult
                 An object like AsyncResult, but which reassembles the sequence of results
                 into a single list. AsyncMapResults can be iterated through before all
                 results are complete.
             else:
                 the result of map(f,*sequences)
 
         """
 
         # default
         block = kwargs.get('block', self.block)
         chunksize = kwargs.get('chunksize', 1)
         ordered = kwargs.get('ordered', True)
 
         keyset = set(kwargs.keys())
         extra_keys = keyset.difference_update(set(['block', 'chunksize']))
         if extra_keys:
             raise TypeError("Invalid kwargs: %s"%list(extra_keys))
 
         assert len(sequences) > 0, "must have some sequences to map onto!"
 
         pf = ParallelFunction(self, f, block=block, chunksize=chunksize, ordered=ordered)
         return pf.map(*sequences)
 
 __all__ = ['LoadBalancedView', 'DirectView']

IPython/parallel/controller/dependency.py

@@ -1,226 +1,230 b''
 """Dependency utilities
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2013  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 from types import ModuleType
 
 from IPython.parallel.client.asyncresult import AsyncResult
 from IPython.parallel.error import UnmetDependency
 from IPython.parallel.util import interactive
 from IPython.utils import py3compat
 from IPython.utils.py3compat import string_types
 from IPython.utils.pickleutil import can, uncan
 
 class depend(object):
     """Dependency decorator, for use with tasks.
     
     `@depend` lets you define a function for engine dependencies
     just like you use `apply` for tasks.
     
     
     Examples
     --------
     ::
     
         @depend(df, a,b, c=5)
         def f(m,n,p)
         
         view.apply(f, 1,2,3)
     
     will call df(a,b,c=5) on the engine, and if it returns False or
     raises an UnmetDependency error, then the task will not be run
     and another engine will be tried.
     """
     def __init__(self, f, *args, **kwargs):
         self.f = f
         self.args = args
         self.kwargs = kwargs
     
     def __call__(self, f):
         return dependent(f, self.f, *self.args, **self.kwargs)
 
 class dependent(object):
     """A function that depends on another function.
     This is an object to prevent the closure used
     in traditional decorators, which are not picklable.
     """
 
     def __init__(self, f, df, *dargs, **dkwargs):
         self.f = f
-        self.__name__ = getattr(f, '__name__', 'f')
+        name = getattr(f, '__name__', 'f')
+        if py3compat.PY3:
+            self.__name__ = name
+        else:
+            self.func_name = name
         self.df = df
         self.dargs = dargs
         self.dkwargs = dkwargs
     
     def check_dependency(self):
         if self.df(*self.dargs, **self.dkwargs) is False:
             raise UnmetDependency()
     
     def __call__(self, *args, **kwargs):
         return self.f(*args, **kwargs)
     
     if not py3compat.PY3:
         @property
         def __name__(self):
             return self.__name__
 
 @interactive
 def _require(*modules, **mapping):
     """Helper for @require decorator."""
     from IPython.parallel.error import UnmetDependency
     from IPython.utils.pickleutil import uncan
     user_ns = globals()
     for name in modules:
         try:
             exec('import %s' % name, user_ns)
         except ImportError:
             raise UnmetDependency(name)
             
     for name, cobj in mapping.items():
         user_ns[name] = uncan(cobj, user_ns)
     return True
 
 def require(*objects, **mapping):
     """Simple decorator for requiring local objects and modules to be available
     when the decorated function is called on the engine.
     
     Modules specified by name or passed directly will be imported
     prior to calling the decorated function.
     
     Objects other than modules will be pushed as a part of the task.
     Functions can be passed positionally,
     and will be pushed to the engine with their __name__.
     Other objects can be passed by keyword arg.
     
     Examples
     --------
     
     In [1]: @require('numpy')
        ...: def norm(a):
        ...:     return numpy.linalg.norm(a,2)
 
     In [2]: foo = lambda x: x*x
     In [3]: @require(foo)
        ...: def bar(a):
        ...:     return foo(1-a)
     """
     names = []
     for obj in objects:
         if isinstance(obj, ModuleType):
             obj = obj.__name__
         
         if isinstance(obj, string_types):
             names.append(obj)
         elif hasattr(obj, '__name__'):
             mapping[obj.__name__] = obj
         else:
             raise TypeError("Objects other than modules and functions "
                 "must be passed by kwarg, but got: %s" % type(obj)
             )
     
     for name, obj in mapping.items():
         mapping[name] = can(obj)
     return depend(_require, *names, **mapping)
 
 class Dependency(set):
     """An object for representing a set of msg_id dependencies.
     
     Subclassed from set().
     
     Parameters
     ----------
     dependencies: list/set of msg_ids or AsyncResult objects or output of Dependency.as_dict()
         The msg_ids to depend on
     all : bool [default True]
         Whether the dependency should be considered met when *all* depending tasks have completed
         or only when *any* have been completed.
     success : bool [default True]
         Whether to consider successes as fulfilling dependencies.
     failure : bool [default False]
         Whether to consider failures as fulfilling dependencies.
     
     If `all=success=True` and `failure=False`, then the task will fail with an ImpossibleDependency
         as soon as the first depended-upon task fails.
     """
     
     all=True
     success=True
     failure=True
     
     def __init__(self, dependencies=[], all=True, success=True, failure=False):
         if isinstance(dependencies, dict):
             # load from dict
             all = dependencies.get('all', True)
             success = dependencies.get('success', success)
             failure = dependencies.get('failure', failure)
             dependencies = dependencies.get('dependencies', [])
         ids = []
         
         # extract ids from various sources:
         if isinstance(dependencies, string_types + (AsyncResult,)):
             dependencies = [dependencies]
         for d in dependencies:
             if isinstance(d, string_types):
                 ids.append(d)
             elif isinstance(d, AsyncResult):
                 ids.extend(d.msg_ids)
             else:
                 raise TypeError("invalid dependency type: %r"%type(d))
         
         set.__init__(self, ids)
         self.all = all
         if not (success or failure):
             raise ValueError("Must depend on at least one of successes or failures!")
         self.success=success
         self.failure = failure
     
     def check(self, completed, failed=None):
         """check whether our dependencies have been met."""
         if len(self) == 0:
             return True
         against = set()
         if self.success:
             against = completed
         if failed is not None and self.failure:
             against = against.union(failed)
         if self.all:
             return self.issubset(against)
         else:
             return not self.isdisjoint(against)
     
     def unreachable(self, completed, failed=None):
         """return whether this dependency has become impossible."""
         if len(self) == 0:
             return False
         against = set()
         if not self.success:
             against = completed
         if failed is not None and not self.failure:
             against = against.union(failed)
         if self.all:
             return not self.isdisjoint(against)
         else:
             return self.issubset(against)
         
     
     def as_dict(self):
         """Represent this dependency as a dict. For json compatibility."""
         return dict(
             dependencies=list(self),
             all=self.all,
             success=self.success,
             failure=self.failure
         )
 
 
 __all__ = ['depend', 'require', 'dependent', 'Dependency']
 

IPython/parallel/controller/heartmonitor.py

@@ -1,190 +1,192 b''
 #!/usr/bin/env python
 """
 A multi-heart Heartbeat system using PUB and ROUTER sockets. pings are sent out on the PUB,
 and hearts are tracked based on their DEALER identities.
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 from __future__ import print_function
 import time
 import uuid
 
 import zmq
 from zmq.devices import ThreadDevice, ThreadMonitoredQueue
 from zmq.eventloop import ioloop, zmqstream
 
 from IPython.config.configurable import LoggingConfigurable
 from IPython.utils.py3compat import str_to_bytes
 from IPython.utils.traitlets import Set, Instance, CFloat, Integer, Dict
 
 from IPython.parallel.util import log_errors
 
 class Heart(object):
     """A basic heart object for responding to a HeartMonitor.
     This is a simple wrapper with defaults for the most common
     Device model for responding to heartbeats.
 
     It simply builds a threadsafe zmq.FORWARDER Device, defaulting to using
     SUB/DEALER for in/out.
 
     You can specify the DEALER's IDENTITY via the optional heart_id argument."""
     device=None
     id=None
     def __init__(self, in_addr, out_addr, mon_addr=None, in_type=zmq.SUB, out_type=zmq.DEALER, mon_type=zmq.PUB, heart_id=None):
         if mon_addr is None:
             self.device = ThreadDevice(zmq.FORWARDER, in_type, out_type)
         else:
             self.device = ThreadMonitoredQueue(in_type, out_type, mon_type, in_prefix=b"", out_prefix=b"")
         # do not allow the device to share global Context.instance,
         # which is the default behavior in pyzmq > 2.1.10
         self.device.context_factory = zmq.Context
         
         self.device.daemon=True
         self.device.connect_in(in_addr)
         self.device.connect_out(out_addr)
         if mon_addr is not None:
             self.device.connect_mon(mon_addr)
         if in_type == zmq.SUB:
             self.device.setsockopt_in(zmq.SUBSCRIBE, b"")
         if heart_id is None:
             heart_id = uuid.uuid4().bytes
         self.device.setsockopt_out(zmq.IDENTITY, heart_id)
         self.id = heart_id
 
     def start(self):
         return self.device.start()
 
 
 class HeartMonitor(LoggingConfigurable):
     """A basic HeartMonitor class
     pingstream: a PUB stream
     pongstream: an ROUTER stream
     period: the period of the heartbeat in milliseconds"""
 
     period = Integer(3000, config=True,
         help='The frequency at which the Hub pings the engines for heartbeats '
         '(in ms)',
     )
     max_heartmonitor_misses = Integer(10, config=True,
         help='Allowed consecutive missed pings from controller Hub to engine before unregistering.',
     )
 
     pingstream=Instance('zmq.eventloop.zmqstream.ZMQStream')
     pongstream=Instance('zmq.eventloop.zmqstream.ZMQStream')
     loop = Instance('zmq.eventloop.ioloop.IOLoop')
     def _loop_default(self):
         return ioloop.IOLoop.instance()
 
     # not settable:
     hearts=Set()
     responses=Set()
     on_probation=Dict()
     last_ping=CFloat(0)
     _new_handlers = Set()
     _failure_handlers = Set()
     lifetime = CFloat(0)
     tic = CFloat(0)
 
     def __init__(self, **kwargs):
         super(HeartMonitor, self).__init__(**kwargs)
 
         self.pongstream.on_recv(self.handle_pong)
 
     def start(self):
         self.tic = time.time()
         self.caller = ioloop.PeriodicCallback(self.beat, self.period, self.loop)
         self.caller.start()
 
     def add_new_heart_handler(self, handler):
         """add a new handler for new hearts"""
         self.log.debug("heartbeat::new_heart_handler: %s", handler)
         self._new_handlers.add(handler)
 
     def add_heart_failure_handler(self, handler):
         """add a new handler for heart failure"""
         self.log.debug("heartbeat::new heart failure handler: %s", handler)
         self._failure_handlers.add(handler)
 
     def beat(self):
         self.pongstream.flush()
         self.last_ping = self.lifetime
 
         toc = time.time()
         self.lifetime += toc-self.tic
         self.tic = toc
         self.log.debug("heartbeat::sending %s", self.lifetime)
         goodhearts = self.hearts.intersection(self.responses)
         missed_beats = self.hearts.difference(goodhearts)
         newhearts = self.responses.difference(goodhearts)
-        map(self.handle_new_heart, newhearts)
+        for heart in newhearts:
+            self.handle_new_heart(heart)
         heartfailures, on_probation = self._check_missed(missed_beats, self.on_probation,
                                                          self.hearts)
-        map(self.handle_heart_failure, heartfailures)
+        for failure in heartfailures:
+            self.handle_heart_failure(failure)
         self.on_probation = on_probation
         self.responses = set()
         #print self.on_probation, self.hearts
         # self.log.debug("heartbeat::beat %.3f, %i beating hearts", self.lifetime, len(self.hearts))
         self.pingstream.send(str_to_bytes(str(self.lifetime)))
         # flush stream to force immediate socket send
         self.pingstream.flush()
 
     def _check_missed(self, missed_beats, on_probation, hearts):
         """Update heartbeats on probation, identifying any that have too many misses.
         """
         failures = []
         new_probation = {}
         for cur_heart in (b for b in missed_beats if b in hearts):
             miss_count = on_probation.get(cur_heart, 0) + 1
             self.log.info("heartbeat::missed %s : %s" % (cur_heart, miss_count))
             if miss_count > self.max_heartmonitor_misses:
                 failures.append(cur_heart)
             else:
                 new_probation[cur_heart] = miss_count
         return failures, new_probation
 
     def handle_new_heart(self, heart):
         if self._new_handlers:
             for handler in self._new_handlers:
                 handler(heart)
         else:
             self.log.info("heartbeat::yay, got new heart %s!", heart)
         self.hearts.add(heart)
 
     def handle_heart_failure(self, heart):
         if self._failure_handlers:
             for handler in self._failure_handlers:
                 try:
                     handler(heart)
                 except Exception as e:
                     self.log.error("heartbeat::Bad Handler! %s", handler, exc_info=True)
                     pass
         else:
             self.log.info("heartbeat::Heart %s failed :(", heart)
         self.hearts.remove(heart)
 
 
     @log_errors
     def handle_pong(self, msg):
         "a heart just beat"
         current = str_to_bytes(str(self.lifetime))
         last = str_to_bytes(str(self.last_ping))
         if msg[1] == current:
             delta = time.time()-self.tic
             # self.log.debug("heartbeat::heart %r took %.2f ms to respond"%(msg[0], 1000*delta))
             self.responses.add(msg[0])
         elif msg[1] == last:
             delta = time.time()-self.tic + (self.lifetime-self.last_ping)
             self.log.warn("heartbeat::heart %r missed a beat, and took %.2f ms to respond", msg[0], 1000*delta)
             self.responses.add(msg[0])
         else:
             self.log.warn("heartbeat::got bad heartbeat (possibly old?): %s (current=%.3f)", msg[1], self.lifetime)
 

IPython/parallel/controller/hub.py

@@ -1,1421 +1,1421 b''
 """The IPython Controller Hub with 0MQ
 This is the master object that handles connections from engines and clients,
 and monitors traffic through the various queues.
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 from __future__ import print_function
 
 import json
 import os
 import sys
 import time
 from datetime import datetime
 
 import zmq
 from zmq.eventloop import ioloop
 from zmq.eventloop.zmqstream import ZMQStream
 
 # internal:
 from IPython.utils.importstring import import_item
 from IPython.utils.localinterfaces import localhost
 from IPython.utils.py3compat import cast_bytes, unicode_type, iteritems
 from IPython.utils.traitlets import (
         HasTraits, Instance, Integer, Unicode, Dict, Set, Tuple, CBytes, DottedObjectName
         )
 
 from IPython.parallel import error, util
 from IPython.parallel.factory import RegistrationFactory
 
 from IPython.kernel.zmq.session import SessionFactory
 
 from .heartmonitor import HeartMonitor
 
 #-----------------------------------------------------------------------------
 # Code
 #-----------------------------------------------------------------------------
 
 def _passer(*args, **kwargs):
     return
 
 def _printer(*args, **kwargs):
     print (args)
     print (kwargs)
 
 def empty_record():
     """Return an empty dict with all record keys."""
     return {
         'msg_id' : None,
         'header' : None,
         'metadata' : None,
         'content': None,
         'buffers': None,
         'submitted': None,
         'client_uuid' : None,
         'engine_uuid' : None,
         'started': None,
         'completed': None,
         'resubmitted': None,
         'received': None,
         'result_header' : None,
         'result_metadata' : None,
         'result_content' : None,
         'result_buffers' : None,
         'queue' : None,
         'pyin' : None,
         'pyout': None,
         'pyerr': None,
         'stdout': '',
         'stderr': '',
     }
 
 def init_record(msg):
     """Initialize a TaskRecord based on a request."""
     header = msg['header']
     return {
         'msg_id' : header['msg_id'],
         'header' : header,
         'content': msg['content'],
         'metadata': msg['metadata'],
         'buffers': msg['buffers'],
         'submitted': header['date'],
         'client_uuid' : None,
         'engine_uuid' : None,
         'started': None,
         'completed': None,
         'resubmitted': None,
         'received': None,
         'result_header' : None,
         'result_metadata': None,
         'result_content' : None,
         'result_buffers' : None,
         'queue' : None,
         'pyin' : None,
         'pyout': None,
         'pyerr': None,
         'stdout': '',
         'stderr': '',
     }
 
 
 class EngineConnector(HasTraits):
     """A simple object for accessing the various zmq connections of an object.
     Attributes are:
     id (int): engine ID
     uuid (unicode): engine UUID
     pending: set of msg_ids
     stallback: DelayedCallback for stalled registration
     """
     
     id = Integer(0)
     uuid = Unicode()
     pending = Set()
     stallback = Instance(ioloop.DelayedCallback)
 
 
 _db_shortcuts = {
     'sqlitedb' : 'IPython.parallel.controller.sqlitedb.SQLiteDB',
     'mongodb'  : 'IPython.parallel.controller.mongodb.MongoDB',
     'dictdb'   : 'IPython.parallel.controller.dictdb.DictDB',
     'nodb'     : 'IPython.parallel.controller.dictdb.NoDB',
 }
 
 class HubFactory(RegistrationFactory):
     """The Configurable for setting up a Hub."""
 
     # port-pairs for monitoredqueues:
     hb = Tuple(Integer,Integer,config=True,
         help="""PUB/ROUTER Port pair for Engine heartbeats""")
     def _hb_default(self):
         return tuple(util.select_random_ports(2))
 
     mux = Tuple(Integer,Integer,config=True,
         help="""Client/Engine Port pair for MUX queue""")
 
     def _mux_default(self):
         return tuple(util.select_random_ports(2))
 
     task = Tuple(Integer,Integer,config=True,
         help="""Client/Engine Port pair for Task queue""")
     def _task_default(self):
         return tuple(util.select_random_ports(2))
 
     control = Tuple(Integer,Integer,config=True,
         help="""Client/Engine Port pair for Control queue""")
 
     def _control_default(self):
         return tuple(util.select_random_ports(2))
 
     iopub = Tuple(Integer,Integer,config=True,
         help="""Client/Engine Port pair for IOPub relay""")
 
     def _iopub_default(self):
         return tuple(util.select_random_ports(2))
 
     # single ports:
     mon_port = Integer(config=True,
         help="""Monitor (SUB) port for queue traffic""")
 
     def _mon_port_default(self):
         return util.select_random_ports(1)[0]
 
     notifier_port = Integer(config=True,
         help="""PUB port for sending engine status notifications""")
 
     def _notifier_port_default(self):
         return util.select_random_ports(1)[0]
 
     engine_ip = Unicode(config=True,
         help="IP on which to listen for engine connections. [default: loopback]")
     def _engine_ip_default(self):
         return localhost()
     engine_transport = Unicode('tcp', config=True,
         help="0MQ transport for engine connections. [default: tcp]")
 
     client_ip = Unicode(config=True,
         help="IP on which to listen for client connections. [default: loopback]")
     client_transport = Unicode('tcp', config=True,
         help="0MQ transport for client connections. [default : tcp]")
 
     monitor_ip = Unicode(config=True,
         help="IP on which to listen for monitor messages. [default: loopback]")
     monitor_transport = Unicode('tcp', config=True,
         help="0MQ transport for monitor messages. [default : tcp]")
     
     _client_ip_default = _monitor_ip_default = _engine_ip_default
     
 
     monitor_url = Unicode('')
 
     db_class = DottedObjectName('NoDB',
         config=True, help="""The class to use for the DB backend
         
         Options include:
         
         SQLiteDB: SQLite
         MongoDB : use MongoDB
         DictDB  : in-memory storage (fastest, but be mindful of memory growth of the Hub)
         NoDB    : disable database altogether (default)
         
         """)
 
     # not configurable
     db = Instance('IPython.parallel.controller.dictdb.BaseDB')
     heartmonitor = Instance('IPython.parallel.controller.heartmonitor.HeartMonitor')
 
     def _ip_changed(self, name, old, new):
         self.engine_ip = new
         self.client_ip = new
         self.monitor_ip = new
         self._update_monitor_url()
 
     def _update_monitor_url(self):
         self.monitor_url = "%s://%s:%i" % (self.monitor_transport, self.monitor_ip, self.mon_port)
 
     def _transport_changed(self, name, old, new):
         self.engine_transport = new
         self.client_transport = new
         self.monitor_transport = new
         self._update_monitor_url()
 
     def __init__(self, **kwargs):
         super(HubFactory, self).__init__(**kwargs)
         self._update_monitor_url()
 
 
     def construct(self):
         self.init_hub()
 
     def start(self):
         self.heartmonitor.start()
         self.log.info("Heartmonitor started")
 
     def client_url(self, channel):
         """return full zmq url for a named client channel"""
         return "%s://%s:%i" % (self.client_transport, self.client_ip, self.client_info[channel])
     
     def engine_url(self, channel):
         """return full zmq url for a named engine channel"""
         return "%s://%s:%i" % (self.engine_transport, self.engine_ip, self.engine_info[channel])
     
     def init_hub(self):
         """construct Hub object"""
 
         ctx = self.context
         loop = self.loop
         if 'TaskScheduler.scheme_name' in self.config:
             scheme = self.config.TaskScheduler.scheme_name
         else:
             from .scheduler import TaskScheduler
             scheme = TaskScheduler.scheme_name.get_default_value()
         
         # build connection dicts
         engine = self.engine_info = {
             'interface'     : "%s://%s" % (self.engine_transport, self.engine_ip),
             'registration'  : self.regport,
             'control'       : self.control[1],
             'mux'           : self.mux[1],
             'hb_ping'       : self.hb[0],
             'hb_pong'       : self.hb[1],
             'task'          : self.task[1],
             'iopub'         : self.iopub[1],
             }
 
         client = self.client_info = {
             'interface'     : "%s://%s" % (self.client_transport, self.client_ip),
             'registration'  : self.regport,
             'control'       : self.control[0],
             'mux'           : self.mux[0],
             'task'          : self.task[0],
             'task_scheme'   : scheme,
             'iopub'         : self.iopub[0],
             'notification'  : self.notifier_port,
             }
         
         self.log.debug("Hub engine addrs: %s", self.engine_info)
         self.log.debug("Hub client addrs: %s", self.client_info)
         
         # Registrar socket
         q = ZMQStream(ctx.socket(zmq.ROUTER), loop)
         util.set_hwm(q, 0)
         q.bind(self.client_url('registration'))
         self.log.info("Hub listening on %s for registration.", self.client_url('registration'))
         if self.client_ip != self.engine_ip:
             q.bind(self.engine_url('registration'))
             self.log.info("Hub listening on %s for registration.", self.engine_url('registration'))
 
         ### Engine connections ###
 
         # heartbeat
         hpub = ctx.socket(zmq.PUB)
         hpub.bind(self.engine_url('hb_ping'))
         hrep = ctx.socket(zmq.ROUTER)
         util.set_hwm(hrep, 0)
         hrep.bind(self.engine_url('hb_pong'))
         self.heartmonitor = HeartMonitor(loop=loop, parent=self, log=self.log,
                                 pingstream=ZMQStream(hpub,loop),
                                 pongstream=ZMQStream(hrep,loop)
                             )
 
