Update changelog for v0.6.1

xld - use native plistlib instead of biplist

rutracker - use new Result object

seacher - small tweaks

transmission - str no longer required

CHANGELOG.md

@@ -1,5 +1,21 @@
 # Changelog
 
+## v0.6.1
+Released 26 November 2023
+
+Highlights:
+* Dependency updates to work with > Python 3.11
+
+The full list of commits can be found [here](https://github.com/rembo10/headphones/compare/v0.6.0...v0.6.1).
+
+## v0.6.0
+Released 13 November 2022
+
+Highlights:
+* Updated to python 3
+
+The full list of commits can be found [here](https://github.com/rembo10/headphones/compare/v0.5.20...v0.6.0).
+
 ## v0.5.20
 Released 15 October 2021
 

data/interfaces/default/config.html

@@ -1654,6 +1654,16 @@
                                 </div>
                             </fieldset>
 
+                            <fieldset>
+                                <legend>Last.fm</legend>
+                                <div id="lastfmoptions">
+                                    <div class="row">
+                                        <label>API Key</label>
+                                        <input type="text" name="lastfm_apikey" value="${config['lastfm_apikey']}" size="40" />
+                                    </div>
+                                </div>
+                            </fieldset>
+
                             <fieldset>
                                 <legend>Songkick</legend>
                                 <div class="row checkbox">

headphones/config.py

@@ -155,6 +155,7 @@ _CONFIG_DEFINITIONS = {
     'KEEP_TORRENT_FILES': (int, 'General', 0),
     'KEEP_TORRENT_FILES_DIR': (path, 'General', ''),
     'LASTFM_USERNAME': (str, 'General', ''),
+    'LASTFM_APIKEY': (str, 'General', ''),
     'LAUNCH_BROWSER': (int, 'General', 1),
     'LIBRARYSCAN': (int, 'General', 1),
     'LIBRARYSCAN_INTERVAL': (int, 'General', 24),

headphones/getXldProfile.py

@@ -1,6 +1,6 @@
 import os.path
 
-import biplist
+import plistlib
 from headphones import logger
 
 
@@ -14,8 +14,9 @@ def getXldProfile(xldProfile):
 
     # Get xld preferences plist
     try:
-        preferences = biplist.readPlist(expanded)
-    except (biplist.InvalidPlistException, biplist.NotBinaryPlistException) as e:
+        with open(expanded, 'rb') as _f:
+            preferences = plistlib.load(_f)
+    except Exception as e:
         logger.error("Error reading xld preferences plist: %s", e)
         return (xldProfileNotFound, None, None)
 

headphones/importer.py

@@ -102,13 +102,7 @@ def artistlist_to_mbids(artistlist, forced=False):
             myDB.action('DELETE from newartists WHERE ArtistName=?', [artist])
 
     # Update the similar artist tag cloud:
-    # TODO: Fix last.fm api
-    # logger.info('Updating artist information from Last.fm')
-
-    # try:
-    #     lastfm.getSimilar()
-    # except Exception as e:
-    #     logger.warn('Failed to update artist information from Last.fm: %s' % e)
+    lastfm.getSimilar()
 
 
 def addArtistIDListToDB(artistidlist):

headphones/lastfm.py

@@ -23,7 +23,7 @@ from headphones import db, logger, request
 TIMEOUT = 60.0  # seconds
 REQUEST_LIMIT = 1.0 / 5  # seconds
 ENTRY_POINT = "https://ws.audioscrobbler.com/2.0/"
-API_KEY = "395e6ec6bb557382fc41fde867bce66f"
+APP_API_KEY = "395e6ec6bb557382fc41fde867bce66f"
 
 # Required for API request limit
 lastfm_lock = headphones.lock.TimedLock(REQUEST_LIMIT)
@@ -31,7 +31,7 @@ lastfm_lock = headphones.lock.TimedLock(REQUEST_LIMIT)
 
 def request_lastfm(method, **kwargs):
     """
-    Call a Last.FM API method. Automatically sets the method and API key. Method
+    Call a Last.fm API method. Automatically sets the method and API key. Method
     will return the result if no error occured.
 
     By default, this method will request the JSON format, since it is more
@@ -40,32 +40,39 @@ def request_lastfm(method, **kwargs):
 
     # Prepare request
     kwargs["method"] = method
-    kwargs.setdefault("api_key", API_KEY)
+    kwargs.setdefault("api_key", headphones.CONFIG.LASTFM_APIKEY or APP_API_KEY)
     kwargs.setdefault("format", "json")
 
     # Send request
-    logger.debug("Calling Last.FM method: %s", method)
-    logger.debug("Last.FM call parameters: %s", kwargs)
+    logger.debug("Calling Last.fm method: %s", method)
+    logger.debug("Last.fm call parameters: %s", kwargs)
 
     data = request.request_json(ENTRY_POINT, timeout=TIMEOUT, params=kwargs, lock=lastfm_lock)
 
     # Parse response and check for errors.
     if not data:
-        logger.error("Error calling Last.FM method: %s", method)
+        logger.error("Error calling Last.fm method: %s", method)
         return
 
     if "error" in data:
-        logger.debug("Last.FM returned an error: %s", data["message"])
+        logger.debug("Last.fm returned an error: %s", data["message"])
         return
 
     return data
 
 
 def getSimilar():
+    if not headphones.CONFIG.LASTFM_APIKEY:
+        logger.info(
+            'To update the Similar Artists cloud tag, create a Last.fm application api key '
+            'and add it under the Advanced config tab'
+        )
+        return
+
     myDB = db.DBConnection()
     results = myDB.select("SELECT ArtistID from artists ORDER BY HaveTracks DESC LIMIT 10")
 
-    logger.info("Fetching similar artists from Last.FM for tag cloud")
+    logger.info("Fetching similar artists from Last.fm for tag cloud")
     artistlist = []
 
     for result in results:
@@ -85,7 +92,7 @@ def getSimilar():
                     artistlist.append((artist_name, artist_mbid))
 
     # Add new artists to tag cloud
-    logger.debug("Fetched %d artists from Last.FM", len(artistlist))
+    logger.debug("Fetched %d artists from Last.fm", len(artistlist))
     count = defaultdict(int)
 
     for artist, mbid in artistlist:
@@ -103,7 +110,7 @@ def getSimilar():
 
         myDB.action("INSERT INTO lastfmcloud VALUES( ?, ?, ?)", [artist_name, artist_mbid, count])
 
-    logger.debug("Inserted %d artists into Last.FM tag cloud", len(top_list))
+    logger.debug("Inserted %d artists into Last.fm tag cloud", len(top_list))
 
 
 def getArtists():
@@ -111,16 +118,16 @@ def getArtists():
     results = myDB.select("SELECT ArtistID from artists")
 
     if not headphones.CONFIG.LASTFM_USERNAME:
-        logger.warn("Last.FM username not set, not importing artists.")
+        logger.warn("Last.fm username not set, not importing artists.")
         return
 