         ### Client connections ###
         
         # Notifier socket
         n = ZMQStream(ctx.socket(zmq.PUB), loop)
         n.bind(self.client_url('notification'))
 
         ### build and launch the queues ###
 
         # monitor socket
         sub = ctx.socket(zmq.SUB)
         sub.setsockopt(zmq.SUBSCRIBE, b"")
         sub.bind(self.monitor_url)
         sub.bind('inproc://monitor')
         sub = ZMQStream(sub, loop)
 
         # connect the db
         db_class = _db_shortcuts.get(self.db_class.lower(), self.db_class)
         self.log.info('Hub using DB backend: %r', (db_class.split('.')[-1]))
         self.db = import_item(str(db_class))(session=self.session.session,
                                             parent=self, log=self.log)
         time.sleep(.25)
 
         # resubmit stream
         r = ZMQStream(ctx.socket(zmq.DEALER), loop)
         url = util.disambiguate_url(self.client_url('task'))
         r.connect(url)
 
         self.hub = Hub(loop=loop, session=self.session, monitor=sub, heartmonitor=self.heartmonitor,
                 query=q, notifier=n, resubmit=r, db=self.db,
                 engine_info=self.engine_info, client_info=self.client_info,
                 log=self.log)
 
 
 class Hub(SessionFactory):
     """The IPython Controller Hub with 0MQ connections
 
     Parameters
     ==========
     loop: zmq IOLoop instance
     session: Session object
     <removed> context: zmq context for creating new connections (?)
     queue: ZMQStream for monitoring the command queue (SUB)
     query: ZMQStream for engine registration and client queries requests (ROUTER)
     heartbeat: HeartMonitor object checking the pulse of the engines
     notifier: ZMQStream for broadcasting engine registration changes (PUB)
     db: connection to db for out of memory logging of commands
                 NotImplemented
     engine_info: dict of zmq connection information for engines to connect
                 to the queues.
     client_info: dict of zmq connection information for engines to connect
                 to the queues.
     """
     
     engine_state_file = Unicode()
     
     # internal data structures:
     ids=Set() # engine IDs
     keytable=Dict()
     by_ident=Dict()
     engines=Dict()
     clients=Dict()
     hearts=Dict()
     pending=Set()
     queues=Dict()  # pending msg_ids keyed by engine_id
     tasks=Dict() # pending msg_ids submitted as tasks, keyed by client_id
     completed=Dict() # completed msg_ids keyed by engine_id
     all_completed=Set() # completed msg_ids keyed by engine_id
     dead_engines=Set() # completed msg_ids keyed by engine_id
     unassigned=Set() # set of task msg_ds not yet assigned a destination
     incoming_registrations=Dict()
     registration_timeout=Integer()
     _idcounter=Integer(0)
 
     # objects from constructor:
     query=Instance(ZMQStream)
     monitor=Instance(ZMQStream)
     notifier=Instance(ZMQStream)
     resubmit=Instance(ZMQStream)
     heartmonitor=Instance(HeartMonitor)
     db=Instance(object)
     client_info=Dict()
     engine_info=Dict()
 
 
     def __init__(self, **kwargs):
         """
         # universal:
         loop: IOLoop for creating future connections
         session: streamsession for sending serialized data
         # engine:
         queue: ZMQStream for monitoring queue messages
         query: ZMQStream for engine+client registration and client requests
         heartbeat: HeartMonitor object for tracking engines
         # extra:
         db: ZMQStream for db connection (NotImplemented)
         engine_info: zmq address/protocol dict for engine connections
         client_info: zmq address/protocol dict for client connections
         """
 
         super(Hub, self).__init__(**kwargs)
         self.registration_timeout = max(10000, 5*self.heartmonitor.period)
 
         # register our callbacks
         self.query.on_recv(self.dispatch_query)
         self.monitor.on_recv(self.dispatch_monitor_traffic)
 
         self.heartmonitor.add_heart_failure_handler(self.handle_heart_failure)
         self.heartmonitor.add_new_heart_handler(self.handle_new_heart)
 
         self.monitor_handlers = {b'in' : self.save_queue_request,
                                 b'out': self.save_queue_result,
                                 b'intask': self.save_task_request,
                                 b'outtask': self.save_task_result,
                                 b'tracktask': self.save_task_destination,
                                 b'incontrol': _passer,
                                 b'outcontrol': _passer,
                                 b'iopub': self.save_iopub_message,
         }
 
         self.query_handlers = {'queue_request': self.queue_status,
                                 'result_request': self.get_results,
                                 'history_request': self.get_history,
                                 'db_request': self.db_query,
                                 'purge_request': self.purge_results,
                                 'load_request': self.check_load,
                                 'resubmit_request': self.resubmit_task,
                                 'shutdown_request': self.shutdown_request,
                                 'registration_request' : self.register_engine,
                                 'unregistration_request' : self.unregister_engine,
                                 'connection_request': self.connection_request,
         }
 
         # ignore resubmit replies
         self.resubmit.on_recv(lambda msg: None, copy=False)
 
         self.log.info("hub::created hub")
     
     @property
     def _next_id(self):
         """gemerate a new ID.
 
         No longer reuse old ids, just count from 0."""
         newid = self._idcounter
         self._idcounter += 1
         return newid
         # newid = 0
         # incoming = [id[0] for id in itervalues(self.incoming_registrations)]
         # # print newid, self.ids, self.incoming_registrations
         # while newid in self.ids or newid in incoming:
         #     newid += 1
         # return newid
     
     #-----------------------------------------------------------------------------
     # message validation
     #-----------------------------------------------------------------------------
 
     def _validate_targets(self, targets):
         """turn any valid targets argument into a list of integer ids"""
         if targets is None:
             # default to all
             return self.ids
 
         if isinstance(targets, (int,str,unicode_type)):
             # only one target specified
             targets = [targets]
         _targets = []
         for t in targets:
             # map raw identities to ids
             if isinstance(t, (str,unicode_type)):
                 t = self.by_ident.get(cast_bytes(t), t)
             _targets.append(t)
         targets = _targets
         bad_targets = [ t for t in targets if t not in self.ids ]
         if bad_targets:
             raise IndexError("No Such Engine: %r" % bad_targets)
         if not targets:
             raise IndexError("No Engines Registered")
         return targets
 
     #-----------------------------------------------------------------------------
     # dispatch methods (1 per stream)
     #-----------------------------------------------------------------------------
 
 
     @util.log_errors
     def dispatch_monitor_traffic(self, msg):
         """all ME and Task queue messages come through here, as well as
         IOPub traffic."""
         self.log.debug("monitor traffic: %r", msg[0])
         switch = msg[0]
         try:
             idents, msg = self.session.feed_identities(msg[1:])
         except ValueError:
             idents=[]
         if not idents:
             self.log.error("Monitor message without topic: %r", msg)
             return
         handler = self.monitor_handlers.get(switch, None)
         if handler is not None:
             handler(idents, msg)
         else:
             self.log.error("Unrecognized monitor topic: %r", switch)
 
 
     @util.log_errors
     def dispatch_query(self, msg):
         """Route registration requests and queries from clients."""
         try:
             idents, msg = self.session.feed_identities(msg)
         except ValueError:
             idents = []
         if not idents:
             self.log.error("Bad Query Message: %r", msg)
             return
         client_id = idents[0]
         try:
             msg = self.session.unserialize(msg, content=True)
         except Exception:
             content = error.wrap_exception()
             self.log.error("Bad Query Message: %r", msg, exc_info=True)
             self.session.send(self.query, "hub_error", ident=client_id,
                     content=content)
             return
         # print client_id, header, parent, content
         #switch on message type:
         msg_type = msg['header']['msg_type']
         self.log.info("client::client %r requested %r", client_id, msg_type)
         handler = self.query_handlers.get(msg_type, None)
         try:
             assert handler is not None, "Bad Message Type: %r" % msg_type
         except:
             content = error.wrap_exception()
             self.log.error("Bad Message Type: %r", msg_type, exc_info=True)
             self.session.send(self.query, "hub_error", ident=client_id,
                     content=content)
             return
 
         else:
             handler(idents, msg)
 
     def dispatch_db(self, msg):
         """"""
         raise NotImplementedError
 
     #---------------------------------------------------------------------------
     # handler methods (1 per event)
     #---------------------------------------------------------------------------
 
     #----------------------- Heartbeat --------------------------------------
 
     def handle_new_heart(self, heart):
         """handler to attach to heartbeater.
         Called when a new heart starts to beat.
         Triggers completion of registration."""
         self.log.debug("heartbeat::handle_new_heart(%r)", heart)
         if heart not in self.incoming_registrations:
             self.log.info("heartbeat::ignoring new heart: %r", heart)
         else:
             self.finish_registration(heart)
 
 
     def handle_heart_failure(self, heart):
         """handler to attach to heartbeater.
         called when a previously registered heart fails to respond to beat request.
         triggers unregistration"""
         self.log.debug("heartbeat::handle_heart_failure(%r)", heart)
         eid = self.hearts.get(heart, None)
         uuid = self.engines[eid].uuid
         if eid is None or self.keytable[eid] in self.dead_engines:
             self.log.info("heartbeat::ignoring heart failure %r (not an engine or already dead)", heart)
         else:
             self.unregister_engine(heart, dict(content=dict(id=eid, queue=uuid)))
 
     #----------------------- MUX Queue Traffic ------------------------------
 
     def save_queue_request(self, idents, msg):
         if len(idents) < 2:
             self.log.error("invalid identity prefix: %r", idents)
             return
         queue_id, client_id = idents[:2]
         try:
             msg = self.session.unserialize(msg)
         except Exception:
             self.log.error("queue::client %r sent invalid message to %r: %r", client_id, queue_id, msg, exc_info=True)
             return
 
         eid = self.by_ident.get(queue_id, None)
         if eid is None:
             self.log.error("queue::target %r not registered", queue_id)
             self.log.debug("queue::    valid are: %r", self.by_ident.keys())
             return
         record = init_record(msg)
         msg_id = record['msg_id']
         self.log.info("queue::client %r submitted request %r to %s", client_id, msg_id, eid)
         # Unicode in records
         record['engine_uuid'] = queue_id.decode('ascii')
         record['client_uuid'] = msg['header']['session']
         record['queue'] = 'mux'
 
         try:
             # it's posible iopub arrived first:
             existing = self.db.get_record(msg_id)
             for key,evalue in iteritems(existing):
                 rvalue = record.get(key, None)
                 if evalue and rvalue and evalue != rvalue:
                     self.log.warn("conflicting initial state for record: %r:%r <%r> %r", msg_id, rvalue, key, evalue)
                 elif evalue and not rvalue:
                     record[key] = evalue
             try:
                 self.db.update_record(msg_id, record)
             except Exception:
                 self.log.error("DB Error updating record %r", msg_id, exc_info=True)
         except KeyError:
             try:
                 self.db.add_record(msg_id, record)
             except Exception:
                 self.log.error("DB Error adding record %r", msg_id, exc_info=True)
 
 
         self.pending.add(msg_id)
         self.queues[eid].append(msg_id)
 
     def save_queue_result(self, idents, msg):
         if len(idents) < 2:
             self.log.error("invalid identity prefix: %r", idents)
             return
 
         client_id, queue_id = idents[:2]
         try:
             msg = self.session.unserialize(msg)
         except Exception:
             self.log.error("queue::engine %r sent invalid message to %r: %r",
                     queue_id, client_id, msg, exc_info=True)
             return
 
         eid = self.by_ident.get(queue_id, None)
         if eid is None:
             self.log.error("queue::unknown engine %r is sending a reply: ", queue_id)
             return
 
         parent = msg['parent_header']
         if not parent:
             return
         msg_id = parent['msg_id']
         if msg_id in self.pending:
             self.pending.remove(msg_id)
             self.all_completed.add(msg_id)
             self.queues[eid].remove(msg_id)
             self.completed[eid].append(msg_id)
             self.log.info("queue::request %r completed on %s", msg_id, eid)
         elif msg_id not in self.all_completed:
             # it could be a result from a dead engine that died before delivering the
             # result
             self.log.warn("queue:: unknown msg finished %r", msg_id)
             return
         # update record anyway, because the unregistration could have been premature
         rheader = msg['header']
         md = msg['metadata']
         completed = rheader['date']
         started = md.get('started', None)
         result = {
             'result_header' : rheader,
             'result_metadata': md,
             'result_content': msg['content'],
             'received': datetime.now(),
             'started' : started,
             'completed' : completed
         }
 
         result['result_buffers'] = msg['buffers']
         try:
             self.db.update_record(msg_id, result)
         except Exception:
             self.log.error("DB Error updating record %r", msg_id, exc_info=True)
 
 
     #--------------------- Task Queue Traffic ------------------------------
 
     def save_task_request(self, idents, msg):
         """Save the submission of a task."""
         client_id = idents[0]
 
         try:
             msg = self.session.unserialize(msg)
         except Exception:
             self.log.error("task::client %r sent invalid task message: %r",
                     client_id, msg, exc_info=True)
             return
         record = init_record(msg)
 
         record['client_uuid'] = msg['header']['session']
         record['queue'] = 'task'
         header = msg['header']
         msg_id = header['msg_id']
         self.pending.add(msg_id)
         self.unassigned.add(msg_id)
         try:
             # it's posible iopub arrived first:
             existing = self.db.get_record(msg_id)
             if existing['resubmitted']:
                 for key in ('submitted', 'client_uuid', 'buffers'):
                     # don't clobber these keys on resubmit
                     # submitted and client_uuid should be different
                     # and buffers might be big, and shouldn't have changed
                     record.pop(key)
                     # still check content,header which should not change
                     # but are not expensive to compare as buffers
 
             for key,evalue in iteritems(existing):
                 if key.endswith('buffers'):
                     # don't compare buffers
                     continue
                 rvalue = record.get(key, None)
                 if evalue and rvalue and evalue != rvalue:
                     self.log.warn("conflicting initial state for record: %r:%r <%r> %r", msg_id, rvalue, key, evalue)
                 elif evalue and not rvalue:
                     record[key] = evalue
             try:
                 self.db.update_record(msg_id, record)
             except Exception:
                 self.log.error("DB Error updating record %r", msg_id, exc_info=True)
         except KeyError:
             try:
                 self.db.add_record(msg_id, record)
             except Exception:
                 self.log.error("DB Error adding record %r", msg_id, exc_info=True)
         except Exception:
             self.log.error("DB Error saving task request %r", msg_id, exc_info=True)
 
     def save_task_result(self, idents, msg):
         """save the result of a completed task."""
         client_id = idents[0]
         try:
             msg = self.session.unserialize(msg)
         except Exception:
             self.log.error("task::invalid task result message send to %r: %r",
                     client_id, msg, exc_info=True)
             return
 
         parent = msg['parent_header']
         if not parent:
             # print msg
             self.log.warn("Task %r had no parent!", msg)
             return
         msg_id = parent['msg_id']
         if msg_id in self.unassigned:
             self.unassigned.remove(msg_id)
 
         header = msg['header']
         md = msg['metadata']
         engine_uuid = md.get('engine', u'')
         eid = self.by_ident.get(cast_bytes(engine_uuid), None)
         
         status = md.get('status', None)
 
         if msg_id in self.pending:
             self.log.info("task::task %r finished on %s", msg_id, eid)
             self.pending.remove(msg_id)
             self.all_completed.add(msg_id)
             if eid is not None:
                 if status != 'aborted':
                     self.completed[eid].append(msg_id)
                 if msg_id in self.tasks[eid]:
                     self.tasks[eid].remove(msg_id)
             completed = header['date']
             started = md.get('started', None)
             result = {
                 'result_header' : header,
                 'result_metadata': msg['metadata'],
                 'result_content': msg['content'],
                 'started' : started,
                 'completed' : completed,
                 'received' : datetime.now(),
                 'engine_uuid': engine_uuid,
             }
 
             result['result_buffers'] = msg['buffers']
             try:
                 self.db.update_record(msg_id, result)
             except Exception:
                 self.log.error("DB Error saving task request %r", msg_id, exc_info=True)
 
         else:
             self.log.debug("task::unknown task %r finished", msg_id)
 
     def save_task_destination(self, idents, msg):
         try:
             msg = self.session.unserialize(msg, content=True)
         except Exception:
             self.log.error("task::invalid task tracking message", exc_info=True)
             return
         content = msg['content']
         # print (content)
         msg_id = content['msg_id']
         engine_uuid = content['engine_id']
         eid = self.by_ident[cast_bytes(engine_uuid)]
 
         self.log.info("task::task %r arrived on %r", msg_id, eid)
         if msg_id in self.unassigned:
             self.unassigned.remove(msg_id)
         # else:
         #     self.log.debug("task::task %r not listed as MIA?!"%(msg_id))
 
         self.tasks[eid].append(msg_id)
         # self.pending[msg_id][1].update(received=datetime.now(),engine=(eid,engine_uuid))
         try:
             self.db.update_record(msg_id, dict(engine_uuid=engine_uuid))
         except Exception:
             self.log.error("DB Error saving task destination %r", msg_id, exc_info=True)
 
 
     def mia_task_request(self, idents, msg):
         raise NotImplementedError
         client_id = idents[0]
         # content = dict(mia=self.mia,status='ok')
         # self.session.send('mia_reply', content=content, idents=client_id)
 
 
     #--------------------- IOPub Traffic ------------------------------
 
     def save_iopub_message(self, topics, msg):
         """save an iopub message into the db"""
         # print (topics)
         try:
             msg = self.session.unserialize(msg, content=True)
         except Exception:
             self.log.error("iopub::invalid IOPub message", exc_info=True)
             return
 
         parent = msg['parent_header']
         if not parent:
             self.log.warn("iopub::IOPub message lacks parent: %r", msg)
             return
         msg_id = parent['msg_id']
         msg_type = msg['header']['msg_type']
         content = msg['content']
 
         # ensure msg_id is in db
         try:
             rec = self.db.get_record(msg_id)
         except KeyError:
             rec = empty_record()
             rec['msg_id'] = msg_id
             self.db.add_record(msg_id, rec)
         # stream
         d = {}
         if msg_type == 'stream':
             name = content['name']
             s = rec[name] or ''
             d[name] = s + content['data']
 
         elif msg_type == 'pyerr':
             d['pyerr'] = content
         elif msg_type == 'pyin':
             d['pyin'] = content['code']
         elif msg_type in ('display_data', 'pyout'):
             d[msg_type] = content
         elif msg_type == 'status':
             pass
         elif msg_type == 'data_pub':
             self.log.info("ignored data_pub message for %s" % msg_id)
         else:
             self.log.warn("unhandled iopub msg_type: %r", msg_type)
 
         if not d:
             return
 
         try:
             self.db.update_record(msg_id, d)
         except Exception:
             self.log.error("DB Error saving iopub message %r", msg_id, exc_info=True)
 
 
 
     #-------------------------------------------------------------------------
     # Registration requests
     #-------------------------------------------------------------------------
 
     def connection_request(self, client_id, msg):
         """Reply with connection addresses for clients."""
         self.log.info("client::client %r connected", client_id)
         content = dict(status='ok')
         jsonable = {}
         for k,v in iteritems(self.keytable):
             if v not in self.dead_engines:
                 jsonable[str(k)] = v
         content['engines'] = jsonable
         self.session.send(self.query, 'connection_reply', content, parent=msg, ident=client_id)
 
     def register_engine(self, reg, msg):
         """Register a new engine."""
         content = msg['content']
         try:
             uuid = content['uuid']
         except KeyError:
             self.log.error("registration::queue not specified", exc_info=True)
             return
 
         eid = self._next_id
 
         self.log.debug("registration::register_engine(%i, %r)", eid, uuid)
 
         content = dict(id=eid,status='ok',hb_period=self.heartmonitor.period)
         # check if requesting available IDs:
         if cast_bytes(uuid) in self.by_ident:
             try:
                 raise KeyError("uuid %r in use" % uuid)
             except:
                 content = error.wrap_exception()
                 self.log.error("uuid %r in use", uuid, exc_info=True)
         else:
             for h, ec in iteritems(self.incoming_registrations):
                 if uuid == h:
                     try:
                         raise KeyError("heart_id %r in use" % uuid)
                     except:
                         self.log.error("heart_id %r in use", uuid, exc_info=True)
                         content = error.wrap_exception()
                     break
                 elif uuid == ec.uuid:
                     try:
                         raise KeyError("uuid %r in use" % uuid)
                     except:
                         self.log.error("uuid %r in use", uuid, exc_info=True)
                         content = error.wrap_exception()
                     break
 
         msg = self.session.send(self.query, "registration_reply",
                 content=content,
                 ident=reg)
 
         heart = cast_bytes(uuid)
 
         if content['status'] == 'ok':
             if heart in self.heartmonitor.hearts:
                 # already beating
                 self.incoming_registrations[heart] = EngineConnector(id=eid,uuid=uuid)
                 self.finish_registration(heart)
             else:
                 purge = lambda : self._purge_stalled_registration(heart)
                 dc = ioloop.DelayedCallback(purge, self.registration_timeout, self.loop)
                 dc.start()
                 self.incoming_registrations[heart] = EngineConnector(id=eid,uuid=uuid,stallback=dc)
         else:
             self.log.error("registration::registration %i failed: %r", eid, content['evalue'])
         
         return eid
 
     def unregister_engine(self, ident, msg):
         """Unregister an engine that explicitly requested to leave."""
         try:
             eid = msg['content']['id']
         except:
             self.log.error("registration::bad engine id for unregistration: %r", ident, exc_info=True)
             return
         self.log.info("registration::unregister_engine(%r)", eid)
         # print (eid)
         uuid = self.keytable[eid]
         content=dict(id=eid, uuid=uuid)
         self.dead_engines.add(uuid)
         # self.ids.remove(eid)
         # uuid = self.keytable.pop(eid)
         #
         # ec = self.engines.pop(eid)
         # self.hearts.pop(ec.heartbeat)
         # self.by_ident.pop(ec.queue)
         # self.completed.pop(eid)
         handleit = lambda : self._handle_stranded_msgs(eid, uuid)
         dc = ioloop.DelayedCallback(handleit, self.registration_timeout, self.loop)
         dc.start()
         ############## TODO: HANDLE IT ################
         
         self._save_engine_state()
 
         if self.notifier:
             self.session.send(self.notifier, "unregistration_notification", content=content)
 
     def _handle_stranded_msgs(self, eid, uuid):
         """Handle messages known to be on an engine when the engine unregisters.
 