-    logger.info("Fetching artists from Last.FM for username: %s", headphones.CONFIG.LASTFM_USERNAME)
+    logger.info("Fetching artists from Last.fm for username: %s", headphones.CONFIG.LASTFM_USERNAME)
     data = request_lastfm("library.getartists", limit=1000, user=headphones.CONFIG.LASTFM_USERNAME)
 
     if data and "artists" in data:
         artistlist = []
         artists = data["artists"]["artist"]
-        logger.debug("Fetched %d artists from Last.FM", len(artists))
+        logger.debug("Fetched %d artists from Last.fm", len(artists))
 
         for artist in artists:
             artist_mbid = artist["mbid"]
@@ -133,20 +140,20 @@ def getArtists():
         for artistid in artistlist:
             importer.addArtisttoDB(artistid)
 
-        logger.info("Imported %d new artists from Last.FM", len(artistlist))
+        logger.info("Imported %d new artists from Last.fm", len(artistlist))
 
 
 def getTagTopArtists(tag, limit=50):
     myDB = db.DBConnection()
     results = myDB.select("SELECT ArtistID from artists")
 
-    logger.info("Fetching top artists from Last.FM for tag: %s", tag)
+    logger.info("Fetching top artists from Last.fm for tag: %s", tag)
     data = request_lastfm("tag.gettopartists", limit=limit, tag=tag)
 
     if data and "topartists" in data:
         artistlist = []
         artists = data["topartists"]["artist"]
-        logger.debug("Fetched %d artists from Last.FM", len(artists))
+        logger.debug("Fetched %d artists from Last.fm", len(artists))
 
         for artist in artists:
             try:
@@ -162,4 +169,4 @@ def getTagTopArtists(tag, limit=50):
         for artistid in artistlist:
             importer.addArtisttoDB(artistid)
 
-        logger.debug("Added %d new artists from Last.FM", len(artistlist))
+        logger.debug("Added %d new artists from Last.fm", len(artistlist))

headphones/librarysync.py

@@ -442,9 +442,8 @@ def libraryScan(dir=None, append=False, ArtistID=None, ArtistName=None,
     # if not append:
     #     update_album_status()
 
-    # TODO: Fix last.fm api calls
-    #if not append and not artistScan:
-        #lastfm.getSimilar()
+    if not append and not artistScan:
+        lastfm.getSimilar()
 
     if ArtistName:
         logger.info('Scanning complete for artist: %s', ArtistName)

headphones/rutracker.py

@@ -11,6 +11,7 @@ from bs4 import BeautifulSoup
 
 import headphones
 from headphones import logger
+from headphones.types import Result
 
 
 class Rutracker(object):
@@ -160,7 +161,7 @@ class Rutracker(object):
                     torrent_id = dict([part.split('=') for part in urlparse(url)[4].split('&')])[
                         't']
                     topicurl = 'https://rutracker.org/forum/viewtopic.php?t=' + torrent_id
-                    rulist.append((title, size, topicurl, 'rutracker.org', 'torrent', True))
+                    rulist.append(Result(title, size, url, 'rutracker.org', 'torrent', True))
                 else:
                     logger.info("%s is larger than the maxsize or has too little seeders for this category, "
                                 "skipping. (Size: %i bytes, Seeders: %i)" % (title, size, int(seeds)))

headphones/searcher.py

@@ -1055,13 +1055,29 @@ def send_to_downloader(data, result, album):
             result.title,
             result.size,
             result.url,
-            "Seed_Snatched" if seed_ratio and torrentid else "Snatched",
+            "Snatched",
             folder_name,
             kind,
             torrentid
         ]
     )
 
+    # Additional record for post processing or scheduled job to remove the torrent when finished seeding
+    if seed_ratio is not None and seed_ratio != 0 and torrentid:
+        myDB.action(
+            "INSERT INTO snatched VALUES (?, ?, ?, ?, DATETIME('NOW', 'localtime'), "
+            "?, ?, ?, ?)", [
+                album['AlbumID'],
+                result.title,
+                result.size,
+                result.url,
+                "Seed_Snatched",
+                folder_name,
+                kind,
+                torrentid
+            ]
+        )
+
     # notify
     artist = album[1]
     albumname = album[2]
@@ -1902,14 +1918,6 @@ def preprocess(resultlist):
             headers = {'User-Agent': USER_AGENT}
 
         if result.kind == 'torrent':
-            # Get out of here if we're using Transmission or Deluge
-            # if not a magnet link still need the .torrent to generate hash... uTorrent support labeling
-            if headphones.CONFIG.TORRENT_DOWNLOADER in [1, 3]:
-                return True, result
-
-            # Get out of here if it's a magnet link
-            if result.url.lower().startswith("magnet:"):
-                return True, result
 
             # rutracker always needs the torrent data
             if result.provider == 'rutracker.org':
@@ -1944,6 +1952,14 @@ def preprocess(resultlist):
                     else:
                         return r.content, result
 
+            # Get out of here if we're using Transmission or Deluge
+            # if not a magnet link still need the .torrent to generate hash... uTorrent support labeling
+            if headphones.CONFIG.TORRENT_DOWNLOADER in [1, 3]:
+                return True, result
+
+            # Get out of here if it's a magnet link
+            if result.url.lower().startswith("magnet:"):
+                return True, result
 
             # Download the torrent file
             return request.request_content(url=result.url, headers=headers), result

headphones/transmission.py

@@ -15,7 +15,7 @@
 
 import time
 import json
-import base64
+from base64 import b64encode
 import urllib.parse
 import os
 
@@ -36,10 +36,10 @@ def addTorrent(link, data=None):
 
     if link.endswith('.torrent') and not link.startswith(('http', 'magnet')) or data:
         if data:
-            metainfo = str(base64.b64encode(data))
+            metainfo = b64encode(data).decode("utf-8")
         else:
             with open(link, 'rb') as f:
-                metainfo = str(base64.b64encode(f.read()))
+                metainfo = b64encode(f.read()).decode("utf-8")
         arguments = {'metainfo': metainfo, 'download-dir': headphones.CONFIG.DOWNLOAD_TORRENT_DIR}
     else:
         arguments = {'filename': link, 'download-dir': headphones.CONFIG.DOWNLOAD_TORRENT_DIR}

headphones/webserve.py

@@ -1385,6 +1385,7 @@ class WebInterface(object):
             "custompass": headphones.CONFIG.CUSTOMPASS,
             "hpuser": headphones.CONFIG.HPUSER,
             "hppass": headphones.CONFIG.HPPASS,
+            "lastfm_apikey": headphones.CONFIG.LASTFM_APIKEY,
             "songkick_enabled": checked(headphones.CONFIG.SONGKICK_ENABLED),
             "songkick_apikey": headphones.CONFIG.SONGKICK_APIKEY,
             "songkick_location": headphones.CONFIG.SONGKICK_LOCATION,

lib/apscheduler/__init__.py

@@ -1,5 +1,10 @@
-version_info = (3, 0, 1)
-version = '3.0.1'
-release = '3.0.1'
+from pkg_resources import get_distribution, DistributionNotFound
 