         It is possible that this will fire prematurely - that is, an engine will
         go down after completing a result, and the client will be notified
         that the result failed and later receive the actual result.
         """
 
         outstanding = self.queues[eid]
 
         for msg_id in outstanding:
             self.pending.remove(msg_id)
             self.all_completed.add(msg_id)
             try:
                 raise error.EngineError("Engine %r died while running task %r" % (eid, msg_id))
             except:
                 content = error.wrap_exception()
             # build a fake header:
             header = {}
             header['engine'] = uuid
             header['date'] = datetime.now()
             rec = dict(result_content=content, result_header=header, result_buffers=[])
             rec['completed'] = header['date']
             rec['engine_uuid'] = uuid
             try:
                 self.db.update_record(msg_id, rec)
             except Exception:
                 self.log.error("DB Error handling stranded msg %r", msg_id, exc_info=True)
 
 
     def finish_registration(self, heart):
         """Second half of engine registration, called after our HeartMonitor
         has received a beat from the Engine's Heart."""
         try:
             ec = self.incoming_registrations.pop(heart)
         except KeyError:
             self.log.error("registration::tried to finish nonexistant registration", exc_info=True)
             return
         self.log.info("registration::finished registering engine %i:%s", ec.id, ec.uuid)
         if ec.stallback is not None:
             ec.stallback.stop()
         eid = ec.id
         self.ids.add(eid)
         self.keytable[eid] = ec.uuid
         self.engines[eid] = ec
         self.by_ident[cast_bytes(ec.uuid)] = ec.id
         self.queues[eid] = list()
         self.tasks[eid] = list()
         self.completed[eid] = list()
         self.hearts[heart] = eid
         content = dict(id=eid, uuid=self.engines[eid].uuid)
         if self.notifier:
             self.session.send(self.notifier, "registration_notification", content=content)
         self.log.info("engine::Engine Connected: %i", eid)
         
         self._save_engine_state()
 
     def _purge_stalled_registration(self, heart):
         if heart in self.incoming_registrations:
             ec = self.incoming_registrations.pop(heart)
             self.log.info("registration::purging stalled registration: %i", ec.id)
         else:
             pass
 
     #-------------------------------------------------------------------------
     # Engine State
     #-------------------------------------------------------------------------
 
 
     def _cleanup_engine_state_file(self):
         """cleanup engine state mapping"""
         
         if os.path.exists(self.engine_state_file):
             self.log.debug("cleaning up engine state: %s", self.engine_state_file)
             try:
                 os.remove(self.engine_state_file)
             except IOError:
                 self.log.error("Couldn't cleanup file: %s", self.engine_state_file, exc_info=True)
 
 
     def _save_engine_state(self):
         """save engine mapping to JSON file"""
         if not self.engine_state_file:
             return
         self.log.debug("save engine state to %s" % self.engine_state_file)
         state = {}
         engines = {}
         for eid, ec in iteritems(self.engines):
             if ec.uuid not in self.dead_engines:
                 engines[eid] = ec.uuid
         
         state['engines'] = engines
         
         state['next_id'] = self._idcounter
         
         with open(self.engine_state_file, 'w') as f:
             json.dump(state, f)
 
 
     def _load_engine_state(self):
         """load engine mapping from JSON file"""
         if not os.path.exists(self.engine_state_file):
             return
         
         self.log.info("loading engine state from %s" % self.engine_state_file)
         
         with open(self.engine_state_file) as f:
             state = json.load(f)
         
         save_notifier = self.notifier
         self.notifier = None
         for eid, uuid in iteritems(state['engines']):
             heart = uuid.encode('ascii')
             # start with this heart as current and beating:
             self.heartmonitor.responses.add(heart)
             self.heartmonitor.hearts.add(heart)
             
             self.incoming_registrations[heart] = EngineConnector(id=int(eid), uuid=uuid)
             self.finish_registration(heart)
         
         self.notifier = save_notifier
         
         self._idcounter = state['next_id']
 
     #-------------------------------------------------------------------------
     # Client Requests
     #-------------------------------------------------------------------------
 
     def shutdown_request(self, client_id, msg):
         """handle shutdown request."""
         self.session.send(self.query, 'shutdown_reply', content={'status': 'ok'}, ident=client_id)
         # also notify other clients of shutdown
         self.session.send(self.notifier, 'shutdown_notice', content={'status': 'ok'})
         dc = ioloop.DelayedCallback(lambda : self._shutdown(), 1000, self.loop)
         dc.start()
 
     def _shutdown(self):
         self.log.info("hub::hub shutting down.")
         time.sleep(0.1)
         sys.exit(0)
 
 
     def check_load(self, client_id, msg):
         content = msg['content']
         try:
             targets = content['targets']
             targets = self._validate_targets(targets)
         except:
             content = error.wrap_exception()
             self.session.send(self.query, "hub_error",
                     content=content, ident=client_id)
             return
 
         content = dict(status='ok')
         # loads = {}
         for t in targets:
             content[bytes(t)] = len(self.queues[t])+len(self.tasks[t])
         self.session.send(self.query, "load_reply", content=content, ident=client_id)
 
 
     def queue_status(self, client_id, msg):
         """Return the Queue status of one or more targets.
         if verbose: return the msg_ids
         else: return len of each type.
         keys: queue (pending MUX jobs)
             tasks (pending Task jobs)
             completed (finished jobs from both queues)"""
         content = msg['content']
         targets = content['targets']
         try:
             targets = self._validate_targets(targets)
         except:
             content = error.wrap_exception()
             self.session.send(self.query, "hub_error",
                     content=content, ident=client_id)
             return
         verbose = content.get('verbose', False)
         content = dict(status='ok')
         for t in targets:
             queue = self.queues[t]
             completed = self.completed[t]
             tasks = self.tasks[t]
             if not verbose:
                 queue = len(queue)
                 completed = len(completed)
                 tasks = len(tasks)
             content[str(t)] = {'queue': queue, 'completed': completed , 'tasks': tasks}
         content['unassigned'] = list(self.unassigned) if verbose else len(self.unassigned)
         # print (content)
         self.session.send(self.query, "queue_reply", content=content, ident=client_id)
 
     def purge_results(self, client_id, msg):
         """Purge results from memory. This method is more valuable before we move
         to a DB based message storage mechanism."""
         content = msg['content']
         self.log.info("Dropping records with %s", content)
         msg_ids = content.get('msg_ids', [])
         reply = dict(status='ok')
         if msg_ids == 'all':
             try:
                 self.db.drop_matching_records(dict(completed={'$ne':None}))
             except Exception:
                 reply = error.wrap_exception()
         else:
-            pending = filter(lambda m: m in self.pending, msg_ids)
+            pending = [m for m in msg_ids if (m in self.pending)]
             if pending:
                 try:
                     raise IndexError("msg pending: %r" % pending[0])
                 except:
                     reply = error.wrap_exception()
             else:
                 try:
                     self.db.drop_matching_records(dict(msg_id={'$in':msg_ids}))
                 except Exception:
                     reply = error.wrap_exception()
 
             if reply['status'] == 'ok':
                 eids = content.get('engine_ids', [])
                 for eid in eids:
                     if eid not in self.engines:
                         try:
                             raise IndexError("No such engine: %i" % eid)
                         except:
                             reply = error.wrap_exception()
                         break
                     uid = self.engines[eid].uuid
                     try:
                         self.db.drop_matching_records(dict(engine_uuid=uid, completed={'$ne':None}))
                     except Exception:
                         reply = error.wrap_exception()
                         break
 
         self.session.send(self.query, 'purge_reply', content=reply, ident=client_id)
 
     def resubmit_task(self, client_id, msg):
         """Resubmit one or more tasks."""
         def finish(reply):
             self.session.send(self.query, 'resubmit_reply', content=reply, ident=client_id)
 
         content = msg['content']
         msg_ids = content['msg_ids']
         reply = dict(status='ok')
         try:
             records = self.db.find_records({'msg_id' : {'$in' : msg_ids}}, keys=[
                 'header', 'content', 'buffers'])
         except Exception:
             self.log.error('db::db error finding tasks to resubmit', exc_info=True)
             return finish(error.wrap_exception())
 
         # validate msg_ids
         found_ids = [ rec['msg_id'] for rec in records ]
         pending_ids = [ msg_id for msg_id in found_ids if msg_id in self.pending ]
         if len(records) > len(msg_ids):
             try:
                 raise RuntimeError("DB appears to be in an inconsistent state."
                     "More matching records were found than should exist")
             except Exception:
                 return finish(error.wrap_exception())
         elif len(records) < len(msg_ids):
             missing = [ m for m in msg_ids if m not in found_ids ]
             try:
                 raise KeyError("No such msg(s): %r" % missing)
             except KeyError:
                 return finish(error.wrap_exception())
         elif pending_ids:
             pass
             # no need to raise on resubmit of pending task, now that we
             # resubmit under new ID, but do we want to raise anyway?
             # msg_id = invalid_ids[0]
             # try:
             #     raise ValueError("Task(s) %r appears to be inflight" % )
             # except Exception:
             #     return finish(error.wrap_exception())
 
         # mapping of original IDs to resubmitted IDs
         resubmitted = {}
 
         # send the messages
         for rec in records:
             header = rec['header']
             msg = self.session.msg(header['msg_type'], parent=header)
             msg_id = msg['msg_id']
             msg['content'] = rec['content']
             
             # use the old header, but update msg_id and timestamp
             fresh = msg['header']
             header['msg_id'] = fresh['msg_id']
             header['date'] = fresh['date']
             msg['header'] = header
 
             self.session.send(self.resubmit, msg, buffers=rec['buffers'])
 
             resubmitted[rec['msg_id']] = msg_id
             self.pending.add(msg_id)
             msg['buffers'] = rec['buffers']
             try:
                 self.db.add_record(msg_id, init_record(msg))
             except Exception:
                 self.log.error("db::DB Error updating record: %s", msg_id, exc_info=True)
                 return finish(error.wrap_exception())
 
         finish(dict(status='ok', resubmitted=resubmitted))
         
         # store the new IDs in the Task DB
         for msg_id, resubmit_id in iteritems(resubmitted):
             try:
                 self.db.update_record(msg_id, {'resubmitted' : resubmit_id})
             except Exception:
                 self.log.error("db::DB Error updating record: %s", msg_id, exc_info=True)
 
 
     def _extract_record(self, rec):
         """decompose a TaskRecord dict into subsection of reply for get_result"""
         io_dict = {}
         for key in ('pyin', 'pyout', 'pyerr', 'stdout', 'stderr'):
                 io_dict[key] = rec[key]
         content = { 
             'header': rec['header'],
             'metadata': rec['metadata'],
             'result_metadata': rec['result_metadata'],
             'result_header' : rec['result_header'],
             'result_content': rec['result_content'],
             'received' : rec['received'],
             'io' : io_dict,
         }
         if rec['result_buffers']:
-            buffers = map(bytes, rec['result_buffers'])
+            buffers = list(map(bytes, rec['result_buffers']))
         else:
             buffers = []
 
         return content, buffers
 
     def get_results(self, client_id, msg):
         """Get the result of 1 or more messages."""
         content = msg['content']
         msg_ids = sorted(set(content['msg_ids']))
         statusonly = content.get('status_only', False)
         pending = []
         completed = []
         content = dict(status='ok')
         content['pending'] = pending
         content['completed'] = completed
         buffers = []
         if not statusonly:
             try:
                 matches = self.db.find_records(dict(msg_id={'$in':msg_ids}))
                 # turn match list into dict, for faster lookup
                 records = {}
                 for rec in matches:
                     records[rec['msg_id']] = rec
             except Exception:
                 content = error.wrap_exception()
                 self.session.send(self.query, "result_reply", content=content,
                                                     parent=msg, ident=client_id)
                 return
         else:
             records = {}
         for msg_id in msg_ids:
             if msg_id in self.pending:
                 pending.append(msg_id)
             elif msg_id in self.all_completed:
                 completed.append(msg_id)
                 if not statusonly:
                     c,bufs = self._extract_record(records[msg_id])
                     content[msg_id] = c
                     buffers.extend(bufs)
             elif msg_id in records:
                 if rec['completed']:
                     completed.append(msg_id)
                     c,bufs = self._extract_record(records[msg_id])
                     content[msg_id] = c
                     buffers.extend(bufs)
                 else:
                     pending.append(msg_id)
             else:
                 try:
                     raise KeyError('No such message: '+msg_id)
                 except:
                     content = error.wrap_exception()
                 break
         self.session.send(self.query, "result_reply", content=content,
                                             parent=msg, ident=client_id,
                                             buffers=buffers)
 
     def get_history(self, client_id, msg):
         """Get a list of all msg_ids in our DB records"""
         try:
             msg_ids = self.db.get_history()
         except Exception as e:
             content = error.wrap_exception()
         else:
             content = dict(status='ok', history=msg_ids)
 
         self.session.send(self.query, "history_reply", content=content,
                                             parent=msg, ident=client_id)
 
     def db_query(self, client_id, msg):
         """Perform a raw query on the task record database."""
         content = msg['content']
         query = content.get('query', {})
         keys = content.get('keys', None)
         buffers = []
         empty = list()
         try:
             records = self.db.find_records(query, keys)
         except Exception as e:
             content = error.wrap_exception()
         else:
             # extract buffers from reply content:
             if keys is not None:
                 buffer_lens = [] if 'buffers' in keys else None
                 result_buffer_lens = [] if 'result_buffers' in keys else None
             else:
                 buffer_lens = None
                 result_buffer_lens = None
 
             for rec in records:
                 # buffers may be None, so double check
                 b = rec.pop('buffers', empty) or empty
                 if buffer_lens is not None:
                     buffer_lens.append(len(b))
                     buffers.extend(b)
                 rb = rec.pop('result_buffers', empty) or empty
                 if result_buffer_lens is not None:
                     result_buffer_lens.append(len(rb))
                     buffers.extend(rb)
             content = dict(status='ok', records=records, buffer_lens=buffer_lens,
                                     result_buffer_lens=result_buffer_lens)
         # self.log.debug (content)
         self.session.send(self.query, "db_reply", content=content,
                                             parent=msg, ident=client_id,
                                             buffers=buffers)
 

IPython/parallel/controller/mongodb.py

@@ -1,122 +1,122 b''
 """A TaskRecord backend using mongodb
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 from pymongo import Connection
 
 # bson.Binary import moved
 try:
     from bson.binary import Binary
 except ImportError:
     from bson import Binary
 
 from IPython.utils.traitlets import Dict, List, Unicode, Instance
 
 from .dictdb import BaseDB
 
 #-----------------------------------------------------------------------------
 # MongoDB class
 #-----------------------------------------------------------------------------
 
 class MongoDB(BaseDB):
     """MongoDB TaskRecord backend."""
     
     connection_args = List(config=True,
         help="""Positional arguments to be passed to pymongo.Connection.  Only
         necessary if the default mongodb configuration does not point to your
         mongod instance.""")
     connection_kwargs = Dict(config=True,
         help="""Keyword arguments to be passed to pymongo.Connection.  Only
         necessary if the default mongodb configuration does not point to your
         mongod instance."""
     )
     database = Unicode("ipython-tasks", config=True,
         help="""The MongoDB database name to use for storing tasks for this session. If unspecified,
         a new database will be created with the Hub's IDENT.  Specifying the database will result
         in tasks from previous sessions being available via Clients' db_query and
         get_result methods.""")
 
     _connection = Instance(Connection) # pymongo connection
     
     def __init__(self, **kwargs):
         super(MongoDB, self).__init__(**kwargs)
         if self._connection is None:
             self._connection = Connection(*self.connection_args, **self.connection_kwargs)
         if not self.database:
             self.database = self.session
         self._db = self._connection[self.database]
         self._records = self._db['task_records']
         self._records.ensure_index('msg_id', unique=True)
         self._records.ensure_index('submitted') # for sorting history
         # for rec in self._records.find
     
     def _binary_buffers(self, rec):
         for key in ('buffers', 'result_buffers'):
             if rec.get(key, None):
-                rec[key] = map(Binary, rec[key])
+                rec[key] = list(map(Binary, rec[key]))
         return rec
     
     def add_record(self, msg_id, rec):
         """Add a new Task Record, by msg_id."""
         # print rec
         rec = self._binary_buffers(rec)
         self._records.insert(rec)
     
     def get_record(self, msg_id):
         """Get a specific Task Record, by msg_id."""
         r = self._records.find_one({'msg_id': msg_id})
         if not r:
             # r will be '' if nothing is found
             raise KeyError(msg_id)
         return r
     
     def update_record(self, msg_id, rec):
         """Update the data in an existing record."""
         rec = self._binary_buffers(rec)
 
         self._records.update({'msg_id':msg_id}, {'$set': rec})
     
     def drop_matching_records(self, check):
         """Remove a record from the DB."""
         self._records.remove(check)
         
     def drop_record(self, msg_id):
         """Remove a record from the DB."""
         self._records.remove({'msg_id':msg_id})
     
     def find_records(self, check, keys=None):
         """Find records matching a query dict, optionally extracting subset of keys.
         
         Returns list of matching records.
         
         Parameters
         ----------
         
         check: dict
             mongodb-style query argument
         keys: list of strs [optional]
             if specified, the subset of keys to extract.  msg_id will *always* be
             included.
         """
         if keys and 'msg_id' not in keys:
             keys.append('msg_id')
         matches = list(self._records.find(check,keys))
         for rec in matches:
             rec.pop('_id')
         return matches
 
     def get_history(self):
         """get all msg_ids, ordered by time submitted."""
         cursor = self._records.find({},{'msg_id':1}).sort('submitted')
         return [ rec['msg_id'] for rec in cursor ]
 
 

IPython/parallel/controller/scheduler.py

@@ -1,860 +1,859 b''
 """The Python scheduler for rich scheduling.
 
 The Pure ZMQ scheduler does not allow routing schemes other than LRU,
 nor does it check msg_id DAG dependencies. For those, a slightly slower
 Python Scheduler exists.
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #----------------------------------------------------------------------
 # Imports
 #----------------------------------------------------------------------
 
 import logging
 import sys
 import time
 
 from collections import deque
 from datetime import datetime
 from random import randint, random
 from types import FunctionType
 
 try:
     import numpy
 except ImportError:
     numpy = None
 
 import zmq
 from zmq.eventloop import ioloop, zmqstream
 
 # local imports
 from IPython.external.decorator import decorator
 from IPython.config.application import Application
 from IPython.config.loader import Config
 from IPython.utils.traitlets import Instance, Dict, List, Set, Integer, Enum, CBytes
 from IPython.utils.py3compat import cast_bytes
 
 from IPython.parallel import error, util
 from IPython.parallel.factory import SessionFactory
 from IPython.parallel.util import connect_logger, local_logger
 
 from .dependency import Dependency
 
 @decorator
 def logged(f,self,*args,**kwargs):
     # print ("#--------------------")
     self.log.debug("scheduler::%s(*%s,**%s)", f.__name__, args, kwargs)
     # print ("#--")
     return f(self,*args, **kwargs)
 
 #----------------------------------------------------------------------
 # Chooser functions
 #----------------------------------------------------------------------
 
 def plainrandom(loads):
     """Plain random pick."""
     n = len(loads)
     return randint(0,n-1)
 
 def lru(loads):
     """Always pick the front of the line.
 
     The content of `loads` is ignored.
 
     Assumes LRU ordering of loads, with oldest first.
     """
     return 0
 
 def twobin(loads):
     """Pick two at random, use the LRU of the two.
 
     The content of loads is ignored.
 
     Assumes LRU ordering of loads, with oldest first.
     """
     n = len(loads)
     a = randint(0,n-1)
     b = randint(0,n-1)
     return min(a,b)
 
 def weighted(loads):
     """Pick two at random using inverse load as weight.
 
     Return the less loaded of the two.
     """
     # weight 0 a million times more than 1:
     weights = 1./(1e-6+numpy.array(loads))
     sums = weights.cumsum()
     t = sums[-1]
     x = random()*t
     y = random()*t
     idx = 0
     idy = 0
     while sums[idx] < x:
         idx += 1
     while sums[idy] < y:
         idy += 1
     if weights[idy] > weights[idx]:
         return idy
     else:
         return idx
 
 def leastload(loads):
     """Always choose the lowest load.
 
     If the lowest load occurs more than once, the first
     occurance will be used.  If loads has LRU ordering, this means
     the LRU of those with the lowest load is chosen.
     """
     return loads.index(min(loads))
 
 #---------------------------------------------------------------------
 # Classes
 #---------------------------------------------------------------------
 
 
 # store empty default dependency:
 MET = Dependency([])
 
 
 class Job(object):
     """Simple container for a job"""
     def __init__(self, msg_id, raw_msg, idents, msg, header, metadata,
                     targets, after, follow, timeout):
         self.msg_id = msg_id
         self.raw_msg = raw_msg
         self.idents = idents
         self.msg = msg
         self.header = header
         self.metadata = metadata
         self.targets = targets
         self.after = after
         self.follow = follow
         self.timeout = timeout
         
         self.removed = False # used for lazy-delete from sorted queue
         self.timestamp = time.time()
         self.timeout_id = 0
         self.blacklist = set()
 
     def __lt__(self, other):
         return self.timestamp < other.timestamp
 
     def __cmp__(self, other):
         return cmp(self.timestamp, other.timestamp)
 
     @property
     def dependents(self):
         return self.follow.union(self.after)
 
 
 class TaskScheduler(SessionFactory):
     """Python TaskScheduler object.
 
     This is the simplest object that supports msg_id based
     DAG dependencies. *Only* task msg_ids are checked, not
     msg_ids of jobs submitted via the MUX queue.
 
     """
 
     hwm = Integer(1, config=True,
         help="""specify the High Water Mark (HWM) for the downstream
         socket in the Task scheduler. This is the maximum number
         of allowed outstanding tasks on each engine.
         
         The default (1) means that only one task can be outstanding on each
         engine.  Setting TaskScheduler.hwm=0 means there is no limit, and the
         engines continue to be assigned tasks while they are working,
         effectively hiding network latency behind computation, but can result
         in an imbalance of work when submitting many heterogenous tasks all at
         once.  Any positive value greater than one is a compromise between the
         two.
 
         """
     )
     scheme_name = Enum(('leastload', 'pure', 'lru', 'plainrandom', 'weighted', 'twobin'),
         'leastload', config=True, allow_none=False,
         help="""select the task scheduler scheme  [default: Python LRU]
         Options are: 'pure', 'lru', 'plainrandom', 'weighted', 'twobin','leastload'"""
     )
     def _scheme_name_changed(self, old, new):
         self.log.debug("Using scheme %r"%new)
         self.scheme = globals()[new]
 
     # input arguments:
     scheme = Instance(FunctionType) # function for determining the destination
     def _scheme_default(self):
         return leastload
     client_stream = Instance(zmqstream.ZMQStream) # client-facing stream
     engine_stream = Instance(zmqstream.ZMQStream) # engine-facing stream
     notifier_stream = Instance(zmqstream.ZMQStream) # hub-facing sub stream
     mon_stream = Instance(zmqstream.ZMQStream) # hub-facing pub stream
     query_stream = Instance(zmqstream.ZMQStream) # hub-facing DEALER stream
 
     # internals:
     queue = Instance(deque) # sorted list of Jobs
     def _queue_default(self):
         return deque()
     queue_map = Dict() # dict by msg_id of Jobs (for O(1) access to the Queue)
     graph = Dict() # dict by msg_id of [ msg_ids that depend on key ]
     retries = Dict() # dict by msg_id of retries remaining (non-neg ints)
     # waiting = List() # list of msg_ids ready to run, but haven't due to HWM
     pending = Dict() # dict by engine_uuid of submitted tasks
     completed = Dict() # dict by engine_uuid of completed tasks
     failed = Dict() # dict by engine_uuid of failed tasks
     destinations = Dict() # dict by msg_id of engine_uuids where jobs ran (reverse of completed+failed)
     clients = Dict() # dict by msg_id for who submitted the task
     targets = List() # list of target IDENTs
     loads = List() # list of engine loads
     # full = Set() # set of IDENTs that have HWM outstanding tasks
     all_completed = Set() # set of all completed tasks
     all_failed = Set() # set of all failed tasks
     all_done = Set() # set of all finished tasks=union(completed,failed)
     all_ids = Set() # set of all submitted task IDs
 
     ident = CBytes() # ZMQ identity. This should just be self.session.session
                      # but ensure Bytes
     def _ident_default(self):
         return self.session.bsession
 
     def start(self):
         self.query_stream.on_recv(self.dispatch_query_reply)
         self.session.send(self.query_stream, "connection_request", {})
         
         self.engine_stream.on_recv(self.dispatch_result, copy=False)
         self.client_stream.on_recv(self.dispatch_submission, copy=False)
 
         self._notification_handlers = dict(
             registration_notification = self._register_engine,
             unregistration_notification = self._unregister_engine
         )
         self.notifier_stream.on_recv(self.dispatch_notification)
         self.log.info("Scheduler started [%s]" % self.scheme_name)
 
     def resume_receiving(self):
         """Resume accepting jobs."""
         self.client_stream.on_recv(self.dispatch_submission, copy=False)
 
     def stop_receiving(self):
         """Stop accepting jobs while there are no engines.
         Leave them in the ZMQ queue."""
         self.client_stream.on_recv(None)
 
     #-----------------------------------------------------------------------
     # [Un]Registration Handling
     #-----------------------------------------------------------------------
     
     
     def dispatch_query_reply(self, msg):
         """handle reply to our initial connection request"""
         try:
             idents,msg = self.session.feed_identities(msg)
         except ValueError:
             self.log.warn("task::Invalid Message: %r",msg)
             return
         try:
             msg = self.session.unserialize(msg)
         except ValueError:
             self.log.warn("task::Unauthorized message from: %r"%idents)
             return
         
         content = msg['content']
         for uuid in content.get('engines', {}).values():
             self._register_engine(cast_bytes(uuid))
 
     
     @util.log_errors
     def dispatch_notification(self, msg):
         """dispatch register/unregister events."""
         try:
             idents,msg = self.session.feed_identities(msg)
         except ValueError:
             self.log.warn("task::Invalid Message: %r",msg)
             return
         try:
             msg = self.session.unserialize(msg)
         except ValueError:
             self.log.warn("task::Unauthorized message from: %r"%idents)
             return
 
         msg_type = msg['header']['msg_type']
 
         handler = self._notification_handlers.get(msg_type, None)
         if handler is None:
             self.log.error("Unhandled message type: %r"%msg_type)
         else:
             try:
                 handler(cast_bytes(msg['content']['uuid']))
             except Exception:
                 self.log.error("task::Invalid notification msg: %r", msg, exc_info=True)
 
     def _register_engine(self, uid):
         """New engine with ident `uid` became available."""
         # head of the line:
         self.targets.insert(0,uid)
         self.loads.insert(0,0)
 
         # initialize sets
         self.completed[uid] = set()
         self.failed[uid] = set()
         self.pending[uid] = {}
 
         # rescan the graph:
         self.update_graph(None)
 
     def _unregister_engine(self, uid):
         """Existing engine with ident `uid` became unavailable."""
         if len(self.targets) == 1:
             # this was our only engine
             pass
 
         # handle any potentially finished tasks:
         self.engine_stream.flush()
 
         # don't pop destinations, because they might be used later
         # map(self.destinations.pop, self.completed.pop(uid))
         # map(self.destinations.pop, self.failed.pop(uid))
 
         # prevent this engine from receiving work
         idx = self.targets.index(uid)
         self.targets.pop(idx)
         self.loads.pop(idx)
 
         # wait 5 seconds before cleaning up pending jobs, since the results might
         # still be incoming
         if self.pending[uid]:
             dc = ioloop.DelayedCallback(lambda : self.handle_stranded_tasks(uid), 5000, self.loop)
             dc.start()
         else:
             self.completed.pop(uid)
             self.failed.pop(uid)
 
 
     def handle_stranded_tasks(self, engine):
         """Deal with jobs resident in an engine that died."""
         lost = self.pending[engine]
         for msg_id in lost.keys():
             if msg_id not in self.pending[engine]:
                 # prevent double-handling of messages
                 continue
 
             raw_msg = lost[msg_id].raw_msg
             idents,msg = self.session.feed_identities(raw_msg, copy=False)
             parent = self.session.unpack(msg[1].bytes)
             idents = [engine, idents[0]]
 
             # build fake error reply
             try:
                 raise error.EngineError("Engine %r died while running task %r"%(engine, msg_id))
             except:
                 content = error.wrap_exception()
             # build fake metadata
             md = dict(
                 status=u'error',
                 engine=engine.decode('ascii'),
                 date=datetime.now(),
             )
             msg = self.session.msg('apply_reply', content, parent=parent, metadata=md)
-            raw_reply = map(zmq.Message, self.session.serialize(msg, ident=idents))
+            raw_reply = list(map(zmq.Message, self.session.serialize(msg, ident=idents)))
             # and dispatch it
             self.dispatch_result(raw_reply)
 
         # finally scrub completed/failed lists
         self.completed.pop(engine)
         self.failed.pop(engine)
 
 
     #-----------------------------------------------------------------------
     # Job Submission
     #-----------------------------------------------------------------------
     
 
     @util.log_errors
     def dispatch_submission(self, raw_msg):
         """Dispatch job submission to appropriate handlers."""
         # ensure targets up to date:
         self.notifier_stream.flush()
         try:
             idents, msg = self.session.feed_identities(raw_msg, copy=False)
             msg = self.session.unserialize(msg, content=False, copy=False)
         except Exception:
             self.log.error("task::Invaid task msg: %r"%raw_msg, exc_info=True)
             return
 
 
         # send to monitor
         self.mon_stream.send_multipart([b'intask']+raw_msg, copy=False)
 
         header = msg['header']
         md = msg['metadata']
         msg_id = header['msg_id']
         self.all_ids.add(msg_id)
 
         # get targets as a set of bytes objects
         # from a list of unicode objects
         targets = md.get('targets', [])
-        targets = map(cast_bytes, targets)
-        targets = set(targets)
+        targets = set(map(cast_bytes, targets))
 
         retries = md.get('retries', 0)
         self.retries[msg_id] = retries
 
         # time dependencies
         after = md.get('after', None)
         if after:
             after = Dependency(after)
             if after.all:
                 if after.success:
                     after = Dependency(after.difference(self.all_completed),
                                 success=after.success,
                                 failure=after.failure,
                                 all=after.all,
                     )
                 if after.failure:
                     after = Dependency(after.difference(self.all_failed),
                                 success=after.success,
                                 failure=after.failure,
                                 all=after.all,
                     )
             if after.check(self.all_completed, self.all_failed):
                 # recast as empty set, if `after` already met,
                 # to prevent unnecessary set comparisons
                 after = MET
         else:
             after = MET
 
         # location dependencies
         follow = Dependency(md.get('follow', []))
 
         timeout = md.get('timeout', None)
         if timeout:
             timeout = float(timeout)
 
         job = Job(msg_id=msg_id, raw_msg=raw_msg, idents=idents, msg=msg,
                  header=header, targets=targets, after=after, follow=follow,
                  timeout=timeout, metadata=md,
         )
         # validate and reduce dependencies:
         for dep in after,follow:
             if not dep: # empty dependency
                 continue
             # check valid:
             if msg_id in dep or dep.difference(self.all_ids):
                 self.queue_map[msg_id] = job
                 return self.fail_unreachable(msg_id, error.InvalidDependency)
             # check if unreachable:
             if dep.unreachable(self.all_completed, self.all_failed):
                 self.queue_map[msg_id] = job
                 return self.fail_unreachable(msg_id)
 
         if after.check(self.all_completed, self.all_failed):
             # time deps already met, try to run
             if not self.maybe_run(job):
                 # can't run yet
                 if msg_id not in self.all_failed:
                     # could have failed as unreachable
                     self.save_unmet(job)
         else:
             self.save_unmet(job)
 
     def job_timeout(self, job, timeout_id):
         """callback for a job's timeout.
         
         The job may or may not have been run at this point.
         """
         if job.timeout_id != timeout_id:
             # not the most recent call
             return
         now = time.time()
         if job.timeout >= (now + 1):
             self.log.warn("task %s timeout fired prematurely: %s > %s",
                 job.msg_id, job.timeout, now
             )
         if job.msg_id in self.queue_map:
             # still waiting, but ran out of time
             self.log.info("task %r timed out", job.msg_id)
             self.fail_unreachable(job.msg_id, error.TaskTimeout)
 
     def fail_unreachable(self, msg_id, why=error.ImpossibleDependency):
         """a task has become unreachable, send a reply with an ImpossibleDependency
         error."""
         if msg_id not in self.queue_map:
             self.log.error("task %r already failed!", msg_id)
             return
         job = self.queue_map.pop(msg_id)
         # lazy-delete from the queue
         job.removed = True
         for mid in job.dependents:
             if mid in self.graph:
                 self.graph[mid].remove(msg_id)
 
         try:
             raise why()
         except:
             content = error.wrap_exception()
         self.log.debug("task %r failing as unreachable with: %s", msg_id, content['ename'])
 
         self.all_done.add(msg_id)
         self.all_failed.add(msg_id)
 
         msg = self.session.send(self.client_stream, 'apply_reply', content,
                                                 parent=job.header, ident=job.idents)
         self.session.send(self.mon_stream, msg, ident=[b'outtask']+job.idents)
 
         self.update_graph(msg_id, success=False)
 
     def available_engines(self):
         """return a list of available engine indices based on HWM"""
         if not self.hwm:
-            return range(len(self.targets))
+            return list(range(len(self.targets)))
         available = []
         for idx in range(len(self.targets)):
             if self.loads[idx] < self.hwm:
                 available.append(idx)
         return available
 
     def maybe_run(self, job):
         """check location dependencies, and run if they are met."""
         msg_id = job.msg_id
         self.log.debug("Attempting to assign task %s", msg_id)
         available = self.available_engines()
         if not available:
             # no engines, definitely can't run
             return False
         
         if job.follow or job.targets or job.blacklist or self.hwm:
             # we need a can_run filter
             def can_run(idx):
                 # check hwm
                 if self.hwm and self.loads[idx] == self.hwm:
                     return False
                 target = self.targets[idx]
                 # check blacklist
                 if target in job.blacklist:
                     return False
                 # check targets
                 if job.targets and target not in job.targets:
                     return False
                 # check follow
                 return job.follow.check(self.completed[target], self.failed[target])
 
-            indices = filter(can_run, available)
+            indices = list(filter(can_run, available))
 
             if not indices:
                 # couldn't run
                 if job.follow.all:
                     # check follow for impossibility
                     dests = set()
                     relevant = set()
                     if job.follow.success:
                         relevant = self.all_completed
                     if job.follow.failure:
                         relevant = relevant.union(self.all_failed)
                     for m in job.follow.intersection(relevant):
                         dests.add(self.destinations[m])
                     if len(dests) > 1:
                         self.queue_map[msg_id] = job
                         self.fail_unreachable(msg_id)
                         return False
                 if job.targets:
                     # check blacklist+targets for impossibility
                     job.targets.difference_update(job.blacklist)
                     if not job.targets or not job.targets.intersection(self.targets):
                         self.queue_map[msg_id] = job
                         self.fail_unreachable(msg_id)
                         return False
                 return False
         else:
             indices = None
 
         self.submit_task(job, indices)
         return True
 
     def save_unmet(self, job):
         """Save a message for later submission when its dependencies are met."""
         msg_id = job.msg_id
         self.log.debug("Adding task %s to the queue", msg_id)
         self.queue_map[msg_id] = job
         self.queue.append(job)
         # track the ids in follow or after, but not those already finished
         for dep_id in job.after.union(job.follow).difference(self.all_done):
             if dep_id not in self.graph:
                 self.graph[dep_id] = set()
             self.graph[dep_id].add(msg_id)
         
         # schedule timeout callback
         if job.timeout:
             timeout_id = job.timeout_id = job.timeout_id + 1
             self.loop.add_timeout(time.time() + job.timeout,
                 lambda : self.job_timeout(job, timeout_id)
             )
         
 
     def submit_task(self, job, indices=None):
         """Submit a task to any of a subset of our targets."""
         if indices:
             loads = [self.loads[i] for i in indices]
         else:
             loads = self.loads
         idx = self.scheme(loads)
         if indices:
             idx = indices[idx]
         target = self.targets[idx]
         # print (target, map(str, msg[:3]))
         # send job to the engine
         self.engine_stream.send(target, flags=zmq.SNDMORE, copy=False)
         self.engine_stream.send_multipart(job.raw_msg, copy=False)
         # update load
         self.add_job(idx)
         self.pending[target][job.msg_id] = job
         # notify Hub
         content = dict(msg_id=job.msg_id, engine_id=target.decode('ascii'))
         self.session.send(self.mon_stream, 'task_destination', content=content,
                         ident=[b'tracktask',self.ident])
 
 
     #-----------------------------------------------------------------------
     # Result Handling
     #-----------------------------------------------------------------------
     
     
     @util.log_errors
     def dispatch_result(self, raw_msg):
         """dispatch method for result replies"""
         try:
             idents,msg = self.session.feed_identities(raw_msg, copy=False)
             msg = self.session.unserialize(msg, content=False, copy=False)
             engine = idents[0]
             try:
                 idx = self.targets.index(engine)
             except ValueError:
                 pass # skip load-update for dead engines
             else:
                 self.finish_job(idx)
         except Exception:
             self.log.error("task::Invalid result: %r", raw_msg, exc_info=True)
             return
 
         md = msg['metadata']
         parent = msg['parent_header']
         if md.get('dependencies_met', True):
             success = (md['status'] == 'ok')
             msg_id = parent['msg_id']
             retries = self.retries[msg_id]
             if not success and retries > 0:
                 # failed
                 self.retries[msg_id] = retries - 1
                 self.handle_unmet_dependency(idents, parent)
             else:
                 del self.retries[msg_id]
                 # relay to client and update graph
                 self.handle_result(idents, parent, raw_msg, success)
                 # send to Hub monitor
                 self.mon_stream.send_multipart([b'outtask']+raw_msg, copy=False)
         else:
             self.handle_unmet_dependency(idents, parent)
 
     def handle_result(self, idents, parent, raw_msg, success=True):
         """handle a real task result, either success or failure"""
         # first, relay result to client
         engine = idents[0]
         client = idents[1]
         # swap_ids for ROUTER-ROUTER mirror
         raw_msg[:2] = [client,engine]
         # print (map(str, raw_msg[:4]))
         self.client_stream.send_multipart(raw_msg, copy=False)
         # now, update our data structures
         msg_id = parent['msg_id']
         self.pending[engine].pop(msg_id)
         if success:
             self.completed[engine].add(msg_id)
             self.all_completed.add(msg_id)
         else:
             self.failed[engine].add(msg_id)
             self.all_failed.add(msg_id)
         self.all_done.add(msg_id)
         self.destinations[msg_id] = engine
 
         self.update_graph(msg_id, success)
 
     def handle_unmet_dependency(self, idents, parent):
         """handle an unmet dependency"""
         engine = idents[0]
         msg_id = parent['msg_id']
 
         job = self.pending[engine].pop(msg_id)
         job.blacklist.add(engine)
 
         if job.blacklist == job.targets:
             self.queue_map[msg_id] = job
             self.fail_unreachable(msg_id)
         elif not self.maybe_run(job):
             # resubmit failed
             if msg_id not in self.all_failed:
                 # put it back in our dependency tree
                 self.save_unmet(job)
 
         if self.hwm:
             try:
                 idx = self.targets.index(engine)
             except ValueError:
                 pass # skip load-update for dead engines
             else:
                 if self.loads[idx] == self.hwm-1:
                     self.update_graph(None)
 
     def update_graph(self, dep_id=None, success=True):
         """dep_id just finished. Update our dependency
         graph and submit any jobs that just became runnable.
 
         Called with dep_id=None to update entire graph for hwm, but without finishing a task.
         """
         # print ("\n\n***********")
         # pprint (dep_id)
         # pprint (self.graph)
         # pprint (self.queue_map)
         # pprint (self.all_completed)
         # pprint (self.all_failed)
         # print ("\n\n***********\n\n")
         # update any jobs that depended on the dependency
         msg_ids = self.graph.pop(dep_id, [])
 
         # recheck *all* jobs if
         # a) we have HWM and an engine just become no longer full
         # or b) dep_id was given as None
         
         if dep_id is None or self.hwm and any( [ load==self.hwm-1 for load in self.loads ]):
             jobs = self.queue
             using_queue = True
         else:
             using_queue = False
             jobs = deque(sorted( self.queue_map[msg_id] for msg_id in msg_ids ))
         
         to_restore = []
         while jobs:
             job = jobs.popleft()
             if job.removed:
                 continue
             msg_id = job.msg_id
             
             put_it_back = True
             
             if job.after.unreachable(self.all_completed, self.all_failed)\
                     or job.follow.unreachable(self.all_completed, self.all_failed):
                 self.fail_unreachable(msg_id)
                 put_it_back = False
 
             elif job.after.check(self.all_completed, self.all_failed): # time deps met, maybe run
                 if self.maybe_run(job):
                     put_it_back = False
                     self.queue_map.pop(msg_id)
                     for mid in job.dependents:
                         if mid in self.graph:
                             self.graph[mid].remove(msg_id)
                     
                     # abort the loop if we just filled up all of our engines.
                     # avoids an O(N) operation in situation of full queue,
                     # where graph update is triggered as soon as an engine becomes
                     # non-full, and all tasks after the first are checked,
                     # even though they can't run.
                     if not self.available_engines():
                         break
             
             if using_queue and put_it_back:
                 # popped a job from the queue but it neither ran nor failed,
                 # so we need to put it back when we are done
                 # make sure to_restore preserves the same ordering
                 to_restore.append(job)
         
         # put back any tasks we popped but didn't run
         if using_queue:
             self.queue.extendleft(to_restore)
     
     #----------------------------------------------------------------------
     # methods to be overridden by subclasses
     #----------------------------------------------------------------------
 
     def add_job(self, idx):
         """Called after self.targets[idx] just got the job with header.
         Override with subclasses.  The default ordering is simple LRU.
         The default loads are the number of outstanding jobs."""
         self.loads[idx] += 1
         for lis in (self.targets, self.loads):
             lis.append(lis.pop(idx))
 
 
     def finish_job(self, idx):
         """Called after self.targets[idx] just finished a job.
         Override with subclasses."""
         self.loads[idx] -= 1
 
 
 
 def launch_scheduler(in_addr, out_addr, mon_addr, not_addr, reg_addr, config=None,
                         logname='root', log_url=None, loglevel=logging.DEBUG,
                         identity=b'task', in_thread=False):
 
     ZMQStream = zmqstream.ZMQStream
 
     if config:
         # unwrap dict back into Config
         config = Config(config)
 
     if in_thread:
         # use instance() to get the same Context/Loop as our parent
         ctx = zmq.Context.instance()
         loop = ioloop.IOLoop.instance()
     else:
         # in a process, don't use instance()
         # for safety with multiprocessing
         ctx = zmq.Context()
         loop = ioloop.IOLoop()
     ins = ZMQStream(ctx.socket(zmq.ROUTER),loop)
     util.set_hwm(ins, 0)
     ins.setsockopt(zmq.IDENTITY, identity + b'_in')
     ins.bind(in_addr)
 
     outs = ZMQStream(ctx.socket(zmq.ROUTER),loop)
     util.set_hwm(outs, 0)
     outs.setsockopt(zmq.IDENTITY, identity + b'_out')
     outs.bind(out_addr)
     mons = zmqstream.ZMQStream(ctx.socket(zmq.PUB),loop)
     util.set_hwm(mons, 0)
     mons.connect(mon_addr)
     nots = zmqstream.ZMQStream(ctx.socket(zmq.SUB),loop)
     nots.setsockopt(zmq.SUBSCRIBE, b'')
     nots.connect(not_addr)
     
     querys = ZMQStream(ctx.socket(zmq.DEALER),loop)
     querys.connect(reg_addr)
     
     # setup logging.
     if in_thread:
         log = Application.instance().log
     else:
         if log_url:
             log = connect_logger(logname, ctx, log_url, root="scheduler", loglevel=loglevel)
         else:
             log = local_logger(logname, loglevel)
 
     scheduler = TaskScheduler(client_stream=ins, engine_stream=outs,
                             mon_stream=mons, notifier_stream=nots,
                             query_stream=querys,
                             loop=loop, log=log,
                             config=config)
     scheduler.start()
     if not in_thread:
         try:
             loop.start()
         except KeyboardInterrupt:
             scheduler.log.critical("Interrupted, exiting...")
 