-__version__ = release  # PEP 396
+try:
+    release = get_distribution('APScheduler').version.split('-')[0]
+except DistributionNotFound:
+    release = '3.5.0'
+
+version_info = tuple(int(x) if x.isdigit() else x for x in release.split('.'))
+version = __version__ = '.'.join(str(x) for x in version_info[:3])
+del get_distribution, DistributionNotFound

lib/apscheduler/events.py

@@ -1,25 +1,33 @@
-__all__ = ('EVENT_SCHEDULER_START', 'EVENT_SCHEDULER_SHUTDOWN', 'EVENT_EXECUTOR_ADDED', 'EVENT_EXECUTOR_REMOVED',
-           'EVENT_JOBSTORE_ADDED', 'EVENT_JOBSTORE_REMOVED', 'EVENT_ALL_JOBS_REMOVED', 'EVENT_JOB_ADDED',
-           'EVENT_JOB_REMOVED', 'EVENT_JOB_MODIFIED', 'EVENT_JOB_EXECUTED', 'EVENT_JOB_ERROR', 'EVENT_JOB_MISSED',
-           'SchedulerEvent', 'JobEvent', 'JobExecutionEvent')
+__all__ = ('EVENT_SCHEDULER_STARTED', 'EVENT_SCHEDULER_SHUTDOWN', 'EVENT_SCHEDULER_PAUSED',
+           'EVENT_SCHEDULER_RESUMED', 'EVENT_EXECUTOR_ADDED', 'EVENT_EXECUTOR_REMOVED',
+           'EVENT_JOBSTORE_ADDED', 'EVENT_JOBSTORE_REMOVED', 'EVENT_ALL_JOBS_REMOVED',
+           'EVENT_JOB_ADDED', 'EVENT_JOB_REMOVED', 'EVENT_JOB_MODIFIED', 'EVENT_JOB_EXECUTED',
+           'EVENT_JOB_ERROR', 'EVENT_JOB_MISSED', 'EVENT_JOB_SUBMITTED', 'EVENT_JOB_MAX_INSTANCES',
+           'SchedulerEvent', 'JobEvent', 'JobExecutionEvent', 'JobSubmissionEvent')
 
 
-EVENT_SCHEDULER_START = 1
-EVENT_SCHEDULER_SHUTDOWN = 2
-EVENT_EXECUTOR_ADDED = 4
-EVENT_EXECUTOR_REMOVED = 8
-EVENT_JOBSTORE_ADDED = 16
-EVENT_JOBSTORE_REMOVED = 32
-EVENT_ALL_JOBS_REMOVED = 64
-EVENT_JOB_ADDED = 128
-EVENT_JOB_REMOVED = 256
-EVENT_JOB_MODIFIED = 512
-EVENT_JOB_EXECUTED = 1024
-EVENT_JOB_ERROR = 2048
-EVENT_JOB_MISSED = 4096
-EVENT_ALL = (EVENT_SCHEDULER_START | EVENT_SCHEDULER_SHUTDOWN | EVENT_JOBSTORE_ADDED | EVENT_JOBSTORE_REMOVED |
+EVENT_SCHEDULER_STARTED = EVENT_SCHEDULER_START = 2 ** 0
+EVENT_SCHEDULER_SHUTDOWN = 2 ** 1
+EVENT_SCHEDULER_PAUSED = 2 ** 2
+EVENT_SCHEDULER_RESUMED = 2 ** 3
+EVENT_EXECUTOR_ADDED = 2 ** 4
+EVENT_EXECUTOR_REMOVED = 2 ** 5
+EVENT_JOBSTORE_ADDED = 2 ** 6
+EVENT_JOBSTORE_REMOVED = 2 ** 7
+EVENT_ALL_JOBS_REMOVED = 2 ** 8
+EVENT_JOB_ADDED = 2 ** 9
+EVENT_JOB_REMOVED = 2 ** 10
+EVENT_JOB_MODIFIED = 2 ** 11
+EVENT_JOB_EXECUTED = 2 ** 12
+EVENT_JOB_ERROR = 2 ** 13
+EVENT_JOB_MISSED = 2 ** 14
+EVENT_JOB_SUBMITTED = 2 ** 15
+EVENT_JOB_MAX_INSTANCES = 2 ** 16
+EVENT_ALL = (EVENT_SCHEDULER_STARTED | EVENT_SCHEDULER_SHUTDOWN | EVENT_SCHEDULER_PAUSED |
+             EVENT_SCHEDULER_RESUMED | EVENT_EXECUTOR_ADDED | EVENT_EXECUTOR_REMOVED |
+             EVENT_JOBSTORE_ADDED | EVENT_JOBSTORE_REMOVED | EVENT_ALL_JOBS_REMOVED |
              EVENT_JOB_ADDED | EVENT_JOB_REMOVED | EVENT_JOB_MODIFIED | EVENT_JOB_EXECUTED |
-             EVENT_JOB_ERROR | EVENT_JOB_MISSED)
+             EVENT_JOB_ERROR | EVENT_JOB_MISSED | EVENT_JOB_SUBMITTED | EVENT_JOB_MAX_INSTANCES)
 
 
 class SchedulerEvent(object):
@@ -55,9 +63,21 @@ class JobEvent(SchedulerEvent):
         self.jobstore = jobstore
 
 
+class JobSubmissionEvent(JobEvent):
+    """
+    An event that concerns the submission of a job to its executor.
+
+    :ivar scheduled_run_times: a list of datetimes when the job was intended to run
+    """
+
+    def __init__(self, code, job_id, jobstore, scheduled_run_times):
+        super(JobSubmissionEvent, self).__init__(code, job_id, jobstore)
+        self.scheduled_run_times = scheduled_run_times
+
+
 class JobExecutionEvent(JobEvent):
     """
-    An event that concerns the execution of individual jobs.
+    An event that concerns the running of a job within its executor.
 
     :ivar scheduled_run_time: the time when the job was scheduled to be run
     :ivar retval: the return value of the successfully executed job
@@ -65,7 +85,8 @@ class JobExecutionEvent(JobEvent):
     :ivar traceback: a formatted traceback for the exception
     """
 
-    def __init__(self, code, job_id, jobstore, scheduled_run_time, retval=None, exception=None, traceback=None):
+    def __init__(self, code, job_id, jobstore, scheduled_run_time, retval=None, exception=None,
+                 traceback=None):
         super(JobExecutionEvent, self).__init__(code, job_id, jobstore)
         self.scheduled_run_time = scheduled_run_time
         self.retval = retval

lib/apscheduler/executors/asyncio.py

@@ -1,28 +1,52 @@
+from __future__ import absolute_import
 
 import sys
 
 from apscheduler.executors.base import BaseExecutor, run_job
+from apscheduler.executors.base_py3 import run_coroutine_job
+from apscheduler.util import iscoroutinefunction_partial
 
 
 class AsyncIOExecutor(BaseExecutor):
     """
     Runs jobs in the default executor of the event loop.
 
+    If the job function is a native coroutine function, it is scheduled to be run directly in the
+    event loop as soon as possible. All other functions are run in the event loop's default
+    executor which is usually a thread pool.
+
     Plugin alias: ``asyncio``
     """
 
     def start(self, scheduler, alias):
         super(AsyncIOExecutor, self).start(scheduler, alias)
         self._eventloop = scheduler._eventloop
+        self._pending_futures = set()
+
+    def shutdown(self, wait=True):
+        # There is no way to honor wait=True without converting this method into a coroutine method
+        for f in self._pending_futures:
+            if not f.done():
+                f.cancel()
+
+        self._pending_futures.clear()
 
     def _do_submit_job(self, job, run_times):
         def callback(f):
+            self._pending_futures.discard(f)
             try:
                 events = f.result()
-            except:
+            except BaseException:
                 self._run_job_error(job.id, *sys.exc_info()[1:])
             else:
                 self._run_job_success(job.id, events)
 
-        f = self._eventloop.run_in_executor(None, run_job, job, job._jobstore_alias, run_times, self._logger.name)
+        if iscoroutinefunction_partial(job.func):
+            coro = run_coroutine_job(job, job._jobstore_alias, run_times, self._logger.name)
+            f = self._eventloop.create_task(coro)
+        else:
+            f = self._eventloop.run_in_executor(None, run_job, job, job._jobstore_alias, run_times,
+                                                self._logger.name)
+
         f.add_done_callback(callback)
+        self._pending_futures.add(f)

lib/apscheduler/executors/base.py

@@ -8,13 +8,15 @@ import sys
 from pytz import utc
 import six
 
-from apscheduler.events import JobExecutionEvent, EVENT_JOB_MISSED, EVENT_JOB_ERROR, EVENT_JOB_EXECUTED
+from apscheduler.events import (
+    JobExecutionEvent, EVENT_JOB_MISSED, EVENT_JOB_ERROR, EVENT_JOB_EXECUTED)
 
 
 class MaxInstancesReachedError(Exception):
     def __init__(self, job):
         super(MaxInstancesReachedError, self).__init__(
-            'Job "%s" has already reached its maximum number of instances (%d)' % (job.id, job.max_instances))
+            'Job "%s" has already reached its maximum number of instances (%d)' %
+            (job.id, job.max_instances))
 
 
 class BaseExecutor(six.with_metaclass(ABCMeta, object)):
@@ -30,13 +32,14 @@ class BaseExecutor(six.with_metaclass(ABCMeta, object)):
 
     def start(self, scheduler, alias):
         """
-        Called by the scheduler when the scheduler is being started or when the executor is being added to an already
-        running scheduler.
+        Called by the scheduler when the scheduler is being started or when the executor is being
+        added to an already running scheduler.
 
-        :param apscheduler.schedulers.base.BaseScheduler scheduler: the scheduler that is starting this executor
+        :param apscheduler.schedulers.base.BaseScheduler scheduler: the scheduler that is starting
+            this executor
         :param str|unicode alias: alias of this executor as it was assigned to the scheduler
-        """
 
+        """
         self._scheduler = scheduler
         self._lock = scheduler._create_lock()
         self._logger = logging.getLogger('apscheduler.executors.%s' % alias)
@@ -45,7 +48,8 @@ class BaseExecutor(six.with_metaclass(ABCMeta, object)):
         """
         Shuts down this executor.
 
-        :param bool wait: ``True`` to wait until all submitted jobs have been executed
+        :param bool wait: ``True`` to wait until all submitted jobs
+            have been executed
         """
 
     def submit_job(self, job, run_times):
@@ -53,10 +57,12 @@ class BaseExecutor(six.with_metaclass(ABCMeta, object)):
         Submits job for execution.
 
         :param Job job: job to execute
-        :param list[datetime] run_times: list of datetimes specifying when the job should have been run
-        :raises MaxInstancesReachedError: if the maximum number of allowed instances for this job has been reached
-        """
+        :param list[datetime] run_times: list of datetimes specifying
+            when the job should have been run
+        :raises MaxInstancesReachedError: if the maximum number of
+            allowed instances for this job has been reached
 
+        """
         assert self._lock is not None, 'This executor has not been started yet'
         with self._lock:
             if self._instances[job.id] >= job.max_instances:
@@ -70,50 +76,71 @@ class BaseExecutor(six.with_metaclass(ABCMeta, object)):
         """Performs the actual task of scheduling `run_job` to be called."""
 
     def _run_job_success(self, job_id, events):
-        """Called by the executor with the list of generated events when `run_job` has been successfully called."""
+        """
+        Called by the executor with the list of generated events when :func:`run_job` has been
+        successfully called.
 
+        """
         with self._lock:
             self._instances[job_id] -= 1
+            if self._instances[job_id] == 0:
+                del self._instances[job_id]
 
         for event in events:
             self._scheduler._dispatch_event(event)
 
     def _run_job_error(self, job_id, exc, traceback=None):
-        """Called by the executor with the exception if there is an error calling `run_job`."""
-
+        """Called by the executor with the exception if there is an error  calling `run_job`."""
         with self._lock:
             self._instances[job_id] -= 1
+            if self._instances[job_id] == 0:
+                del self._instances[job_id]
 
         exc_info = (exc.__class__, exc, traceback)
         self._logger.error('Error running job %s', job_id, exc_info=exc_info)
 
 
 def run_job(job, jobstore_alias, run_times, logger_name):
-    """Called by executors to run the job. Returns a list of scheduler events to be dispatched by the scheduler."""
+    """
+    Called by executors to run the job. Returns a list of scheduler events to be dispatched by the
+    scheduler.
 