IPython/parallel/controller/sqlitedb.py

@@ -1,422 +1,422 b''
 """A TaskRecord backend using sqlite3
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 import json
 import os
 try:
     import cPickle as pickle
 except ImportError:
     import pickle
 from datetime import datetime
 
 try:
     import sqlite3
 except ImportError:
     sqlite3 = None
 
 from zmq.eventloop import ioloop
 
 from IPython.utils.traitlets import Unicode, Instance, List, Dict
 from .dictdb import BaseDB
 from IPython.utils.jsonutil import date_default, extract_dates, squash_dates
 from IPython.utils.py3compat import iteritems
 
 #-----------------------------------------------------------------------------
 # SQLite operators, adapters, and converters
 #-----------------------------------------------------------------------------
 
 try:
     buffer
 except NameError:
     # py3k
     buffer = memoryview
 
 operators = {
  '$lt' : "<",
  '$gt' : ">",
  # null is handled weird with ==,!=
  '$eq' : "=",
  '$ne' : "!=",
  '$lte': "<=",
  '$gte': ">=",
  '$in' : ('=', ' OR '),
  '$nin': ('!=', ' AND '),
  # '$all': None,
  # '$mod': None,
  # '$exists' : None
 }
 null_operators = {
 '=' : "IS NULL",
 '!=' : "IS NOT NULL",
 }
 
 def _adapt_dict(d):
     return json.dumps(d, default=date_default)
 
 def _convert_dict(ds):
     if ds is None:
         return ds
     else:
         if isinstance(ds, bytes):
             # If I understand the sqlite doc correctly, this will always be utf8
             ds = ds.decode('utf8')
         return extract_dates(json.loads(ds))
 
 def _adapt_bufs(bufs):
     # this is *horrible*
     # copy buffers into single list and pickle it:
     if bufs and isinstance(bufs[0], (bytes, buffer)):
-        return sqlite3.Binary(pickle.dumps(map(bytes, bufs),-1))
+        return sqlite3.Binary(pickle.dumps(list(map(bytes, bufs)),-1))
     elif bufs:
         return bufs
     else:
         return None
 
 def _convert_bufs(bs):
     if bs is None:
         return []
     else:
         return pickle.loads(bytes(bs))
 
 #-----------------------------------------------------------------------------
 # SQLiteDB class
 #-----------------------------------------------------------------------------
 
 class SQLiteDB(BaseDB):
     """SQLite3 TaskRecord backend."""
 
     filename = Unicode('tasks.db', config=True,
         help="""The filename of the sqlite task database. [default: 'tasks.db']""")
     location = Unicode('', config=True,
         help="""The directory containing the sqlite task database.  The default
         is to use the cluster_dir location.""")
     table = Unicode("ipython-tasks", config=True,
         help="""The SQLite Table to use for storing tasks for this session. If unspecified,
         a new table will be created with the Hub's IDENT.  Specifying the table will result
         in tasks from previous sessions being available via Clients' db_query and
         get_result methods.""")
 
     if sqlite3 is not None:
         _db = Instance('sqlite3.Connection')
     else:
         _db = None
     # the ordered list of column names
     _keys = List(['msg_id' ,
             'header' ,
             'metadata',
             'content',
             'buffers',
             'submitted',
             'client_uuid' ,
             'engine_uuid' ,
             'started',
             'completed',
             'resubmitted',
             'received',
             'result_header' ,
             'result_metadata',
             'result_content' ,
             'result_buffers' ,
             'queue' ,
             'pyin' ,
             'pyout',
             'pyerr',
             'stdout',
             'stderr',
         ])
     # sqlite datatypes for checking that db is current format
     _types = Dict({'msg_id' : 'text' ,
             'header' : 'dict text',
             'metadata' : 'dict text',
             'content' : 'dict text',
             'buffers' : 'bufs blob',
             'submitted' : 'timestamp',
             'client_uuid' : 'text',
             'engine_uuid' : 'text',
             'started' : 'timestamp',
             'completed' : 'timestamp',
             'resubmitted' : 'text',
             'received' : 'timestamp',
             'result_header' : 'dict text',
             'result_metadata' : 'dict text',
             'result_content' : 'dict text',
             'result_buffers' : 'bufs blob',
             'queue' : 'text',
             'pyin' : 'text',
             'pyout' : 'text',
             'pyerr' : 'text',
             'stdout' : 'text',
             'stderr' : 'text',
         })
 
     def __init__(self, **kwargs):
         super(SQLiteDB, self).__init__(**kwargs)
         if sqlite3 is None:
             raise ImportError("SQLiteDB requires sqlite3")
         if not self.table:
             # use session, and prefix _, since starting with # is illegal
             self.table = '_'+self.session.replace('-','_')
         if not self.location:
             # get current profile
             from IPython.core.application import BaseIPythonApplication
             if BaseIPythonApplication.initialized():
                 app = BaseIPythonApplication.instance()
                 if app.profile_dir is not None:
                     self.location = app.profile_dir.location
                 else:
                     self.location = u'.'
             else:
                 self.location = u'.'
         self._init_db()
 
         # register db commit as 2s periodic callback
         # to prevent clogging pipes
         # assumes we are being run in a zmq ioloop app
         loop = ioloop.IOLoop.instance()
         pc = ioloop.PeriodicCallback(self._db.commit, 2000, loop)
         pc.start()
 
     def _defaults(self, keys=None):
         """create an empty record"""
         d = {}
         keys = self._keys if keys is None else keys
         for key in keys:
             d[key] = None
         return d
 
     def _check_table(self):
         """Ensure that an incorrect table doesn't exist
 
         If a bad (old) table does exist, return False
         """
         cursor = self._db.execute("PRAGMA table_info('%s')"%self.table)
         lines = cursor.fetchall()
         if not lines:
             # table does not exist
             return True
         types = {}
         keys = []
         for line in lines:
             keys.append(line[1])
             types[line[1]] = line[2]
         if self._keys != keys:
             # key mismatch
             self.log.warn('keys mismatch')
             return False
         for key in self._keys:
             if types[key] != self._types[key]:
                 self.log.warn(
                     'type mismatch: %s: %s != %s'%(key,types[key],self._types[key])
                 )
                 return False
         return True
 
     def _init_db(self):
         """Connect to the database and get new session number."""
         # register adapters
         sqlite3.register_adapter(dict, _adapt_dict)
         sqlite3.register_converter('dict', _convert_dict)
         sqlite3.register_adapter(list, _adapt_bufs)
         sqlite3.register_converter('bufs', _convert_bufs)
         # connect to the db
         dbfile = os.path.join(self.location, self.filename)
         self._db = sqlite3.connect(dbfile, detect_types=sqlite3.PARSE_DECLTYPES,
             # isolation_level = None)#,
              cached_statements=64)
         # print dir(self._db)
         first_table = previous_table = self.table
         i=0
         while not self._check_table():
             i+=1
             self.table = first_table+'_%i'%i
             self.log.warn(
                 "Table %s exists and doesn't match db format, trying %s"%
                 (previous_table, self.table)
             )
             previous_table = self.table
 
         self._db.execute("""CREATE TABLE IF NOT EXISTS '%s'
                 (msg_id text PRIMARY KEY,
                 header dict text,
                 metadata dict text,
                 content dict text,
                 buffers bufs blob,
                 submitted timestamp,
                 client_uuid text,
                 engine_uuid text,
                 started timestamp,
                 completed timestamp,
                 resubmitted text,
                 received timestamp,
                 result_header dict text,
                 result_metadata dict text,
                 result_content dict text,
                 result_buffers bufs blob,
                 queue text,
                 pyin text,
                 pyout text,
                 pyerr text,
                 stdout text,
                 stderr text)
                 """%self.table)
         self._db.commit()
 
     def _dict_to_list(self, d):
         """turn a mongodb-style record dict into a list."""
 
         return [ d[key] for key in self._keys ]
 
     def _list_to_dict(self, line, keys=None):
         """Inverse of dict_to_list"""
         keys = self._keys if keys is None else keys
         d = self._defaults(keys)
         for key,value in zip(keys, line):
             d[key] = value
 
         return d
 
     def _render_expression(self, check):
         """Turn a mongodb-style search dict into an SQL query."""
         expressions = []
         args = []
 
         skeys = set(check.keys())
         skeys.difference_update(set(self._keys))
         skeys.difference_update(set(['buffers', 'result_buffers']))
         if skeys:
             raise KeyError("Illegal testing key(s): %s"%skeys)
 
         for name,sub_check in iteritems(check):
             if isinstance(sub_check, dict):
                 for test,value in iteritems(sub_check):
                     try:
                         op = operators[test]
                     except KeyError:
                         raise KeyError("Unsupported operator: %r"%test)
                     if isinstance(op, tuple):
                         op, join = op
 
                     if value is None and op in null_operators:
                         expr = "%s %s" % (name, null_operators[op])
                     else:
                         expr = "%s %s ?"%(name, op)
                         if isinstance(value, (tuple,list)):
                             if op in null_operators and any([v is None for v in value]):
                                 # equality tests don't work with NULL
                                 raise ValueError("Cannot use %r test with NULL values on SQLite backend"%test)
                             expr = '( %s )'%( join.join([expr]*len(value)) )
                             args.extend(value)
                         else:
                             args.append(value)
                     expressions.append(expr)
             else:
                 # it's an equality check
                 if sub_check is None:
                     expressions.append("%s IS NULL" % name)
                 else:
                     expressions.append("%s = ?"%name)
                     args.append(sub_check)
 
         expr = " AND ".join(expressions)
         return expr, args
 
     def add_record(self, msg_id, rec):
         """Add a new Task Record, by msg_id."""
         d = self._defaults()
         d.update(rec)
         d['msg_id'] = msg_id
         line = self._dict_to_list(d)
         tups = '(%s)'%(','.join(['?']*len(line)))
         self._db.execute("INSERT INTO '%s' VALUES %s"%(self.table, tups), line)
         # self._db.commit()
 
     def get_record(self, msg_id):
         """Get a specific Task Record, by msg_id."""
         cursor = self._db.execute("""SELECT * FROM '%s' WHERE msg_id==?"""%self.table, (msg_id,))
         line = cursor.fetchone()
         if line is None:
             raise KeyError("No such msg: %r"%msg_id)
         return self._list_to_dict(line)
 
     def update_record(self, msg_id, rec):
         """Update the data in an existing record."""
         query = "UPDATE '%s' SET "%self.table
         sets = []
         keys = sorted(rec.keys())
         values = []
         for key in keys:
             sets.append('%s = ?'%key)
             values.append(rec[key])
         query += ', '.join(sets)
         query += ' WHERE msg_id == ?'
         values.append(msg_id)
         self._db.execute(query, values)
         # self._db.commit()
 
     def drop_record(self, msg_id):
         """Remove a record from the DB."""
         self._db.execute("""DELETE FROM '%s' WHERE msg_id==?"""%self.table, (msg_id,))
         # self._db.commit()
 
     def drop_matching_records(self, check):
         """Remove a record from the DB."""
         expr,args = self._render_expression(check)
         query = "DELETE FROM '%s' WHERE %s"%(self.table, expr)
         self._db.execute(query,args)
         # self._db.commit()
 
     def find_records(self, check, keys=None):
         """Find records matching a query dict, optionally extracting subset of keys.
 
         Returns list of matching records.
 
         Parameters
         ----------
 
         check: dict
             mongodb-style query argument
         keys: list of strs [optional]
             if specified, the subset of keys to extract.  msg_id will *always* be
             included.
         """
         if keys:
             bad_keys = [ key for key in keys if key not in self._keys ]
             if bad_keys:
                 raise KeyError("Bad record key(s): %s"%bad_keys)
 
         if keys:
             # ensure msg_id is present and first:
             if 'msg_id' in keys:
                 keys.remove('msg_id')
             keys.insert(0, 'msg_id')
             req = ', '.join(keys)
         else:
             req = '*'
         expr,args = self._render_expression(check)
         query = """SELECT %s FROM '%s' WHERE %s"""%(req, self.table, expr)
         cursor = self._db.execute(query, args)
         matches = cursor.fetchall()
         records = []
         for line in matches:
             rec = self._list_to_dict(line, keys)
             records.append(rec)
         return records
 
     def get_history(self):
         """get all msg_ids, ordered by time submitted."""
         query = """SELECT msg_id FROM '%s' ORDER by submitted ASC"""%self.table
         cursor = self._db.execute(query)
         # will be a list of length 1 tuples
         return [ tup[0] for tup in cursor.fetchall()]
 
 __all__ = ['SQLiteDB']
  No newline at end of file

IPython/parallel/tests/test_asyncresult.py

@@ -1,326 +1,326 b''
 """Tests for asyncresult.py
 
 Authors:
 
 * Min RK
 """
 
 #-------------------------------------------------------------------------------
 #  Copyright (C) 2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-------------------------------------------------------------------------------
 
 #-------------------------------------------------------------------------------
 # Imports
 #-------------------------------------------------------------------------------
 
 import time
 
 import nose.tools as nt
 
 from IPython.utils.io import capture_output
 
 from IPython.parallel.error import TimeoutError
 from IPython.parallel import error, Client
 from IPython.parallel.tests import add_engines
 from .clienttest import ClusterTestCase
 from IPython.utils.py3compat import iteritems
 
 def setup():
     add_engines(2, total=True)
 
 def wait(n):
     import time
     time.sleep(n)
     return n
 
 def echo(x):
     return x
 
 class AsyncResultTest(ClusterTestCase):
 
     def test_single_result_view(self):
         """various one-target views get the right value for single_result"""
         eid = self.client.ids[-1]
         ar = self.client[eid].apply_async(lambda : 42)
         self.assertEqual(ar.get(), 42)
         ar = self.client[[eid]].apply_async(lambda : 42)
         self.assertEqual(ar.get(), [42])
         ar = self.client[-1:].apply_async(lambda : 42)
         self.assertEqual(ar.get(), [42])
 
     def test_get_after_done(self):
         ar = self.client[-1].apply_async(lambda : 42)
         ar.wait()
         self.assertTrue(ar.ready())
         self.assertEqual(ar.get(), 42)
         self.assertEqual(ar.get(), 42)
 
     def test_get_before_done(self):
         ar = self.client[-1].apply_async(wait, 0.1)
         self.assertRaises(TimeoutError, ar.get, 0)
         ar.wait(0)
         self.assertFalse(ar.ready())
         self.assertEqual(ar.get(), 0.1)
 
     def test_get_after_error(self):
         ar = self.client[-1].apply_async(lambda : 1/0)
         ar.wait(10)
         self.assertRaisesRemote(ZeroDivisionError, ar.get)
         self.assertRaisesRemote(ZeroDivisionError, ar.get)
         self.assertRaisesRemote(ZeroDivisionError, ar.get_dict)
     
     def test_get_dict(self):
         n = len(self.client)
         ar = self.client[:].apply_async(lambda : 5)
         self.assertEqual(ar.get(), [5]*n)
         d = ar.get_dict()
         self.assertEqual(sorted(d.keys()), sorted(self.client.ids))
         for eid,r in iteritems(d):
             self.assertEqual(r, 5)
     
     def test_get_dict_single(self):
         view = self.client[-1]
-        for v in (range(5), 5, ('abc', 'def'), 'string'):
+        for v in (list(range(5)), 5, ('abc', 'def'), 'string'):
             ar = view.apply_async(echo, v)
             self.assertEqual(ar.get(), v)
             d = ar.get_dict()
             self.assertEqual(d, {view.targets : v})
     
     def test_get_dict_bad(self):
         ar = self.client[:].apply_async(lambda : 5)
         ar2 = self.client[:].apply_async(lambda : 5)
         ar = self.client.get_result(ar.msg_ids + ar2.msg_ids)
         self.assertRaises(ValueError, ar.get_dict)
     
     def test_list_amr(self):
         ar = self.client.load_balanced_view().map_async(wait, [0.1]*5)
         rlist = list(ar)
     
     def test_getattr(self):
         ar = self.client[:].apply_async(wait, 0.5)
         self.assertEqual(ar.engine_id, [None] * len(ar))
         self.assertRaises(AttributeError, lambda : ar._foo)
         self.assertRaises(AttributeError, lambda : ar.__length_hint__())
         self.assertRaises(AttributeError, lambda : ar.foo)
         self.assertFalse(hasattr(ar, '__length_hint__'))
         self.assertFalse(hasattr(ar, 'foo'))
         self.assertTrue(hasattr(ar, 'engine_id'))
         ar.get(5)
         self.assertRaises(AttributeError, lambda : ar._foo)
         self.assertRaises(AttributeError, lambda : ar.__length_hint__())
         self.assertRaises(AttributeError, lambda : ar.foo)
         self.assertTrue(isinstance(ar.engine_id, list))
         self.assertEqual(ar.engine_id, ar['engine_id'])
         self.assertFalse(hasattr(ar, '__length_hint__'))
         self.assertFalse(hasattr(ar, 'foo'))
         self.assertTrue(hasattr(ar, 'engine_id'))
 
     def test_getitem(self):
         ar = self.client[:].apply_async(wait, 0.5)
         self.assertEqual(ar['engine_id'], [None] * len(ar))
         self.assertRaises(KeyError, lambda : ar['foo'])
         ar.get(5)
         self.assertRaises(KeyError, lambda : ar['foo'])
         self.assertTrue(isinstance(ar['engine_id'], list))
         self.assertEqual(ar.engine_id, ar['engine_id'])
     
     def test_single_result(self):
         ar = self.client[-1].apply_async(wait, 0.5)
         self.assertRaises(KeyError, lambda : ar['foo'])
         self.assertEqual(ar['engine_id'], None)
         self.assertTrue(ar.get(5) == 0.5)
         self.assertTrue(isinstance(ar['engine_id'], int))
         self.assertTrue(isinstance(ar.engine_id, int))
         self.assertEqual(ar.engine_id, ar['engine_id'])
     
     def test_abort(self):
         e = self.client[-1]
         ar = e.execute('import time; time.sleep(1)', block=False)
         ar2 = e.apply_async(lambda : 2)
         ar2.abort()
         self.assertRaises(error.TaskAborted, ar2.get)
         ar.get()
     
     def test_len(self):
         v = self.client.load_balanced_view()
-        ar = v.map_async(lambda x: x, range(10))
+        ar = v.map_async(lambda x: x, list(range(10)))
         self.assertEqual(len(ar), 10)
-        ar = v.apply_async(lambda x: x, range(10))
+        ar = v.apply_async(lambda x: x, list(range(10)))
         self.assertEqual(len(ar), 1)
-        ar = self.client[:].apply_async(lambda x: x, range(10))
+        ar = self.client[:].apply_async(lambda x: x, list(range(10)))
         self.assertEqual(len(ar), len(self.client.ids))
     
     def test_wall_time_single(self):
         v = self.client.load_balanced_view()
         ar = v.apply_async(time.sleep, 0.25)
         self.assertRaises(TimeoutError, getattr, ar, 'wall_time')
         ar.get(2)
         self.assertTrue(ar.wall_time < 1.)
         self.assertTrue(ar.wall_time > 0.2)
 
     def test_wall_time_multi(self):
         self.minimum_engines(4)
         v = self.client[:]
         ar = v.apply_async(time.sleep, 0.25)
         self.assertRaises(TimeoutError, getattr, ar, 'wall_time')
         ar.get(2)
         self.assertTrue(ar.wall_time < 1.)
         self.assertTrue(ar.wall_time > 0.2)
 
     def test_serial_time_single(self):
         v = self.client.load_balanced_view()
         ar = v.apply_async(time.sleep, 0.25)
         self.assertRaises(TimeoutError, getattr, ar, 'serial_time')
         ar.get(2)
         self.assertTrue(ar.serial_time < 1.)
         self.assertTrue(ar.serial_time > 0.2)
 
     def test_serial_time_multi(self):
         self.minimum_engines(4)
         v = self.client[:]
         ar = v.apply_async(time.sleep, 0.25)
         self.assertRaises(TimeoutError, getattr, ar, 'serial_time')
         ar.get(2)
         self.assertTrue(ar.serial_time < 2.)
         self.assertTrue(ar.serial_time > 0.8)
 
     def test_elapsed_single(self):
         v = self.client.load_balanced_view()
         ar = v.apply_async(time.sleep, 0.25)
         while not ar.ready():
             time.sleep(0.01)
             self.assertTrue(ar.elapsed < 1)
         self.assertTrue(ar.elapsed < 1)
         ar.get(2)
 
     def test_elapsed_multi(self):
         v = self.client[:]
         ar = v.apply_async(time.sleep, 0.25)
         while not ar.ready():
             time.sleep(0.01)
             self.assertTrue(ar.elapsed < 1)
         self.assertTrue(ar.elapsed < 1)
         ar.get(2)
 
     def test_hubresult_timestamps(self):
         self.minimum_engines(4)
         v = self.client[:]
         ar = v.apply_async(time.sleep, 0.25)
         ar.get(2)
         rc2 = Client(profile='iptest')
         # must have try/finally to close second Client, otherwise
         # will have dangling sockets causing problems
         try:
             time.sleep(0.25)
             hr = rc2.get_result(ar.msg_ids)
             self.assertTrue(hr.elapsed > 0., "got bad elapsed: %s" % hr.elapsed)
             hr.get(1)
             self.assertTrue(hr.wall_time < ar.wall_time + 0.2, "got bad wall_time: %s > %s" % (hr.wall_time, ar.wall_time))
             self.assertEqual(hr.serial_time, ar.serial_time)
         finally:
             rc2.close()
 
     def test_display_empty_streams_single(self):
         """empty stdout/err are not displayed (single result)"""
         self.minimum_engines(1)
         
         v = self.client[-1]
         ar = v.execute("print (5555)")
         ar.get(5)
         with capture_output() as io:
             ar.display_outputs()
         self.assertEqual(io.stderr, '')
         self.assertEqual('5555\n', io.stdout)
 
         ar = v.execute("a=5")
         ar.get(5)
         with capture_output() as io:
             ar.display_outputs()
         self.assertEqual(io.stderr, '')
         self.assertEqual(io.stdout, '')
         
     def test_display_empty_streams_type(self):
         """empty stdout/err are not displayed (groupby type)"""
         self.minimum_engines(1)
         
         v = self.client[:]
         ar = v.execute("print (5555)")
         ar.get(5)
         with capture_output() as io:
             ar.display_outputs()
         self.assertEqual(io.stderr, '')
         self.assertEqual(io.stdout.count('5555'), len(v), io.stdout)
         self.assertFalse('\n\n' in io.stdout, io.stdout)
         self.assertEqual(io.stdout.count('[stdout:'), len(v), io.stdout)
 
         ar = v.execute("a=5")
         ar.get(5)
         with capture_output() as io:
             ar.display_outputs()
         self.assertEqual(io.stderr, '')
         self.assertEqual(io.stdout, '')
 
     def test_display_empty_streams_engine(self):
         """empty stdout/err are not displayed (groupby engine)"""
         self.minimum_engines(1)
         
         v = self.client[:]
         ar = v.execute("print (5555)")
         ar.get(5)
         with capture_output() as io:
             ar.display_outputs('engine')
         self.assertEqual(io.stderr, '')
         self.assertEqual(io.stdout.count('5555'), len(v), io.stdout)
         self.assertFalse('\n\n' in io.stdout, io.stdout)
         self.assertEqual(io.stdout.count('[stdout:'), len(v), io.stdout)
 
         ar = v.execute("a=5")
         ar.get(5)
         with capture_output() as io:
             ar.display_outputs('engine')
         self.assertEqual(io.stderr, '')
         self.assertEqual(io.stdout, '')
     
     def test_await_data(self):
         """asking for ar.data flushes outputs"""
         self.minimum_engines(1)
         
         v = self.client[-1]
         ar = v.execute('\n'.join([
             "import time",
             "from IPython.kernel.zmq.datapub import publish_data",
             "for i in range(5):",
             "    publish_data(dict(i=i))",
             "    time.sleep(0.1)",
         ]), block=False)
         found = set()
         tic = time.time()
         # timeout after 10s
         while time.time() <= tic + 10:
             if ar.data:
                 found.add(ar.data['i'])
                 if ar.data['i'] == 4:
                     break
             time.sleep(0.05)
         
         ar.get(5)
         nt.assert_in(4, found)
         self.assertTrue(len(found) > 1, "should have seen data multiple times, but got: %s" % found)
     
     def test_not_single_result(self):
         save_build = self.client._build_targets
         def single_engine(*a, **kw):
             idents, targets = save_build(*a, **kw)
             return idents[:1], targets[:1]
         ids = single_engine('all')[1]
         self.client._build_targets = single_engine
         for targets in ('all', None, ids):
             dv = self.client.direct_view(targets=targets)
             ar = dv.apply_async(lambda : 5)
             self.assertEqual(ar.get(10), [5])
         self.client._build_targets = save_build
 