+    """
     events = []
     logger = logging.getLogger(logger_name)
     for run_time in run_times:
-        # See if the job missed its run time window, and handle possible misfires accordingly
+        # See if the job missed its run time window, and handle
+        # possible misfires accordingly
         if job.misfire_grace_time is not None:
             difference = datetime.now(utc) - run_time
             grace_time = timedelta(seconds=job.misfire_grace_time)
             if difference > grace_time:
-                events.append(JobExecutionEvent(EVENT_JOB_MISSED, job.id, jobstore_alias, run_time))
+                events.append(JobExecutionEvent(EVENT_JOB_MISSED, job.id, jobstore_alias,
+                                                run_time))
                 logger.warning('Run time of job "%s" was missed by %s', job, difference)
                 continue
 
         logger.info('Running job "%s" (scheduled at %s)', job, run_time)
         try:
             retval = job.func(*job.args, **job.kwargs)
-        except:
+        except BaseException:
             exc, tb = sys.exc_info()[1:]
             formatted_tb = ''.join(format_tb(tb))
-            events.append(JobExecutionEvent(EVENT_JOB_ERROR, job.id, jobstore_alias, run_time, exception=exc,
-                                            traceback=formatted_tb))
+            events.append(JobExecutionEvent(EVENT_JOB_ERROR, job.id, jobstore_alias, run_time,
+                                            exception=exc, traceback=formatted_tb))
             logger.exception('Job "%s" raised an exception', job)
+
+            # This is to prevent cyclic references that would lead to memory leaks
+            if six.PY2:
+                sys.exc_clear()
+                del tb
+            else:
+                import traceback
+                traceback.clear_frames(tb)
+                del tb
         else:
-            events.append(JobExecutionEvent(EVENT_JOB_EXECUTED, job.id, jobstore_alias, run_time, retval=retval))
+            events.append(JobExecutionEvent(EVENT_JOB_EXECUTED, job.id, jobstore_alias, run_time,
+                                            retval=retval))
             logger.info('Job "%s" executed successfully', job)
 
     return events

lib/apscheduler/executors/base_py3.py

@@ -0,0 +1,43 @@
+import logging
+import sys
+import traceback
+from datetime import datetime, timedelta
+from traceback import format_tb
+
+from pytz import utc
+
+from apscheduler.events import (
+    JobExecutionEvent, EVENT_JOB_MISSED, EVENT_JOB_ERROR, EVENT_JOB_EXECUTED)
+
+
+async def run_coroutine_job(job, jobstore_alias, run_times, logger_name):
+    """Coroutine version of run_job()."""
+    events = []
+    logger = logging.getLogger(logger_name)
+    for run_time in run_times:
+        # See if the job missed its run time window, and handle possible misfires accordingly
+        if job.misfire_grace_time is not None:
+            difference = datetime.now(utc) - run_time
+            grace_time = timedelta(seconds=job.misfire_grace_time)
+            if difference > grace_time:
+                events.append(JobExecutionEvent(EVENT_JOB_MISSED, job.id, jobstore_alias,
+                                                run_time))
+                logger.warning('Run time of job "%s" was missed by %s', job, difference)
+                continue
+
+        logger.info('Running job "%s" (scheduled at %s)', job, run_time)
+        try:
+            retval = await job.func(*job.args, **job.kwargs)
+        except BaseException:
+            exc, tb = sys.exc_info()[1:]
+            formatted_tb = ''.join(format_tb(tb))
+            events.append(JobExecutionEvent(EVENT_JOB_ERROR, job.id, jobstore_alias, run_time,
+                                            exception=exc, traceback=formatted_tb))
+            logger.exception('Job "%s" raised an exception', job)
+            traceback.clear_frames(tb)
+        else:
+            events.append(JobExecutionEvent(EVENT_JOB_EXECUTED, job.id, jobstore_alias, run_time,
+                                            retval=retval))
+            logger.info('Job "%s" executed successfully', job)
+
+    return events

lib/apscheduler/executors/debug.py

@@ -5,7 +5,8 @@ from apscheduler.executors.base import BaseExecutor, run_job
 
 class DebugExecutor(BaseExecutor):
     """
-    A special executor that executes the target callable directly instead of deferring it to a thread or process.
+    A special executor that executes the target callable directly instead of deferring it to a
+    thread or process.
 
     Plugin alias: ``debug``
     """
@@ -13,7 +14,7 @@ class DebugExecutor(BaseExecutor):
     def _do_submit_job(self, job, run_times):
         try:
             events = run_job(job, job._jobstore_alias, run_times, self._logger.name)
-        except:
+        except BaseException:
             self._run_job_error(job.id, *sys.exc_info()[1:])
         else:
             self._run_job_success(job.id, events)

lib/apscheduler/executors/gevent.py

@@ -1,4 +1,4 @@
-
+from __future__ import absolute_import
 import sys
 
 from apscheduler.executors.base import BaseExecutor, run_job
@@ -21,9 +21,10 @@ class GeventExecutor(BaseExecutor):
         def callback(greenlet):
             try:
                 events = greenlet.get()
-            except:
+            except BaseException:
                 self._run_job_error(job.id, *sys.exc_info()[1:])
             else:
                 self._run_job_success(job.id, events)
 
-        gevent.spawn(run_job, job, job._jobstore_alias, run_times, self._logger.name).link(callback)
+        gevent.spawn(run_job, job, job._jobstore_alias, run_times, self._logger.name).\
+            link(callback)

lib/apscheduler/executors/pool.py

@@ -3,6 +3,11 @@ import concurrent.futures
 
 from apscheduler.executors.base import BaseExecutor, run_job
 
+try:
+    from concurrent.futures.process import BrokenProcessPool
+except ImportError:
+    BrokenProcessPool = None
+
 
 class BasePoolExecutor(BaseExecutor):
     @abstractmethod
@@ -19,7 +24,13 @@ class BasePoolExecutor(BaseExecutor):
             else:
                 self._run_job_success(job.id, f.result())
 
-        f = self._pool.submit(run_job, job, job._jobstore_alias, run_times, self._logger.name)
+        try:
+            f = self._pool.submit(run_job, job, job._jobstore_alias, run_times, self._logger.name)
+        except BrokenProcessPool:
+            self._logger.warning('Process pool is broken; replacing pool with a fresh instance')
+            self._pool = self._pool.__class__(self._pool._max_workers)
+            f = self._pool.submit(run_job, job, job._jobstore_alias, run_times, self._logger.name)
+
         f.add_done_callback(callback)
 
     def shutdown(self, wait=True):
@@ -33,10 +44,13 @@ class ThreadPoolExecutor(BasePoolExecutor):
     Plugin alias: ``threadpool``
 
     :param max_workers: the maximum number of spawned threads.
+    :param pool_kwargs: dict of keyword arguments to pass to the underlying
+        ThreadPoolExecutor constructor
     """
 
-    def __init__(self, max_workers=10):
-        pool = concurrent.futures.ThreadPoolExecutor(int(max_workers))
+    def __init__(self, max_workers=10, pool_kwargs=None):
+        pool_kwargs = pool_kwargs or {}
+        pool = concurrent.futures.ThreadPoolExecutor(int(max_workers), **pool_kwargs)
         super(ThreadPoolExecutor, self).__init__(pool)
 
 
@@ -47,8 +61,11 @@ class ProcessPoolExecutor(BasePoolExecutor):
     Plugin alias: ``processpool``
 
     :param max_workers: the maximum number of spawned processes.
+    :param pool_kwargs: dict of keyword arguments to pass to the underlying
+        ProcessPoolExecutor constructor
     """
 
-    def __init__(self, max_workers=10):
-        pool = concurrent.futures.ProcessPoolExecutor(int(max_workers))
+    def __init__(self, max_workers=10, pool_kwargs=None):
+        pool_kwargs = pool_kwargs or {}
+        pool = concurrent.futures.ProcessPoolExecutor(int(max_workers), **pool_kwargs)
         super(ProcessPoolExecutor, self).__init__(pool)

lib/apscheduler/executors/tornado.py

@@ -0,0 +1,54 @@
+from __future__ import absolute_import
+
+import sys
+from concurrent.futures import ThreadPoolExecutor
+
+from tornado.gen import convert_yielded
+
+from apscheduler.executors.base import BaseExecutor, run_job
+
+try:
+    from apscheduler.executors.base_py3 import run_coroutine_job
+    from apscheduler.util import iscoroutinefunction_partial
+except ImportError:
+    def iscoroutinefunction_partial(func):
+        return False
+
+
+class TornadoExecutor(BaseExecutor):
+    """
+    Runs jobs either in a thread pool or directly on the I/O loop.
+
+    If the job function is a native coroutine function, it is scheduled to be run directly in the
+    I/O loop as soon as possible. All other functions are run in a thread pool.
+
+    Plugin alias: ``tornado``
+
+    :param int max_workers: maximum number of worker threads in the thread pool
+    """
+
+    def __init__(self, max_workers=10):
+        super(TornadoExecutor, self).__init__()
+        self.executor = ThreadPoolExecutor(max_workers)
+
+    def start(self, scheduler, alias):
+        super(TornadoExecutor, self).start(scheduler, alias)
+        self._ioloop = scheduler._ioloop
+
+    def _do_submit_job(self, job, run_times):
+        def callback(f):
+            try:
+                events = f.result()
+            except BaseException:
+                self._run_job_error(job.id, *sys.exc_info()[1:])
+            else:
+                self._run_job_success(job.id, events)
+
+        if iscoroutinefunction_partial(job.func):
+            f = run_coroutine_job(job, job._jobstore_alias, run_times, self._logger.name)
+        else:
+            f = self.executor.submit(run_job, job, job._jobstore_alias, run_times,
+                                     self._logger.name)
+
+        f = convert_yielded(f)
+        f.add_done_callback(callback)

lib/apscheduler/executors/twisted.py

@@ -1,4 +1,4 @@
-
+from __future__ import absolute_import
 
 from apscheduler.executors.base import BaseExecutor, run_job
 
@@ -21,5 +21,5 @@ class TwistedExecutor(BaseExecutor):
             else:
                 self._run_job_error(job.id, result.value, result.tb)
 
-        self._reactor.getThreadPool().callInThreadWithCallback(callback, run_job, job, job._jobstore_alias, run_times,
-                                                               self._logger.name)
+        self._reactor.getThreadPool().callInThreadWithCallback(
+            callback, run_job, job, job._jobstore_alias, run_times, self._logger.name)

lib/apscheduler/job.py

@@ -1,11 +1,17 @@
-from collections.abc import Iterable, Mapping
+from inspect import ismethod, isclass
 from uuid import uuid4
 
 import six
 
 from apscheduler.triggers.base import BaseTrigger
-from apscheduler.util import ref_to_obj, obj_to_ref, datetime_repr, repr_escape, get_callable_name, check_callable_args, \
-    convert_to_datetime
+from apscheduler.util import (
+    ref_to_obj, obj_to_ref, datetime_repr, repr_escape, get_callable_name, check_callable_args,
+    convert_to_datetime)
+
+try:
+    from collections.abc import Iterable, Mapping
+except ImportError:
+    from collections import Iterable, Mapping
 
 
 class Job(object):
@@ -21,13 +27,20 @@ class Job(object):
     :var bool coalesce: whether to only run the job once when several run times are due
     :var trigger: the trigger object that controls the schedule of this job
     :var str executor: the name of the executor that will run this job
-    :var int misfire_grace_time: the time (in seconds) how much this job's execution is allowed to be late
-    :var int max_instances: the maximum number of concurrently executing instances allowed for this job
+    :var int misfire_grace_time: the time (in seconds) how much this job's execution is allowed to
+        be late (``None`` means "allow the job to run no matter how late it is")
+    :var int max_instances: the maximum number of concurrently executing instances allowed for this
+        job
     :var datetime.datetime next_run_time: the next scheduled run time of this job
+
+    .. note::
+        The ``misfire_grace_time`` has some non-obvious effects on job execution. See the
+        :ref:`missed-job-executions` section in the documentation for an in-depth explanation.
     """
 
-    __slots__ = ('_scheduler', '_jobstore_alias', 'id', 'trigger', 'executor', 'func', 'func_ref', 'args', 'kwargs',
-                 'name', 'misfire_grace_time', 'coalesce', 'max_instances', 'next_run_time')
+    __slots__ = ('_scheduler', '_jobstore_alias', 'id', 'trigger', 'executor', 'func', 'func_ref',
+                 'args', 'kwargs', 'name', 'misfire_grace_time', 'coalesce', 'max_instances',
+                 'next_run_time', '__weakref__')
 
     def __init__(self, scheduler, id=None, **kwargs):
         super(Job, self).__init__()
@@ -38,53 +51,69 @@ class Job(object):
     def modify(self, **changes):
         """
         Makes the given changes to this job and saves it in the associated job store.
+
         Accepted keyword arguments are the same as the variables on this class.
 
         .. seealso:: :meth:`~apscheduler.schedulers.base.BaseScheduler.modify_job`
-        """
 
+        :return Job: this job instance
+
+        """
         self._scheduler.modify_job(self.id, self._jobstore_alias, **changes)
+        return self
 
     def reschedule(self, trigger, **trigger_args):
         """
         Shortcut for switching the trigger on this job.
 
         .. seealso:: :meth:`~apscheduler.schedulers.base.BaseScheduler.reschedule_job`
-        """
 
+        :return Job: this job instance
+
+        """
         self._scheduler.reschedule_job(self.id, self._jobstore_alias, trigger, **trigger_args)
+        return self
 
     def pause(self):
         """
         Temporarily suspend the execution of this job.
 
         .. seealso:: :meth:`~apscheduler.schedulers.base.BaseScheduler.pause_job`
-        """
 
+        :return Job: this job instance
+
+        """
         self._scheduler.pause_job(self.id, self._jobstore_alias)
+        return self
 
     def resume(self):
         """
         Resume the schedule of this job if previously paused.
 