 

IPython/parallel/tests/test_client.py

@@ -1,517 +1,517 b''
 """Tests for parallel client.py
 
 Authors:
 
 * Min RK
 """
 
 #-------------------------------------------------------------------------------
 #  Copyright (C) 2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-------------------------------------------------------------------------------
 
 #-------------------------------------------------------------------------------
 # Imports
 #-------------------------------------------------------------------------------
 
 from __future__ import division
 
 import time
 from datetime import datetime
 from tempfile import mktemp
 
 import zmq
 
 from IPython import parallel
 from IPython.parallel.client import client as clientmod
 from IPython.parallel import error
 from IPython.parallel import AsyncResult, AsyncHubResult
 from IPython.parallel import LoadBalancedView, DirectView
 
 from .clienttest import ClusterTestCase, segfault, wait, add_engines
 
 def setup():
     add_engines(4, total=True)
 
 class TestClient(ClusterTestCase):
     
     def test_ids(self):
         n = len(self.client.ids)
         self.add_engines(2)
         self.assertEqual(len(self.client.ids), n+2)
     
     def test_view_indexing(self):
         """test index access for views"""
         self.minimum_engines(4)
         targets = self.client._build_targets('all')[-1]
         v = self.client[:]
         self.assertEqual(v.targets, targets)
         t = self.client.ids[2]
         v = self.client[t]
         self.assertTrue(isinstance(v, DirectView))
         self.assertEqual(v.targets, t)
         t = self.client.ids[2:4]
         v = self.client[t]
         self.assertTrue(isinstance(v, DirectView))
         self.assertEqual(v.targets, t)
         v = self.client[::2]
         self.assertTrue(isinstance(v, DirectView))
         self.assertEqual(v.targets, targets[::2])
         v = self.client[1::3]
         self.assertTrue(isinstance(v, DirectView))
         self.assertEqual(v.targets, targets[1::3])
         v = self.client[:-3]
         self.assertTrue(isinstance(v, DirectView))
         self.assertEqual(v.targets, targets[:-3])
         v = self.client[-1]
         self.assertTrue(isinstance(v, DirectView))
         self.assertEqual(v.targets, targets[-1])
         self.assertRaises(TypeError, lambda : self.client[None])
     
     def test_lbview_targets(self):
         """test load_balanced_view targets"""
         v = self.client.load_balanced_view()
         self.assertEqual(v.targets, None)
         v = self.client.load_balanced_view(-1)
         self.assertEqual(v.targets, [self.client.ids[-1]])
         v = self.client.load_balanced_view('all')
         self.assertEqual(v.targets, None)
     
     def test_dview_targets(self):
         """test direct_view targets"""
         v = self.client.direct_view()
         self.assertEqual(v.targets, 'all')
         v = self.client.direct_view('all')
         self.assertEqual(v.targets, 'all')
         v = self.client.direct_view(-1)
         self.assertEqual(v.targets, self.client.ids[-1])
     
     def test_lazy_all_targets(self):
         """test lazy evaluation of rc.direct_view('all')"""
         v = self.client.direct_view()
         self.assertEqual(v.targets, 'all')
         
         def double(x):
             return x*2
-        seq = range(100)
+        seq = list(range(100))
         ref = [ double(x) for x in seq ]
         
         # add some engines, which should be used
         self.add_engines(1)
         n1 = len(self.client.ids)
         
         # simple apply
         r = v.apply_sync(lambda : 1)
         self.assertEqual(r, [1] * n1)
         
         # map goes through remotefunction
         r = v.map_sync(double, seq)
         self.assertEqual(r, ref)
 
         # add a couple more engines, and try again
         self.add_engines(2)
         n2 = len(self.client.ids)
         self.assertNotEqual(n2, n1)
         
         # apply
         r = v.apply_sync(lambda : 1)
         self.assertEqual(r, [1] * n2)
         
         # map
         r = v.map_sync(double, seq)
         self.assertEqual(r, ref)
     
     def test_targets(self):
         """test various valid targets arguments"""
         build = self.client._build_targets
         ids = self.client.ids
         idents,targets = build(None)
         self.assertEqual(ids, targets)
     
     def test_clear(self):
         """test clear behavior"""
         self.minimum_engines(2)
         v = self.client[:]
         v.block=True
         v.push(dict(a=5))
         v.pull('a')
         id0 = self.client.ids[-1]
         self.client.clear(targets=id0, block=True)
         a = self.client[:-1].get('a')
         self.assertRaisesRemote(NameError, self.client[id0].get, 'a')
         self.client.clear(block=True)
         for i in self.client.ids:
             self.assertRaisesRemote(NameError, self.client[i].get, 'a')
     
     def test_get_result(self):
         """test getting results from the Hub."""
         c = clientmod.Client(profile='iptest')
         t = c.ids[-1]
         ar = c[t].apply_async(wait, 1)
         # give the monitor time to notice the message
         time.sleep(.25)
         ahr = self.client.get_result(ar.msg_ids[0])
         self.assertTrue(isinstance(ahr, AsyncHubResult))
         self.assertEqual(ahr.get(), ar.get())
         ar2 = self.client.get_result(ar.msg_ids[0])
         self.assertFalse(isinstance(ar2, AsyncHubResult))
         c.close()
     
     def test_get_execute_result(self):
         """test getting execute results from the Hub."""
         c = clientmod.Client(profile='iptest')
         t = c.ids[-1]
         cell = '\n'.join([
             'import time',
             'time.sleep(0.25)',
             '5'
         ])
         ar = c[t].execute("import time; time.sleep(1)", silent=False)
         # give the monitor time to notice the message
         time.sleep(.25)
         ahr = self.client.get_result(ar.msg_ids[0])
         self.assertTrue(isinstance(ahr, AsyncHubResult))
         self.assertEqual(ahr.get().pyout, ar.get().pyout)
         ar2 = self.client.get_result(ar.msg_ids[0])
         self.assertFalse(isinstance(ar2, AsyncHubResult))
         c.close()
     
     def test_ids_list(self):
         """test client.ids"""
         ids = self.client.ids
         self.assertEqual(ids, self.client._ids)
         self.assertFalse(ids is self.client._ids)
         ids.remove(ids[-1])
         self.assertNotEqual(ids, self.client._ids)
     
     def test_queue_status(self):
         ids = self.client.ids
         id0 = ids[0]
         qs = self.client.queue_status(targets=id0)
         self.assertTrue(isinstance(qs, dict))
         self.assertEqual(sorted(qs.keys()), ['completed', 'queue', 'tasks'])
         allqs = self.client.queue_status()
         self.assertTrue(isinstance(allqs, dict))
         intkeys = list(allqs.keys())
         intkeys.remove('unassigned')
         self.assertEqual(sorted(intkeys), sorted(self.client.ids))
         unassigned = allqs.pop('unassigned')
         for eid,qs in allqs.items():
             self.assertTrue(isinstance(qs, dict))
             self.assertEqual(sorted(qs.keys()), ['completed', 'queue', 'tasks'])
 
     def test_shutdown(self):
         ids = self.client.ids
         id0 = ids[0]
         self.client.shutdown(id0, block=True)
         while id0 in self.client.ids:
             time.sleep(0.1)
             self.client.spin()
         
         self.assertRaises(IndexError, lambda : self.client[id0])
         
     def test_result_status(self):
         pass
         # to be written
     
     def test_db_query_dt(self):
         """test db query by date"""
         hist = self.client.hub_history()
         middle = self.client.db_query({'msg_id' : hist[len(hist)//2]})[0]
         tic = middle['submitted']
         before = self.client.db_query({'submitted' : {'$lt' : tic}})
         after = self.client.db_query({'submitted' : {'$gte' : tic}})
         self.assertEqual(len(before)+len(after),len(hist))
         for b in before:
             self.assertTrue(b['submitted'] < tic)
         for a in after:
             self.assertTrue(a['submitted'] >= tic)
         same = self.client.db_query({'submitted' : tic})
         for s in same:
             self.assertTrue(s['submitted'] == tic)
     
     def test_db_query_keys(self):
         """test extracting subset of record keys"""
         found = self.client.db_query({'msg_id': {'$ne' : ''}},keys=['submitted', 'completed'])
         for rec in found:
             self.assertEqual(set(rec.keys()), set(['msg_id', 'submitted', 'completed']))
     
     def test_db_query_default_keys(self):
         """default db_query excludes buffers"""
         found = self.client.db_query({'msg_id': {'$ne' : ''}})
         for rec in found:
             keys = set(rec.keys())
             self.assertFalse('buffers' in keys, "'buffers' should not be in: %s" % keys)
             self.assertFalse('result_buffers' in keys, "'result_buffers' should not be in: %s" % keys)
     
     def test_db_query_msg_id(self):
         """ensure msg_id is always in db queries"""
         found = self.client.db_query({'msg_id': {'$ne' : ''}},keys=['submitted', 'completed'])
         for rec in found:
             self.assertTrue('msg_id' in rec.keys())
         found = self.client.db_query({'msg_id': {'$ne' : ''}},keys=['submitted'])
         for rec in found:
             self.assertTrue('msg_id' in rec.keys())
         found = self.client.db_query({'msg_id': {'$ne' : ''}},keys=['msg_id'])
         for rec in found:
             self.assertTrue('msg_id' in rec.keys())
     
     def test_db_query_get_result(self):
         """pop in db_query shouldn't pop from result itself"""
         self.client[:].apply_sync(lambda : 1)
         found = self.client.db_query({'msg_id': {'$ne' : ''}})
         rc2 = clientmod.Client(profile='iptest')
         # If this bug is not fixed, this call will hang:
         ar = rc2.get_result(self.client.history[-1])
         ar.wait(2)
         self.assertTrue(ar.ready())
         ar.get()
         rc2.close()
     
     def test_db_query_in(self):
         """test db query with '$in','$nin' operators"""
         hist = self.client.hub_history()
         even = hist[::2]
         odd = hist[1::2]
         recs = self.client.db_query({ 'msg_id' : {'$in' : even}})
         found = [ r['msg_id'] for r in recs ]
         self.assertEqual(set(even), set(found))
         recs = self.client.db_query({ 'msg_id' : {'$nin' : even}})
         found = [ r['msg_id'] for r in recs ]
         self.assertEqual(set(odd), set(found))
     
     def test_hub_history(self):
         hist = self.client.hub_history()
         recs = self.client.db_query({ 'msg_id' : {"$ne":''}})
         recdict = {}
         for rec in recs:
             recdict[rec['msg_id']] = rec
         
         latest = datetime(1984,1,1)
         for msg_id in hist:
             rec = recdict[msg_id]
             newt = rec['submitted']
             self.assertTrue(newt >= latest)
             latest = newt
         ar = self.client[-1].apply_async(lambda : 1)
         ar.get()
         time.sleep(0.25)
         self.assertEqual(self.client.hub_history()[-1:],ar.msg_ids)
     
     def _wait_for_idle(self):
         """wait for an engine to become idle, according to the Hub"""
         rc = self.client
         
         # step 1. wait for all requests to be noticed
         # timeout 5s, polling every 100ms
         msg_ids = set(rc.history)
         hub_hist = rc.hub_history()
         for i in range(50):
             if msg_ids.difference(hub_hist):
                 time.sleep(0.1)
                 hub_hist = rc.hub_history()
             else:
                 break
         
         self.assertEqual(len(msg_ids.difference(hub_hist)), 0)
         
         # step 2. wait for all requests to be done
         # timeout 5s, polling every 100ms
         qs = rc.queue_status()
         for i in range(50):
             if qs['unassigned'] or any(qs[eid]['tasks'] for eid in rc.ids):
                 time.sleep(0.1)
                 qs = rc.queue_status()
             else:
                 break
         
         # ensure Hub up to date:
         self.assertEqual(qs['unassigned'], 0)
         for eid in rc.ids:
             self.assertEqual(qs[eid]['tasks'], 0)
     
     
     def test_resubmit(self):
         def f():
             import random
             return random.random()
         v = self.client.load_balanced_view()
         ar = v.apply_async(f)
         r1 = ar.get(1)
         # give the Hub a chance to notice:
         self._wait_for_idle()
         ahr = self.client.resubmit(ar.msg_ids)
         r2 = ahr.get(1)
         self.assertFalse(r1 == r2)
 
     def test_resubmit_chain(self):
         """resubmit resubmitted tasks"""
         v = self.client.load_balanced_view()
         ar = v.apply_async(lambda x: x, 'x'*1024)
         ar.get()
         self._wait_for_idle()
         ars = [ar]
         
         for i in range(10):
             ar = ars[-1]
             ar2 = self.client.resubmit(ar.msg_ids)
         
         [ ar.get() for ar in ars ]
 
     def test_resubmit_header(self):
         """resubmit shouldn't clobber the whole header"""
         def f():
             import random
             return random.random()
         v = self.client.load_balanced_view()
         v.retries = 1
         ar = v.apply_async(f)
         r1 = ar.get(1)
         # give the Hub a chance to notice:
         self._wait_for_idle()
         ahr = self.client.resubmit(ar.msg_ids)
         ahr.get(1)
         time.sleep(0.5)
         records = self.client.db_query({'msg_id': {'$in': ar.msg_ids + ahr.msg_ids}}, keys='header')
         h1,h2 = [ r['header'] for r in records ]
         for key in set(h1.keys()).union(set(h2.keys())):
             if key in ('msg_id', 'date'):
                 self.assertNotEqual(h1[key], h2[key])
             else:
                 self.assertEqual(h1[key], h2[key])
 
     def test_resubmit_aborted(self):
         def f():
             import random
             return random.random()
         v = self.client.load_balanced_view()
         # restrict to one engine, so we can put a sleep
         # ahead of the task, so it will get aborted
         eid = self.client.ids[-1]
         v.targets = [eid]
         sleep = v.apply_async(time.sleep, 0.5)
         ar = v.apply_async(f)
         ar.abort()
         self.assertRaises(error.TaskAborted, ar.get)
         # Give the Hub a chance to get up to date:
         self._wait_for_idle()
         ahr = self.client.resubmit(ar.msg_ids)
         r2 = ahr.get(1)
 
     def test_resubmit_inflight(self):
         """resubmit of inflight task"""
         v = self.client.load_balanced_view()
         ar = v.apply_async(time.sleep,1)
         # give the message a chance to arrive
         time.sleep(0.2)
         ahr = self.client.resubmit(ar.msg_ids)
         ar.get(2)
         ahr.get(2)
 
     def test_resubmit_badkey(self):
         """ensure KeyError on resubmit of nonexistant task"""
         self.assertRaisesRemote(KeyError, self.client.resubmit, ['invalid'])
 
     def test_purge_hub_results(self):
         # ensure there are some tasks
         for i in range(5):
             self.client[:].apply_sync(lambda : 1)
         # Wait for the Hub to realise the result is done:
         # This prevents a race condition, where we
         # might purge a result the Hub still thinks is pending.
         self._wait_for_idle()
         rc2 = clientmod.Client(profile='iptest')
         hist = self.client.hub_history()
         ahr = rc2.get_result([hist[-1]])
         ahr.wait(10)
         self.client.purge_hub_results(hist[-1])
         newhist = self.client.hub_history()
         self.assertEqual(len(newhist)+1,len(hist))
         rc2.spin()
         rc2.close()
 
     def test_purge_local_results(self):
         # ensure there are some tasks
         res = []
         for i in range(5):
             res.append(self.client[:].apply_async(lambda : 1))
         self._wait_for_idle()
         self.client.wait(10) # wait for the results to come back
         before = len(self.client.results)
         self.assertEqual(len(self.client.metadata),before)
         self.client.purge_local_results(res[-1])
         self.assertEqual(len(self.client.results),before-len(res[-1]), msg="Not removed from results")
         self.assertEqual(len(self.client.metadata),before-len(res[-1]), msg="Not removed from metadata")
         
     def test_purge_all_hub_results(self):
         self.client.purge_hub_results('all')
         hist = self.client.hub_history()
         self.assertEqual(len(hist), 0)
 
     def test_purge_all_local_results(self):
         self.client.purge_local_results('all')
         self.assertEqual(len(self.client.results), 0, msg="Results not empty")
         self.assertEqual(len(self.client.metadata), 0, msg="metadata not empty")
 
     def test_purge_all_results(self):
         # ensure there are some tasks
         for i in range(5):
             self.client[:].apply_sync(lambda : 1)
         self.client.wait(10)
         self._wait_for_idle()
         self.client.purge_results('all')
         self.assertEqual(len(self.client.results), 0, msg="Results not empty")
         self.assertEqual(len(self.client.metadata), 0, msg="metadata not empty")
         hist = self.client.hub_history()
         self.assertEqual(len(hist), 0, msg="hub history not empty")
         
     def test_purge_everything(self):
         # ensure there are some tasks
         for i in range(5):
             self.client[:].apply_sync(lambda : 1)
         self.client.wait(10)
         self._wait_for_idle()
         self.client.purge_everything()
         # The client results
         self.assertEqual(len(self.client.results), 0, msg="Results not empty")
         self.assertEqual(len(self.client.metadata), 0, msg="metadata not empty")
         # The client "bookkeeping"
         self.assertEqual(len(self.client.session.digest_history), 0, msg="session digest not empty")
         self.assertEqual(len(self.client.history), 0, msg="client history not empty")
         # the hub results
         hist = self.client.hub_history()
         self.assertEqual(len(hist), 0, msg="hub history not empty")
         
     
     def test_spin_thread(self):
         self.client.spin_thread(0.01)
         ar = self.client[-1].apply_async(lambda : 1)
         time.sleep(0.1)
         self.assertTrue(ar.wall_time < 0.1,
             "spin should have kept wall_time < 0.1, but got %f" % ar.wall_time
         )
     
     def test_stop_spin_thread(self):
         self.client.spin_thread(0.01)
         self.client.stop_spin_thread()
         ar = self.client[-1].apply_async(lambda : 1)
         time.sleep(0.15)
         self.assertTrue(ar.wall_time > 0.1,
             "Shouldn't be spinning, but got wall_time=%f" % ar.wall_time
         )
     
     def test_activate(self):
         ip = get_ipython()
         magics = ip.magics_manager.magics
         self.assertTrue('px' in magics['line'])
         self.assertTrue('px' in magics['cell'])
         v0 = self.client.activate(-1, '0')
         self.assertTrue('px0' in magics['line'])
         self.assertTrue('px0' in magics['cell'])
         self.assertEqual(v0.targets, self.client.ids[-1])
         v0 = self.client.activate('all', 'all')
         self.assertTrue('pxall' in magics['line'])
         self.assertTrue('pxall' in magics['cell'])
         self.assertEqual(v0.targets, 'all')

IPython/parallel/tests/test_dependency.py

@@ -1,136 +1,136 b''
 """Tests for dependency.py
 
 Authors:
 
 * Min RK
 """
 
 __docformat__ = "restructuredtext en"
 
 #-------------------------------------------------------------------------------
 #  Copyright (C) 2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-------------------------------------------------------------------------------
 
 #-------------------------------------------------------------------------------
 # Imports
 #-------------------------------------------------------------------------------
 
 # import
 import os
 
 from IPython.utils.pickleutil import can, uncan
 
 import IPython.parallel as pmod
 from IPython.parallel.util import interactive
 
 from IPython.parallel.tests import add_engines
 from .clienttest import ClusterTestCase
 
 def setup():
     add_engines(1, total=True)
 
 @pmod.require('time')
 def wait(n):
     time.sleep(n)
     return n
 
 @pmod.interactive
 def func(x):
     return x*x
 
-mixed = map(str, range(10))
-completed = map(str, range(0,10,2))
-failed = map(str, range(1,10,2))
+mixed = list(map(str, range(10)))
+completed = list(map(str, range(0,10,2)))
+failed = list(map(str, range(1,10,2)))
 
 class DependencyTest(ClusterTestCase):
     
     def setUp(self):
         ClusterTestCase.setUp(self)
         self.user_ns = {'__builtins__' : __builtins__}
         self.view = self.client.load_balanced_view()
         self.dview = self.client[-1]
         self.succeeded = set(map(str, range(0,25,2)))
         self.failed = set(map(str, range(1,25,2)))
     
     def assertMet(self, dep):
         self.assertTrue(dep.check(self.succeeded, self.failed), "Dependency should be met")
         
     def assertUnmet(self, dep):
         self.assertFalse(dep.check(self.succeeded, self.failed), "Dependency should not be met")
         
     def assertUnreachable(self, dep):
         self.assertTrue(dep.unreachable(self.succeeded, self.failed), "Dependency should be unreachable")
     
     def assertReachable(self, dep):
         self.assertFalse(dep.unreachable(self.succeeded, self.failed), "Dependency should be reachable")
     
     def cancan(self, f):
         """decorator to pass through canning into self.user_ns"""
         return uncan(can(f), self.user_ns)
     
     def test_require_imports(self):
         """test that @require imports names"""
         @self.cancan
-        @pmod.require('urllib')
+        @pmod.require('base64')
         @interactive
-        def encode(dikt):
-            return urllib.urlencode(dikt)
+        def encode(arg):
+            return base64.b64encode(arg)
         # must pass through canning to properly connect namespaces
-        self.assertEqual(encode(dict(a=5)), 'a=5')
+        self.assertEqual(encode(b'foo'), b'Zm9v')
     
     def test_success_only(self):
         dep = pmod.Dependency(mixed, success=True, failure=False)
         self.assertUnmet(dep)
         self.assertUnreachable(dep)
         dep.all=False
         self.assertMet(dep)
         self.assertReachable(dep)
         dep = pmod.Dependency(completed, success=True, failure=False)
         self.assertMet(dep)
         self.assertReachable(dep)
         dep.all=False
         self.assertMet(dep)
         self.assertReachable(dep)
 
     def test_failure_only(self):
         dep = pmod.Dependency(mixed, success=False, failure=True)
         self.assertUnmet(dep)
         self.assertUnreachable(dep)
         dep.all=False
         self.assertMet(dep)
         self.assertReachable(dep)
         dep = pmod.Dependency(completed, success=False, failure=True)
         self.assertUnmet(dep)
         self.assertUnreachable(dep)
         dep.all=False
         self.assertUnmet(dep)
         self.assertUnreachable(dep)
     
     def test_require_function(self):
         
         @pmod.interactive
         def bar(a):
             return func(a)
 
         @pmod.require(func)
         @pmod.interactive
         def bar2(a):
             return func(a)
         
         self.client[:].clear()
         self.assertRaisesRemote(NameError, self.view.apply_sync, bar, 5)
         ar = self.view.apply_async(bar2, 5)
         self.assertEqual(ar.get(5), func(5))
 
     def test_require_object(self):
         
         @pmod.require(foo=func)
         @pmod.interactive
         def bar(a):
             return foo(a)
 
         ar = self.view.apply_async(bar, 5)
         self.assertEqual(ar.get(5), func(5))

IPython/parallel/tests/test_lbview.py

@@ -1,221 +1,221 b''
 # -*- coding: utf-8 -*-
 """test LoadBalancedView objects
 
 Authors:
 