         .. seealso:: :meth:`~apscheduler.schedulers.base.BaseScheduler.resume_job`
-        """
 
+        :return Job: this job instance
+
+        """
         self._scheduler.resume_job(self.id, self._jobstore_alias)
+        return self
 
     def remove(self):
         """
         Unschedules this job and removes it from its associated job store.
 
         .. seealso:: :meth:`~apscheduler.schedulers.base.BaseScheduler.remove_job`
-        """
 
+        """
         self._scheduler.remove_job(self.id, self._jobstore_alias)
 
     @property
     def pending(self):
-        """Returns ``True`` if the referenced job is still waiting to be added to its designated job store."""
+        """
+        Returns ``True`` if the referenced job is still waiting to be added to its designated job
+        store.
 
+        """
         return self._jobstore_alias is None
 
     #
@@ -97,8 +126,8 @@ class Job(object):
 
         :type now: datetime.datetime
         :rtype: list[datetime.datetime]
-        """
 
+        """
         run_times = []
         next_run_time = self.next_run_time
         while next_run_time and next_run_time <= now:
@@ -108,8 +137,11 @@ class Job(object):
         return run_times
 
     def _modify(self, **changes):
-        """Validates the changes to the Job and makes the modifications if and only if all of them validate."""
+        """
+        Validates the changes to the Job and makes the modifications if and only if all of them
+        validate.
 
+        """
         approved = {}
 
         if 'id' in changes:
@@ -125,7 +157,7 @@ class Job(object):
             args = changes.pop('args') if 'args' in changes else self.args
             kwargs = changes.pop('kwargs') if 'kwargs' in changes else self.kwargs
 
-            if isinstance(func, str):
+            if isinstance(func, six.string_types):
                 func_ref = func
                 func = ref_to_obj(func)
             elif callable(func):
@@ -177,7 +209,8 @@ class Job(object):
         if 'trigger' in changes:
             trigger = changes.pop('trigger')
             if not isinstance(trigger, BaseTrigger):
-                raise TypeError('Expected a trigger instance, got %s instead' % trigger.__class__.__name__)
+                raise TypeError('Expected a trigger instance, got %s instead' %
+                                trigger.__class__.__name__)
 
             approved['trigger'] = trigger
 
@@ -189,10 +222,12 @@ class Job(object):
 
         if 'next_run_time' in changes:
             value = changes.pop('next_run_time')
-            approved['next_run_time'] = convert_to_datetime(value, self._scheduler.timezone, 'next_run_time')
+            approved['next_run_time'] = convert_to_datetime(value, self._scheduler.timezone,
+                                                            'next_run_time')
 
         if changes:
-            raise AttributeError('The following are not modifiable attributes of Job: %s' % ', '.join(changes))
+            raise AttributeError('The following are not modifiable attributes of Job: %s' %
+                                 ', '.join(changes))
 
         for key, value in six.iteritems(approved):
             setattr(self, key, value)
@@ -200,9 +235,18 @@ class Job(object):
     def __getstate__(self):
         # Don't allow this Job to be serialized if the function reference could not be determined
         if not self.func_ref:
-            raise ValueError('This Job cannot be serialized since the reference to its callable (%r) could not be '
-                             'determined. Consider giving a textual reference (module:function name) instead.' %
-                             (self.func,))
+            raise ValueError(
+                'This Job cannot be serialized since the reference to its callable (%r) could not '
+                'be determined. Consider giving a textual reference (module:function name) '
+                'instead.' % (self.func,))
+
+        # Instance methods cannot survive serialization as-is, so store the "self" argument
+        # explicitly
+        func = self.func
+        if ismethod(func) and not isclass(func.__self__) and obj_to_ref(func) == self.func_ref:
+            args = (func.__self__,) + tuple(self.args)
+        else:
+            args = self.args
 
         return {
             'version': 1,
@@ -210,7 +254,7 @@ class Job(object):
             'func': self.func_ref,
             'trigger': self.trigger,
             'executor': self.executor,
-            'args': self.args,
+            'args': args,
             'kwargs': self.kwargs,
             'name': self.name,
             'misfire_grace_time': self.misfire_grace_time,
@@ -221,7 +265,8 @@ class Job(object):
 
     def __setstate__(self, state):
         if state.get('version', 1) > 1:
-            raise ValueError('Job has version %s, but only version 1 can be handled' % state['version'])
+            raise ValueError('Job has version %s, but only version 1 can be handled' %
+                             state['version'])
 
         self.id = state['id']
         self.func_ref = state['func']
@@ -245,8 +290,13 @@ class Job(object):
         return '<Job (id=%s name=%s)>' % (repr_escape(self.id), repr_escape(self.name))
 
     def __str__(self):
-        return '%s (trigger: %s, next run at: %s)' % (repr_escape(self.name), repr_escape(str(self.trigger)),
-                                                      datetime_repr(self.next_run_time))
+        return repr_escape(self.__unicode__())
 
     def __unicode__(self):
-        return six.u('%s (trigger: %s, next run at: %s)') % (self.name, self.trigger, datetime_repr(self.next_run_time))
+        if hasattr(self, 'next_run_time'):
+            status = ('next run at: ' + datetime_repr(self.next_run_time) if
+                      self.next_run_time else 'paused')
+        else:
+            status = 'pending'
+
+        return u'%s (trigger: %s, %s)' % (self.name, self.trigger, status)

lib/apscheduler/jobstores/base.py

@@ -8,23 +8,27 @@ class JobLookupError(KeyError):
     """Raised when the job store cannot find a job for update or removal."""
 
     def __init__(self, job_id):
-        super(JobLookupError, self).__init__(six.u('No job by the id of %s was found') % job_id)
+        super(JobLookupError, self).__init__(u'No job by the id of %s was found' % job_id)
 
 
 class ConflictingIdError(KeyError):
     """Raised when the uniqueness of job IDs is being violated."""
 
     def __init__(self, job_id):
-        super(ConflictingIdError, self).__init__(six.u('Job identifier (%s) conflicts with an existing job') % job_id)
+        super(ConflictingIdError, self).__init__(
+            u'Job identifier (%s) conflicts with an existing job' % job_id)
 
 
 class TransientJobError(ValueError):
-    """Raised when an attempt to add transient (with no func_ref) job to a persistent job store is detected."""
+    """
+    Raised when an attempt to add transient (with no func_ref) job to a persistent job store is
+    detected.
+    """
 
     def __init__(self, job_id):
         super(TransientJobError, self).__init__(
-            six.u('Job (%s) cannot be added to this job store because a reference to the callable could not be '
-                  'determined.') % job_id)
+            u'Job (%s) cannot be added to this job store because a reference to the callable '
+            u'could not be determined.' % job_id)
 
 
 class BaseJobStore(six.with_metaclass(ABCMeta)):
@@ -36,10 +40,11 @@ class BaseJobStore(six.with_metaclass(ABCMeta)):
 
     def start(self, scheduler, alias):
         """
-        Called by the scheduler when the scheduler is being started or when the job store is being added to an already
-        running scheduler.
+        Called by the scheduler when the scheduler is being started or when the job store is being
+        added to an already running scheduler.
 