 * Min RK
 """
 #-------------------------------------------------------------------------------
 #  Copyright (C) 2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-------------------------------------------------------------------------------
 
 #-------------------------------------------------------------------------------
 # Imports
 #-------------------------------------------------------------------------------
 
 import sys
 import time
 
 import zmq
 from nose import SkipTest
 from nose.plugins.attrib import attr
 
 from IPython import parallel  as pmod
 from IPython.parallel import error
 
 from IPython.parallel.tests import add_engines
 
 from .clienttest import ClusterTestCase, crash, wait, skip_without
 
 def setup():
     add_engines(3, total=True)
 
 class TestLoadBalancedView(ClusterTestCase):
 
     def setUp(self):
         ClusterTestCase.setUp(self)
         self.view = self.client.load_balanced_view()
 
     @attr('crash')
     def test_z_crash_task(self):
         """test graceful handling of engine death (balanced)"""
         # self.add_engines(1)
         ar = self.view.apply_async(crash)
         self.assertRaisesRemote(error.EngineError, ar.get, 10)
         eid = ar.engine_id
         tic = time.time()
         while eid in self.client.ids and time.time()-tic < 5:
             time.sleep(.01)
             self.client.spin()
         self.assertFalse(eid in self.client.ids, "Engine should have died")
 
     def test_map(self):
         def f(x):
             return x**2
-        data = range(16)
+        data = list(range(16))
         r = self.view.map_sync(f, data)
-        self.assertEqual(r, map(f, data))
+        self.assertEqual(r, list(map(f, data)))
     
     def test_map_generator(self):
         def f(x):
             return x**2
         
-        data = range(16)
+        data = list(range(16))
         r = self.view.map_sync(f, iter(data))
-        self.assertEqual(r, map(f, iter(data)))
+        self.assertEqual(r, list(map(f, iter(data))))
     
     def test_map_short_first(self):
         def f(x,y):
             if y is None:
                 return y
             if x is None:
                 return x
             return x*y
-        data = range(10)
-        data2 = range(4)
+        data = list(range(10))
+        data2 = list(range(4))
         
         r = self.view.map_sync(f, data, data2)
-        self.assertEqual(r, map(f, data, data2))
+        self.assertEqual(r, list(map(f, data, data2)))
 
     def test_map_short_last(self):
         def f(x,y):
             if y is None:
                 return y
             if x is None:
                 return x
             return x*y
-        data = range(4)
-        data2 = range(10)
+        data = list(range(4))
+        data2 = list(range(10))
         
         r = self.view.map_sync(f, data, data2)
-        self.assertEqual(r, map(f, data, data2))
+        self.assertEqual(r, list(map(f, data, data2)))
 
     def test_map_unordered(self):
         def f(x):
             return x**2
         def slow_f(x):
             import time
             time.sleep(0.05*x)
             return x**2
-        data = range(16,0,-1)
-        reference = map(f, data)
+        data = list(range(16,0,-1))
+        reference = list(map(f, data))
         
         amr = self.view.map_async(slow_f, data, ordered=False)
         self.assertTrue(isinstance(amr, pmod.AsyncMapResult))
         # check individual elements, retrieved as they come
         # list comprehension uses __iter__
         astheycame = [ r for r in amr ]
         # Ensure that at least one result came out of order:
         self.assertNotEqual(astheycame, reference, "should not have preserved order")
         self.assertEqual(sorted(astheycame, reverse=True), reference, "result corrupted")
 
     def test_map_ordered(self):
         def f(x):
             return x**2
         def slow_f(x):
             import time
             time.sleep(0.05*x)
             return x**2
-        data = range(16,0,-1)
-        reference = map(f, data)
+        data = list(range(16,0,-1))
+        reference = list(map(f, data))
         
         amr = self.view.map_async(slow_f, data)
         self.assertTrue(isinstance(amr, pmod.AsyncMapResult))
         # check individual elements, retrieved as they come
         # list(amr) uses __iter__
         astheycame = list(amr)
         # Ensure that results came in order
         self.assertEqual(astheycame, reference)
         self.assertEqual(amr.result, reference)
 
     def test_map_iterable(self):
         """test map on iterables (balanced)"""
         view = self.view
         # 101 is prime, so it won't be evenly distributed
         arr = range(101)
         # so that it will be an iterator, even in Python 3
         it = iter(arr)
         r = view.map_sync(lambda x:x, arr)
         self.assertEqual(r, list(arr))
 
     
     def test_abort(self):
         view = self.view
         ar = self.client[:].apply_async(time.sleep, .5)
         ar = self.client[:].apply_async(time.sleep, .5)
         time.sleep(0.2)
         ar2 = view.apply_async(lambda : 2)
         ar3 = view.apply_async(lambda : 3)
         view.abort(ar2)
         view.abort(ar3.msg_ids)
         self.assertRaises(error.TaskAborted, ar2.get)
         self.assertRaises(error.TaskAborted, ar3.get)
 
     def test_retries(self):
         self.minimum_engines(3)
         view = self.view
         def fail():
             assert False
         for r in range(len(self.client)-1):
             with view.temp_flags(retries=r):
                 self.assertRaisesRemote(AssertionError, view.apply_sync, fail)
 
         with view.temp_flags(retries=len(self.client), timeout=0.1):
             self.assertRaisesRemote(error.TaskTimeout, view.apply_sync, fail)
 
     def test_short_timeout(self):
         self.minimum_engines(2)
         view = self.view
         def fail():
             import time
             time.sleep(0.25)
             assert False
         with view.temp_flags(retries=1, timeout=0.01):
             self.assertRaisesRemote(AssertionError, view.apply_sync, fail)
 
     def test_invalid_dependency(self):
         view = self.view
         with view.temp_flags(after='12345'):
             self.assertRaisesRemote(error.InvalidDependency, view.apply_sync, lambda : 1)
 
     def test_impossible_dependency(self):
         self.minimum_engines(2)
         view = self.client.load_balanced_view()
         ar1 = view.apply_async(lambda : 1)
         ar1.get()
         e1 = ar1.engine_id
         e2 = e1
         while e2 == e1:
             ar2 = view.apply_async(lambda : 1)
             ar2.get()
             e2 = ar2.engine_id
 
         with view.temp_flags(follow=[ar1, ar2]):
             self.assertRaisesRemote(error.ImpossibleDependency, view.apply_sync, lambda : 1)
 
 
     def test_follow(self):
         ar = self.view.apply_async(lambda : 1)
         ar.get()
         ars = []
         first_id = ar.engine_id
 
         self.view.follow = ar
         for i in range(5):
             ars.append(self.view.apply_async(lambda : 1))
         self.view.wait(ars)
         for ar in ars:
             self.assertEqual(ar.engine_id, first_id)
 
     def test_after(self):
         view = self.view
         ar = view.apply_async(time.sleep, 0.5)
         with view.temp_flags(after=ar):
             ar2 = view.apply_async(lambda : 1)
 
         ar.wait()
         ar2.wait()
         self.assertTrue(ar2.started >= ar.completed, "%s not >= %s"%(ar.started, ar.completed))

IPython/parallel/tests/test_view.py

@@ -1,835 +1,835 b''
 # -*- coding: utf-8 -*-
 """test View objects
 
 Authors:
 
 * Min RK
 """
 #-------------------------------------------------------------------------------
 #  Copyright (C) 2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-------------------------------------------------------------------------------
 
 #-------------------------------------------------------------------------------
 # Imports
 #-------------------------------------------------------------------------------
 
 import base64
 import sys
 import platform
 import time
 from collections import namedtuple
 from tempfile import mktemp
 
 import zmq
 from nose.plugins.attrib import attr
 
 from IPython.testing import decorators as dec
 from IPython.utils.io import capture_output
 from IPython.utils.py3compat import unicode_type
 
 from IPython import parallel  as pmod
 from IPython.parallel import error
 from IPython.parallel import AsyncResult, AsyncHubResult, AsyncMapResult
 from IPython.parallel.util import interactive
 
 from IPython.parallel.tests import add_engines
 
 from .clienttest import ClusterTestCase, crash, wait, skip_without
 
 def setup():
     add_engines(3, total=True)
 
 point = namedtuple("point", "x y")
 
 class TestView(ClusterTestCase):
     
     def setUp(self):
         # On Win XP, wait for resource cleanup, else parallel test group fails
         if platform.system() == "Windows" and platform.win32_ver()[0] == "XP":
             # 1 sec fails. 1.5 sec seems ok. Using 2 sec for margin of safety
             time.sleep(2)
         super(TestView, self).setUp()
 
     @attr('crash')
     def test_z_crash_mux(self):
         """test graceful handling of engine death (direct)"""
         # self.add_engines(1)
         eid = self.client.ids[-1]
         ar = self.client[eid].apply_async(crash)
         self.assertRaisesRemote(error.EngineError, ar.get, 10)
         eid = ar.engine_id
         tic = time.time()
         while eid in self.client.ids and time.time()-tic < 5:
             time.sleep(.01)
             self.client.spin()
         self.assertFalse(eid in self.client.ids, "Engine should have died")
     
     def test_push_pull(self):
         """test pushing and pulling"""
-        data = dict(a=10, b=1.05, c=range(10), d={'e':(1,2),'f':'hi'})
+        data = dict(a=10, b=1.05, c=list(range(10)), d={'e':(1,2),'f':'hi'})
         t = self.client.ids[-1]
         v = self.client[t]
         push = v.push
         pull = v.pull
         v.block=True
         nengines = len(self.client)
         push({'data':data})
         d = pull('data')
         self.assertEqual(d, data)
         self.client[:].push({'data':data})
         d = self.client[:].pull('data', block=True)
         self.assertEqual(d, nengines*[data])
         ar = push({'data':data}, block=False)
         self.assertTrue(isinstance(ar, AsyncResult))
         r = ar.get()
         ar = self.client[:].pull('data', block=False)
         self.assertTrue(isinstance(ar, AsyncResult))
         r = ar.get()
         self.assertEqual(r, nengines*[data])
         self.client[:].push(dict(a=10,b=20))
         r = self.client[:].pull(('a','b'), block=True)
         self.assertEqual(r, nengines*[[10,20]])
     
     def test_push_pull_function(self):
         "test pushing and pulling functions"
         def testf(x):
             return 2.0*x
         
         t = self.client.ids[-1]
         v = self.client[t]
         v.block=True
         push = v.push
         pull = v.pull
         execute = v.execute
         push({'testf':testf})
         r = pull('testf')
         self.assertEqual(r(1.0), testf(1.0))
         execute('r = testf(10)')
         r = pull('r')
         self.assertEqual(r, testf(10))
         ar = self.client[:].push({'testf':testf}, block=False)
         ar.get()
         ar = self.client[:].pull('testf', block=False)
         rlist = ar.get()
         for r in rlist:
             self.assertEqual(r(1.0), testf(1.0))
         execute("def g(x): return x*x")
         r = pull(('testf','g'))
         self.assertEqual((r[0](10),r[1](10)), (testf(10), 100))
     
     def test_push_function_globals(self):
         """test that pushed functions have access to globals"""
         @interactive
         def geta():
             return a
         # self.add_engines(1)
         v = self.client[-1]
         v.block=True
         v['f'] = geta
         self.assertRaisesRemote(NameError, v.execute, 'b=f()')
         v.execute('a=5')
         v.execute('b=f()')
         self.assertEqual(v['b'], 5)
     
     def test_push_function_defaults(self):
         """test that pushed functions preserve default args"""
         def echo(a=10):
             return a
         v = self.client[-1]
         v.block=True
         v['f'] = echo
         v.execute('b=f()')
         self.assertEqual(v['b'], 10)
 
     def test_get_result(self):
         """test getting results from the Hub."""
         c = pmod.Client(profile='iptest')
         # self.add_engines(1)
         t = c.ids[-1]
         v = c[t]
         v2 = self.client[t]
         ar = v.apply_async(wait, 1)
         # give the monitor time to notice the message
         time.sleep(.25)
         ahr = v2.get_result(ar.msg_ids[0])
         self.assertTrue(isinstance(ahr, AsyncHubResult))
         self.assertEqual(ahr.get(), ar.get())
         ar2 = v2.get_result(ar.msg_ids[0])
         self.assertFalse(isinstance(ar2, AsyncHubResult))
         c.spin()
         c.close()
     
     def test_run_newline(self):
         """test that run appends newline to files"""
         tmpfile = mktemp()
         with open(tmpfile, 'w') as f:
             f.write("""def g():
                 return 5
                 """)
         v = self.client[-1]
         v.run(tmpfile, block=True)
         self.assertEqual(v.apply_sync(lambda f: f(), pmod.Reference('g')), 5)
 
     def test_apply_tracked(self):
         """test tracking for apply"""
         # self.add_engines(1)
         t = self.client.ids[-1]
         v = self.client[t]
         v.block=False
         def echo(n=1024*1024, **kwargs):
             with v.temp_flags(**kwargs):
                 return v.apply(lambda x: x, 'x'*n)
         ar = echo(1, track=False)
         self.assertTrue(isinstance(ar._tracker, zmq.MessageTracker))
         self.assertTrue(ar.sent)
         ar = echo(track=True)
         self.assertTrue(isinstance(ar._tracker, zmq.MessageTracker))
         self.assertEqual(ar.sent, ar._tracker.done)
         ar._tracker.wait()
         self.assertTrue(ar.sent)
         
     def test_push_tracked(self):
         t = self.client.ids[-1]
         ns = dict(x='x'*1024*1024)
         v = self.client[t]
         ar = v.push(ns, block=False, track=False)
         self.assertTrue(isinstance(ar._tracker, zmq.MessageTracker))
         self.assertTrue(ar.sent)
         
         ar = v.push(ns, block=False, track=True)
         self.assertTrue(isinstance(ar._tracker, zmq.MessageTracker))
         ar._tracker.wait()
         self.assertEqual(ar.sent, ar._tracker.done)
         self.assertTrue(ar.sent)
         ar.get()
         
     def test_scatter_tracked(self):
         t = self.client.ids
         x='x'*1024*1024
         ar = self.client[t].scatter('x', x, block=False, track=False)
         self.assertTrue(isinstance(ar._tracker, zmq.MessageTracker))
         self.assertTrue(ar.sent)
         
         ar = self.client[t].scatter('x', x, block=False, track=True)
         self.assertTrue(isinstance(ar._tracker, zmq.MessageTracker))
         self.assertEqual(ar.sent, ar._tracker.done)
         ar._tracker.wait()
         self.assertTrue(ar.sent)
         ar.get()
     
     def test_remote_reference(self):
         v = self.client[-1]
         v['a'] = 123
         ra = pmod.Reference('a')
         b = v.apply_sync(lambda x: x, ra)
         self.assertEqual(b, 123)
 
 
     def test_scatter_gather(self):
         view = self.client[:]
-        seq1 = range(16)
+        seq1 = list(range(16))
         view.scatter('a', seq1)
         seq2 = view.gather('a', block=True)
         self.assertEqual(seq2, seq1)
         self.assertRaisesRemote(NameError, view.gather, 'asdf', block=True)
     
     @skip_without('numpy')
     def test_scatter_gather_numpy(self):
         import numpy
         from numpy.testing.utils import assert_array_equal
         view = self.client[:]
         a = numpy.arange(64)
         view.scatter('a', a, block=True)
         b = view.gather('a', block=True)
         assert_array_equal(b, a)
     
     def test_scatter_gather_lazy(self):
         """scatter/gather with targets='all'"""
         view = self.client.direct_view(targets='all')
-        x = range(64)
+        x = list(range(64))
         view.scatter('x', x)
         gathered = view.gather('x', block=True)
         self.assertEqual(gathered, x)
         
 
     @dec.known_failure_py3
     @skip_without('numpy')
     def test_push_numpy_nocopy(self):
         import numpy
         view = self.client[:]
         a = numpy.arange(64)
         view['A'] = a
         @interactive
         def check_writeable(x):
             return x.flags.writeable
         
         for flag in view.apply_sync(check_writeable, pmod.Reference('A')):
             self.assertFalse(flag, "array is writeable, push shouldn't have pickled it")
         
         view.push(dict(B=a))
         for flag in view.apply_sync(check_writeable, pmod.Reference('B')):
             self.assertFalse(flag, "array is writeable, push shouldn't have pickled it")
     
     @skip_without('numpy')
     def test_apply_numpy(self):
         """view.apply(f, ndarray)"""
         import numpy
         from numpy.testing.utils import assert_array_equal
         
         A = numpy.random.random((100,100))
         view = self.client[-1]
         for dt in [ 'int32', 'uint8', 'float32', 'float64' ]:
             B = A.astype(dt)
             C = view.apply_sync(lambda x:x, B)
             assert_array_equal(B,C)
     
     @skip_without('numpy')
     def test_push_pull_recarray(self):
         """push/pull recarrays"""
         import numpy
         from numpy.testing.utils import assert_array_equal
         
         view = self.client[-1]
         
         R = numpy.array([
             (1, 'hi', 0.),
             (2**30, 'there', 2.5),
             (-99999, 'world', -12345.6789),
         ], [('n', int), ('s', '|S10'), ('f', float)])
         
         view['RR'] = R
         R2 = view['RR']
         
         r_dtype, r_shape = view.apply_sync(interactive(lambda : (RR.dtype, RR.shape)))
         self.assertEqual(r_dtype, R.dtype)
         self.assertEqual(r_shape, R.shape)
         self.assertEqual(R2.dtype, R.dtype)
         self.assertEqual(R2.shape, R.shape)
         assert_array_equal(R2, R)
 
     @skip_without('pandas')
     def test_push_pull_timeseries(self):
         """push/pull pandas.TimeSeries"""
         import pandas
         
-        ts = pandas.TimeSeries(range(10))
+        ts = pandas.TimeSeries(list(range(10)))
         
         view = self.client[-1]
         
         view.push(dict(ts=ts), block=True)
         rts = view['ts']
         
         self.assertEqual(type(rts), type(ts))
         self.assertTrue((ts == rts).all())
     
     def test_map(self):
         view = self.client[:]
         def f(x):
             return x**2
-        data = range(16)
+        data = list(range(16))
         r = view.map_sync(f, data)
-        self.assertEqual(r, map(f, data))
+        self.assertEqual(r, list(map(f, data)))
     
     def test_map_iterable(self):
         """test map on iterables (direct)"""
         view = self.client[:]
         # 101 is prime, so it won't be evenly distributed
         arr = range(101)
         # ensure it will be an iterator, even in Python 3
         it = iter(arr)
         r = view.map_sync(lambda x: x, it)
         self.assertEqual(r, list(arr))
 
     @skip_without('numpy')
     def test_map_numpy(self):
         """test map on numpy arrays (direct)"""
         import numpy
         from numpy.testing.utils import assert_array_equal
 
         view = self.client[:]
         # 101 is prime, so it won't be evenly distributed
         arr = numpy.arange(101)
         r = view.map_sync(lambda x: x, arr)
         assert_array_equal(r, arr)
     
     def test_scatter_gather_nonblocking(self):
-        data = range(16)
+        data = list(range(16))
         view = self.client[:]
         view.scatter('a', data, block=False)
         ar = view.gather('a', block=False)
         self.assertEqual(ar.get(), data)
     
     @skip_without('numpy')
     def test_scatter_gather_numpy_nonblocking(self):
         import numpy
         from numpy.testing.utils import assert_array_equal
         a = numpy.arange(64)
         view = self.client[:]
         ar = view.scatter('a', a, block=False)
         self.assertTrue(isinstance(ar, AsyncResult))
         amr = view.gather('a', block=False)
         self.assertTrue(isinstance(amr, AsyncMapResult))
         assert_array_equal(amr.get(), a)
 
     def test_execute(self):
         view = self.client[:]
         # self.client.debug=True
         execute = view.execute
         ar = execute('c=30', block=False)
         self.assertTrue(isinstance(ar, AsyncResult))
         ar = execute('d=[0,1,2]', block=False)
         self.client.wait(ar, 1)
         self.assertEqual(len(ar.get()), len(self.client))
         for c in view['c']:
             self.assertEqual(c, 30)
     
     def test_abort(self):
         view = self.client[-1]
         ar = view.execute('import time; time.sleep(1)', block=False)
         ar2 = view.apply_async(lambda : 2)
         ar3 = view.apply_async(lambda : 3)
         view.abort(ar2)
         view.abort(ar3.msg_ids)
         self.assertRaises(error.TaskAborted, ar2.get)
         self.assertRaises(error.TaskAborted, ar3.get)
     
     def test_abort_all(self):
         """view.abort() aborts all outstanding tasks"""
         view = self.client[-1]
         ars = [ view.apply_async(time.sleep, 0.25) for i in range(10) ]
         view.abort()
         view.wait(timeout=5)
         for ar in ars[5:]:
             self.assertRaises(error.TaskAborted, ar.get)
     
     def test_temp_flags(self):
         view = self.client[-1]
         view.block=True
         with view.temp_flags(block=False):
             self.assertFalse(view.block)
         self.assertTrue(view.block)
     
     @dec.known_failure_py3
     def test_importer(self):
         view = self.client[-1]
         view.clear(block=True)
         with view.importer:
             import re
         
         @interactive
         def findall(pat, s):
             # this globals() step isn't necessary in real code
             # only to prevent a closure in the test
             re = globals()['re']
             return re.findall(pat, s)
         
         self.assertEqual(view.apply_sync(findall, '\w+', 'hello world'), 'hello world'.split())
     
     def test_unicode_execute(self):
         """test executing unicode strings"""
         v = self.client[-1]
         v.block=True
         if sys.version_info[0] >= 3:
             code="a='é'"
         else:
             code=u"a=u'é'"
         v.execute(code)
         self.assertEqual(v['a'], u'é')
         
     def test_unicode_apply_result(self):
         """test unicode apply results"""
         v = self.client[-1]
         r = v.apply_sync(lambda : u'é')
         self.assertEqual(r, u'é')
     
     def test_unicode_apply_arg(self):
         """test passing unicode arguments to apply"""
         v = self.client[-1]
         
         @interactive
         def check_unicode(a, check):
-            assert isinstance(a, unicode_type), "%r is not unicode"%a
+            assert not isinstance(a, bytes), "%r is bytes, not unicode"%a
             assert isinstance(check, bytes), "%r is not bytes"%check
             assert a.encode('utf8') == check, "%s != %s"%(a,check)
         
         for s in [ u'é', u'ßø®∫',u'asdf' ]:
             try:
                 v.apply_sync(check_unicode, s, s.encode('utf8'))
             except error.RemoteError as e:
                 if e.ename == 'AssertionError':
                     self.fail(e.evalue)
                 else:
                     raise e
     
     def test_map_reference(self):
         """view.map(<Reference>, *seqs) should work"""
         v = self.client[:]
         v.scatter('n', self.client.ids, flatten=True)
         v.execute("f = lambda x,y: x*y")
         rf = pmod.Reference('f')
         nlist = list(range(10))
         mlist = nlist[::-1]
         expected = [ m*n for m,n in zip(mlist, nlist) ]
         result = v.map_sync(rf, mlist, nlist)
         self.assertEqual(result, expected)
 
     def test_apply_reference(self):
         """view.apply(<Reference>, *args) should work"""
         v = self.client[:]
         v.scatter('n', self.client.ids, flatten=True)
         v.execute("f = lambda x: n*x")
         rf = pmod.Reference('f')
         result = v.apply_sync(rf, 5)
         expected = [ 5*id for id in self.client.ids ]
         self.assertEqual(result, expected)
     
     def test_eval_reference(self):
         v = self.client[self.client.ids[0]]
-        v['g'] = range(5)
+        v['g'] = list(range(5))
         rg = pmod.Reference('g[0]')
         echo = lambda x:x
         self.assertEqual(v.apply_sync(echo, rg), 0)
     
     def test_reference_nameerror(self):
         v = self.client[self.client.ids[0]]
         r = pmod.Reference('elvis_has_left')
         echo = lambda x:x
         self.assertRaisesRemote(NameError, v.apply_sync, echo, r)
 
     def test_single_engine_map(self):
         e0 = self.client[self.client.ids[0]]
-        r = range(5)
+        r = list(range(5))
         check = [ -1*i for i in r ]
         result = e0.map_sync(lambda x: -1*x, r)
         self.assertEqual(result, check)
     
     def test_len(self):
         """len(view) makes sense"""
         e0 = self.client[self.client.ids[0]]
         self.assertEqual(len(e0), 1)
         v = self.client[:]
         self.assertEqual(len(v), len(self.client.ids))
         v = self.client.direct_view('all')
         self.assertEqual(len(v), len(self.client.ids))
         v = self.client[:2]
         self.assertEqual(len(v), 2)
         v = self.client[:1]
         self.assertEqual(len(v), 1)
         v = self.client.load_balanced_view()
         self.assertEqual(len(v), len(self.client.ids))
 
     
     # begin execute tests
     
     def test_execute_reply(self):
         e0 = self.client[self.client.ids[0]]
         e0.block = True
         ar = e0.execute("5", silent=False)
         er = ar.get()
         self.assertEqual(str(er), "<ExecuteReply[%i]: 5>" % er.execution_count)
         self.assertEqual(er.pyout['data']['text/plain'], '5')
 
     def test_execute_reply_rich(self):
         e0 = self.client[self.client.ids[0]]
         e0.block = True
         e0.execute("from IPython.display import Image, HTML")
         ar = e0.execute("Image(data=b'garbage', format='png', width=10)", silent=False)
         er = ar.get()
         b64data = base64.encodestring(b'garbage').decode('ascii')
         self.assertEqual(er._repr_png_(), (b64data, dict(width=10)))
         ar = e0.execute("HTML('<b>bold</b>')", silent=False)
         er = ar.get()
         self.assertEqual(er._repr_html_(), "<b>bold</b>")
 
     def test_execute_reply_stdout(self):
         e0 = self.client[self.client.ids[0]]
         e0.block = True
         ar = e0.execute("print (5)", silent=False)
         er = ar.get()
         self.assertEqual(er.stdout.strip(), '5')
         
     def test_execute_pyout(self):
         """execute triggers pyout with silent=False"""
         view = self.client[:]
         ar = view.execute("5", silent=False, block=True)
         
         expected = [{'text/plain' : '5'}] * len(view)
         mimes = [ out['data'] for out in ar.pyout ]
         self.assertEqual(mimes, expected)
     
     def test_execute_silent(self):
         """execute does not trigger pyout with silent=True"""
         view = self.client[:]
         ar = view.execute("5", block=True)
         expected = [None] * len(view)
         self.assertEqual(ar.pyout, expected)
     
     def test_execute_magic(self):
         """execute accepts IPython commands"""
         view = self.client[:]
         view.execute("a = 5")
         ar = view.execute("%whos", block=True)
         # this will raise, if that failed
         ar.get(5)
         for stdout in ar.stdout:
             lines = stdout.splitlines()
             self.assertEqual(lines[0].split(), ['Variable', 'Type', 'Data/Info'])
             found = False
             for line in lines[2:]:
                 split = line.split()
                 if split == ['a', 'int', '5']:
                     found = True
                     break
             self.assertTrue(found, "whos output wrong: %s" % stdout)
     
     def test_execute_displaypub(self):
         """execute tracks display_pub output"""
         view = self.client[:]
         view.execute("from IPython.core.display import *")
         ar = view.execute("[ display(i) for i in range(5) ]", block=True)
         
         expected = [ {u'text/plain' : unicode_type(j)} for j in range(5) ]
         for outputs in ar.outputs:
             mimes = [ out['data'] for out in outputs ]
             self.assertEqual(mimes, expected)
     
     def test_apply_displaypub(self):
         """apply tracks display_pub output"""
         view = self.client[:]
         view.execute("from IPython.core.display import *")
         
         @interactive
         def publish():
             [ display(i) for i in range(5) ]
         
         ar = view.apply_async(publish)
         ar.get(5)
         expected = [ {u'text/plain' : unicode_type(j)} for j in range(5) ]
         for outputs in ar.outputs:
             mimes = [ out['data'] for out in outputs ]
             self.assertEqual(mimes, expected)
     
     def test_execute_raises(self):
         """exceptions in execute requests raise appropriately"""
         view = self.client[-1]
         ar = view.execute("1/0")
         self.assertRaisesRemote(ZeroDivisionError, ar.get, 2)
     
     def test_remoteerror_render_exception(self):
         """RemoteErrors get nice tracebacks"""
         view = self.client[-1]
         ar = view.execute("1/0")
         ip = get_ipython()
         ip.user_ns['ar'] = ar
         with capture_output() as io:
             ip.run_cell("ar.get(2)")
         
         self.assertTrue('ZeroDivisionError' in io.stdout, io.stdout)
     
     def test_compositeerror_render_exception(self):
         """CompositeErrors get nice tracebacks"""
         view = self.client[:]
         ar = view.execute("1/0")
         ip = get_ipython()
         ip.user_ns['ar'] = ar
 
         with capture_output() as io:
             ip.run_cell("ar.get(2)")
         
         count = min(error.CompositeError.tb_limit, len(view))
         
         self.assertEqual(io.stdout.count('ZeroDivisionError'), count * 2, io.stdout)
         self.assertEqual(io.stdout.count('by zero'), count, io.stdout)
         self.assertEqual(io.stdout.count(':execute'), count, io.stdout)
     
     def test_compositeerror_truncate(self):
         """Truncate CompositeErrors with many exceptions"""
         view = self.client[:]
         msg_ids = []
         for i in range(10):
             ar = view.execute("1/0")
             msg_ids.extend(ar.msg_ids)
         
         ar = self.client.get_result(msg_ids)
         try:
             ar.get()
         except error.CompositeError as _e:
             e = _e
         else:
             self.fail("Should have raised CompositeError")
         
         lines = e.render_traceback()
         with capture_output() as io:
             e.print_traceback()
         
         self.assertTrue("more exceptions" in lines[-1])
         count = e.tb_limit
         
         self.assertEqual(io.stdout.count('ZeroDivisionError'), 2 * count, io.stdout)
         self.assertEqual(io.stdout.count('by zero'), count, io.stdout)
         self.assertEqual(io.stdout.count(':execute'), count, io.stdout)
     
     @dec.skipif_not_matplotlib
     def test_magic_pylab(self):
         """%pylab works on engines"""
         view = self.client[-1]
         ar = view.execute("%pylab inline")
         # at least check if this raised:
         reply = ar.get(5)
         # include imports, in case user config
         ar = view.execute("plot(rand(100))", silent=False)
         reply = ar.get(5)
         self.assertEqual(len(reply.outputs), 1)
         output = reply.outputs[0]
         self.assertTrue("data" in output)
         data = output['data']
         self.assertTrue("image/png" in data)
     
     def test_func_default_func(self):
         """interactively defined function as apply func default"""
         def foo():
             return 'foo'
         
         def bar(f=foo):
             return f()
         
         view = self.client[-1]
         ar = view.apply_async(bar)
         r = ar.get(10)
         self.assertEqual(r, 'foo')
     def test_data_pub_single(self):
         view = self.client[-1]
         ar = view.execute('\n'.join([
             'from IPython.kernel.zmq.datapub import publish_data',
             'for i in range(5):',
             '  publish_data(dict(i=i))'
         ]), block=False)
         self.assertTrue(isinstance(ar.data, dict))
         ar.get(5)
         self.assertEqual(ar.data, dict(i=4))
 
     def test_data_pub(self):
         view = self.client[:]
         ar = view.execute('\n'.join([
             'from IPython.kernel.zmq.datapub import publish_data',
             'for i in range(5):',
             '  publish_data(dict(i=i))'
         ]), block=False)
         self.assertTrue(all(isinstance(d, dict) for d in ar.data))
         ar.get(5)
         self.assertEqual(ar.data, [dict(i=4)] * len(ar))
     
     def test_can_list_arg(self):
         """args in lists are canned"""
         view = self.client[-1]
         view['a'] = 128
         rA = pmod.Reference('a')
         ar = view.apply_async(lambda x: x, [rA])
         r = ar.get(5)
         self.assertEqual(r, [128])
 
     def test_can_dict_arg(self):
         """args in dicts are canned"""
         view = self.client[-1]
         view['a'] = 128
         rA = pmod.Reference('a')
         ar = view.apply_async(lambda x: x, dict(foo=rA))
         r = ar.get(5)
         self.assertEqual(r, dict(foo=128))
 
     def test_can_list_kwarg(self):
         """kwargs in lists are canned"""
         view = self.client[-1]
         view['a'] = 128
         rA = pmod.Reference('a')
         ar = view.apply_async(lambda x=5: x, x=[rA])
         r = ar.get(5)
         self.assertEqual(r, [128])
 
     def test_can_dict_kwarg(self):
         """kwargs in dicts are canned"""
         view = self.client[-1]
         view['a'] = 128
         rA = pmod.Reference('a')
         ar = view.apply_async(lambda x=5: x, dict(foo=rA))
         r = ar.get(5)
         self.assertEqual(r, dict(foo=128))
     
     def test_map_ref(self):
         """view.map works with references"""
         view = self.client[:]
         ranks = sorted(self.client.ids)
         view.scatter('rank', ranks, flatten=True)
         rrank = pmod.Reference('rank')
         
         amr = view.map_async(lambda x: x*2, [rrank] * len(view))
         drank = amr.get(5)
         self.assertEqual(drank, [ r*2 for r in ranks ])
         
     def test_nested_getitem_setitem(self):
         """get and set with view['a.b']"""
         view = self.client[-1]
         view.execute('\n'.join([
             'class A(object): pass',
             'a = A()',
             'a.b = 128',
             ]), block=True)
         ra = pmod.Reference('a')
 
         r = view.apply_sync(lambda x: x.b, ra)
         self.assertEqual(r, 128)
         self.assertEqual(view['a.b'], 128)
 
         view['a.b'] = 0
 
         r = view.apply_sync(lambda x: x.b, ra)
         self.assertEqual(r, 0)
         self.assertEqual(view['a.b'], 0)
     
     def test_return_namedtuple(self):
         def namedtuplify(x, y):
             from IPython.parallel.tests.test_view import point
             return point(x, y)
         
         view = self.client[-1]
         p = view.apply_sync(namedtuplify, 1, 2)
         self.assertEqual(p.x, 1)
         self.assertEqual(p.y, 2)
 
     def test_apply_namedtuple(self):
         def echoxy(p):
             return p.y, p.x
         
         view = self.client[-1]
         tup = view.apply_sync(echoxy, point(1, 2))
         self.assertEqual(tup, (2,1))
     
     def test_sync_imports(self):
         view = self.client[-1]
         with capture_output() as io:
             with view.sync_imports():
                 import IPython
         self.assertIn("IPython", io.stdout)
         
         @interactive
         def find_ipython():
             return 'IPython' in globals()
         
         assert view.apply_sync(find_ipython)
 
     def test_sync_imports_quiet(self):
         view = self.client[-1]
         with capture_output() as io:
             with view.sync_imports(quiet=True):
                 import IPython
         self.assertEqual(io.stdout, '')
         
         @interactive
         def find_ipython():
             return 'IPython' in globals()
         
         assert view.apply_sync(find_ipython)
 

IPython/parallel/util.py

@@ -1,369 +1,369 b''
 """some generic utilities for dealing with classes, urls, and serialization
 
 Authors:
 
 * Min RK
 """
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2010-2011  The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 # Standard library imports.
 import logging
 import os
 import re
 import stat
 import socket
 import sys
 from signal import signal, SIGINT, SIGABRT, SIGTERM
 try:
     from signal import SIGKILL
 except ImportError:
     SIGKILL=None
 
 try:
     import cPickle
     pickle = cPickle
 except:
     cPickle = None
     import pickle
 
 # System library imports
 import zmq
 from zmq.log import handlers
 
 from IPython.external.decorator import decorator
 
 # IPython imports
 from IPython.config.application import Application
 from IPython.utils.localinterfaces import localhost, is_public_ip, public_ips
 from IPython.utils.py3compat import string_types, iteritems, itervalues
 from IPython.kernel.zmq.log import EnginePUBHandler
 from IPython.kernel.zmq.serialize import (
     unserialize_object, serialize_object, pack_apply_message, unpack_apply_message
 )
 
 #-----------------------------------------------------------------------------
 # Classes
 #-----------------------------------------------------------------------------
 
 class Namespace(dict):
     """Subclass of dict for attribute access to keys."""
     
     def __getattr__(self, key):
         """getattr aliased to getitem"""
         if key in self:
             return self[key]
         else:
             raise NameError(key)
 
     def __setattr__(self, key, value):
         """setattr aliased to setitem, with strict"""
         if hasattr(dict, key):
             raise KeyError("Cannot override dict keys %r"%key)
         self[key] = value
     
 
 class ReverseDict(dict):
     """simple double-keyed subset of dict methods."""
     
     def __init__(self, *args, **kwargs):
         dict.__init__(self, *args, **kwargs)
         self._reverse = dict()
         for key, value in iteritems(self):
             self._reverse[value] = key
     
     def __getitem__(self, key):
         try:
             return dict.__getitem__(self, key)
         except KeyError:
             return self._reverse[key]
     
     def __setitem__(self, key, value):
         if key in self._reverse:
             raise KeyError("Can't have key %r on both sides!"%key)
         dict.__setitem__(self, key, value)
         self._reverse[value] = key
     
     def pop(self, key):
         value = dict.pop(self, key)
         self._reverse.pop(value)
         return value
     
     def get(self, key, default=None):
         try:
             return self[key]
         except KeyError:
             return default
 
 #-----------------------------------------------------------------------------
 # Functions
 #-----------------------------------------------------------------------------
 
 @decorator
 def log_errors(f, self, *args, **kwargs):
     """decorator to log unhandled exceptions raised in a method.
     
     For use wrapping on_recv callbacks, so that exceptions
     do not cause the stream to be closed.
     """
     try:
         return f(self, *args, **kwargs)
     except Exception:
         self.log.error("Uncaught exception in %r" % f, exc_info=True)
     
 
 def is_url(url):
     """boolean check for whether a string is a zmq url"""
     if '://' not in url:
         return False
     proto, addr = url.split('://', 1)
     if proto.lower() not in ['tcp','pgm','epgm','ipc','inproc']:
         return False
     return True
 
 def validate_url(url):
     """validate a url for zeromq"""
     if not isinstance(url, string_types):
         raise TypeError("url must be a string, not %r"%type(url))
     url = url.lower()
     
     proto_addr = url.split('://')
     assert len(proto_addr) == 2, 'Invalid url: %r'%url
     proto, addr = proto_addr
     assert proto in ['tcp','pgm','epgm','ipc','inproc'], "Invalid protocol: %r"%proto
     
     # domain pattern adapted from http://www.regexlib.com/REDetails.aspx?regexp_id=391
     # author: Remi Sabourin
     pat = re.compile(r'^([\w\d]([\w\d\-]{0,61}[\w\d])?\.)*[\w\d]([\w\d\-]{0,61}[\w\d])?$')
     
     if proto == 'tcp':
         lis = addr.split(':')
         assert len(lis) == 2, 'Invalid url: %r'%url
         addr,s_port = lis
         try:
             port = int(s_port)
         except ValueError:
             raise AssertionError("Invalid port %r in url: %r"%(port, url))
         
         assert addr == '*' or pat.match(addr) is not None, 'Invalid url: %r'%url
         
     else:
         # only validate tcp urls currently
         pass
     
     return True
 
 
 def validate_url_container(container):
     """validate a potentially nested collection of urls."""
     if isinstance(container, string_types):
         url = container
         return validate_url(url)
     elif isinstance(container, dict):
         container = itervalues(container)
     
     for element in container:
         validate_url_container(element)
 
 
 def split_url(url):
     """split a zmq url (tcp://ip:port) into ('tcp','ip','port')."""
     proto_addr = url.split('://')
     assert len(proto_addr) == 2, 'Invalid url: %r'%url
     proto, addr = proto_addr
     lis = addr.split(':')
     assert len(lis) == 2, 'Invalid url: %r'%url
     addr,s_port = lis
     return proto,addr,s_port
     
 def disambiguate_ip_address(ip, location=None):
     """turn multi-ip interfaces '0.0.0.0' and '*' into connectable
     ones, based on the location (default interpretation of location is localhost)."""
     if ip in ('0.0.0.0', '*'):
         if location is None or is_public_ip(location) or not public_ips():
             # If location is unspecified or cannot be determined, assume local
             ip = localhost()
         elif location:
             return location
     return ip
 
 def disambiguate_url(url, location=None):
     """turn multi-ip interfaces '0.0.0.0' and '*' into connectable
     ones, based on the location (default interpretation is localhost).
     
     This is for zeromq urls, such as tcp://*:10101."""
     try:
         proto,ip,port = split_url(url)
     except AssertionError:
         # probably not tcp url; could be ipc, etc.
         return url
     
     ip = disambiguate_ip_address(ip,location)
     
     return "%s://%s:%s"%(proto,ip,port)
 
 
 #--------------------------------------------------------------------------
 # helpers for implementing old MEC API via view.apply
 #--------------------------------------------------------------------------
 
 def interactive(f):
     """decorator for making functions appear as interactively defined.
     This results in the function being linked to the user_ns as globals()
     instead of the module globals().
     """
     f.__module__ = '__main__'
     return f
 
 @interactive
 def _push(**ns):
     """helper method for implementing `client.push` via `client.apply`"""
     user_ns = globals()
     tmp = '_IP_PUSH_TMP_'
     while tmp in user_ns:
         tmp = tmp + '_'
     try:
-        for name, value in iteritems(ns):
+        for name, value in ns.items():
             user_ns[tmp] = value
             exec("%s = %s" % (name, tmp), user_ns)
     finally:
         user_ns.pop(tmp, None)
 
 @interactive
 def _pull(keys):
     """helper method for implementing `client.pull` via `client.apply`"""
     if isinstance(keys, (list,tuple, set)):
-        return map(lambda key: eval(key, globals()), keys)
+        return [eval(key, globals()) for key in keys]
     else:
         return eval(keys, globals())
 
 @interactive
 def _execute(code):
     """helper method for implementing `client.execute` via `client.apply`"""
     exec(code, globals())
 
 #--------------------------------------------------------------------------
 # extra process management utilities
 #--------------------------------------------------------------------------
 
 _random_ports = set()
 
 def select_random_ports(n):
     """Selects and return n random ports that are available."""
     ports = []
     for i in range(n):
         sock = socket.socket()
         sock.bind(('', 0))
         while sock.getsockname()[1] in _random_ports:
             sock.close()
             sock = socket.socket()
             sock.bind(('', 0))
         ports.append(sock)
     for i, sock in enumerate(ports):
         port = sock.getsockname()[1]
         sock.close()
         ports[i] = port
         _random_ports.add(port)
     return ports
 
 def signal_children(children):
     """Relay interupt/term signals to children, for more solid process cleanup."""
     def terminate_children(sig, frame):
         log = Application.instance().log
         log.critical("Got signal %i, terminating children..."%sig)
         for child in children:
             child.terminate()
         
         sys.exit(sig != SIGINT)
         # sys.exit(sig)
     for sig in (SIGINT, SIGABRT, SIGTERM):
         signal(sig, terminate_children)
 
 def generate_exec_key(keyfile):
     import uuid
     newkey = str(uuid.uuid4())
     with open(keyfile, 'w') as f:
         # f.write('ipython-key ')
         f.write(newkey+'\n')
     # set user-only RW permissions (0600)
     # this will have no effect on Windows
     os.chmod(keyfile, stat.S_IRUSR|stat.S_IWUSR)
 
 
 def integer_loglevel(loglevel):
     try:
         loglevel = int(loglevel)
     except ValueError:
         if isinstance(loglevel, str):
             loglevel = getattr(logging, loglevel)
     return loglevel
 
 def connect_logger(logname, context, iface, root="ip", loglevel=logging.DEBUG):
     logger = logging.getLogger(logname)
     if any([isinstance(h, handlers.PUBHandler) for h in logger.handlers]):
         # don't add a second PUBHandler
         return
     loglevel = integer_loglevel(loglevel)
     lsock = context.socket(zmq.PUB)
     lsock.connect(iface)
     handler = handlers.PUBHandler(lsock)
     handler.setLevel(loglevel)
     handler.root_topic = root
     logger.addHandler(handler)
     logger.setLevel(loglevel)
 
 def connect_engine_logger(context, iface, engine, loglevel=logging.DEBUG):
     logger = logging.getLogger()
     if any([isinstance(h, handlers.PUBHandler) for h in logger.handlers]):
         # don't add a second PUBHandler
         return
     loglevel = integer_loglevel(loglevel)
     lsock = context.socket(zmq.PUB)
     lsock.connect(iface)
     handler = EnginePUBHandler(engine, lsock)
     handler.setLevel(loglevel)
     logger.addHandler(handler)
     logger.setLevel(loglevel)
     return logger
 
 def local_logger(logname, loglevel=logging.DEBUG):
     loglevel = integer_loglevel(loglevel)
     logger = logging.getLogger(logname)
     if any([isinstance(h, logging.StreamHandler) for h in logger.handlers]):
         # don't add a second StreamHandler
         return
     handler = logging.StreamHandler()
     handler.setLevel(loglevel)
     formatter = logging.Formatter("%(asctime)s.%(msecs).03d [%(name)s] %(message)s",
                 datefmt="%Y-%m-%d %H:%M:%S")
     handler.setFormatter(formatter)
 
     logger.addHandler(handler)
     logger.setLevel(loglevel)
     return logger
 
 def set_hwm(sock, hwm=0):
     """set zmq High Water Mark on a socket
     
     in a way that always works for various pyzmq / libzmq versions.
     """
     import zmq
     
     for key in ('HWM', 'SNDHWM', 'RCVHWM'):
         opt = getattr(zmq, key, None)
         if opt is None:
             continue
         try:
             sock.setsockopt(opt, hwm)
         except zmq.ZMQError:
             pass
 
         
  No newline at end of file