-        :param apscheduler.schedulers.base.BaseScheduler scheduler: the scheduler that is starting this job store
+        :param apscheduler.schedulers.base.BaseScheduler scheduler: the scheduler that is starting
+            this job store
         :param str|unicode alias: alias of this job store as it was assigned to the scheduler
         """
 
@@ -50,13 +55,22 @@ class BaseJobStore(six.with_metaclass(ABCMeta)):
     def shutdown(self):
         """Frees any resources still bound to this job store."""
 
+    def _fix_paused_jobs_sorting(self, jobs):
+        for i, job in enumerate(jobs):
+            if job.next_run_time is not None:
+                if i > 0:
+                    paused_jobs = jobs[:i]
+                    del jobs[:i]
+                    jobs.extend(paused_jobs)
+                break
+
     @abstractmethod
     def lookup_job(self, job_id):
         """
         Returns a specific job, or ``None`` if it isn't found..
 
-        The job store is responsible for setting the ``scheduler`` and ``jobstore`` attributes of the returned job to
-        point to the scheduler and itself, respectively.
+        The job store is responsible for setting the ``scheduler`` and ``jobstore`` attributes of
+        the returned job to point to the scheduler and itself, respectively.
 
         :param str|unicode job_id: identifier of the job
         :rtype: Job
@@ -75,7 +89,8 @@ class BaseJobStore(six.with_metaclass(ABCMeta)):
     @abstractmethod
     def get_next_run_time(self):
         """
-        Returns the earliest run time of all the jobs stored in this job store, or ``None`` if there are no active jobs.
+        Returns the earliest run time of all the jobs stored in this job store, or ``None`` if
+        there are no active jobs.
 
         :rtype: datetime.datetime
         """
@@ -83,11 +98,12 @@ class BaseJobStore(six.with_metaclass(ABCMeta)):
     @abstractmethod
     def get_all_jobs(self):
         """
-        Returns a list of all jobs in this job store. The returned jobs should be sorted by next run time (ascending).
-        Paused jobs (next_run_time is None) should be sorted last.
+        Returns a list of all jobs in this job store.
+        The returned jobs should be sorted by next run time (ascending).
+        Paused jobs (next_run_time == None) should be sorted last.
 
-        The job store is responsible for setting the ``scheduler`` and ``jobstore`` attributes of the returned jobs to
-        point to the scheduler and itself, respectively.
+        The job store is responsible for setting the ``scheduler`` and ``jobstore`` attributes of
+        the returned jobs to point to the scheduler and itself, respectively.
 
         :rtype: list[Job]
         """

lib/apscheduler/jobstores/memory.py

@@ -1,4 +1,4 @@
-
+from __future__ import absolute_import
 
 from apscheduler.jobstores.base import BaseJobStore, JobLookupError, ConflictingIdError
 from apscheduler.util import datetime_to_utc_timestamp
@@ -13,7 +13,8 @@ class MemoryJobStore(BaseJobStore):
 
     def __init__(self):
         super(MemoryJobStore, self).__init__()
-        self._jobs = []  # list of (job, timestamp), sorted by next_run_time and job id (ascending)
+        # list of (job, timestamp), sorted by next_run_time and job id (ascending)
+        self._jobs = []
         self._jobs_index = {}  # id -> (job, timestamp) lookup table
 
     def lookup_job(self, job_id):
@@ -80,13 +81,13 @@ class MemoryJobStore(BaseJobStore):
 
     def _get_job_index(self, timestamp, job_id):
         """
-        Returns the index of the given job, or if it's not found, the index where the job should be inserted based on
-        the given timestamp.
+        Returns the index of the given job, or if it's not found, the index where the job should be
+        inserted based on the given timestamp.
 
         :type timestamp: int
         :type job_id: str
-        """
 
+        """
         lo, hi = 0, len(self._jobs)
         timestamp = float('inf') if timestamp is None else timestamp
         while lo < hi:

lib/apscheduler/jobstores/mongodb.py

@@ -1,11 +1,12 @@
-
+from __future__ import absolute_import
+import warnings
 
 from apscheduler.jobstores.base import BaseJobStore, JobLookupError, ConflictingIdError
 from apscheduler.util import maybe_ref, datetime_to_utc_timestamp, utc_timestamp_to_datetime
 from apscheduler.job import Job
 
 try:
-    import pickle as pickle
+    import cPickle as pickle
 except ImportError:  # pragma: nocover
     import pickle
 
@@ -19,16 +20,18 @@ except ImportError:  # pragma: nocover
 
 class MongoDBJobStore(BaseJobStore):
     """
-    Stores jobs in a MongoDB database. Any leftover keyword arguments are directly passed to pymongo's `MongoClient
+    Stores jobs in a MongoDB database. Any leftover keyword arguments are directly passed to
+    pymongo's `MongoClient
     <http://api.mongodb.org/python/current/api/pymongo/mongo_client.html#pymongo.mongo_client.MongoClient>`_.
 
     Plugin alias: ``mongodb``
 
     :param str database: database to store jobs in
     :param str collection: collection to store jobs in
-    :param client: a :class:`~pymongo.mongo_client.MongoClient` instance to use instead of providing connection
-                   arguments
-    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the highest available
+    :param client: a :class:`~pymongo.mongo_client.MongoClient` instance to use instead of
+        providing connection arguments
+    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the
+        highest available
     """
 
     def __init__(self, database='apscheduler', collection='jobs', client=None,
@@ -42,13 +45,22 @@ class MongoDBJobStore(BaseJobStore):
             raise ValueError('The "collection" parameter must not be empty')
 
         if client:
-            self.connection = maybe_ref(client)
+            self.client = maybe_ref(client)
         else:
             connect_args.setdefault('w', 1)
-            self.connection = MongoClient(**connect_args)
+            self.client = MongoClient(**connect_args)
 
-        self.collection = self.connection[database][collection]
-        self.collection.ensure_index('next_run_time', sparse=True)
+        self.collection = self.client[database][collection]
+
+    def start(self, scheduler, alias):
+        super(MongoDBJobStore, self).start(scheduler, alias)
+        self.collection.create_index('next_run_time', sparse=True)
+
+    @property
+    def connection(self):
+        warnings.warn('The "connection" member is deprecated -- use "client" instead',
+                      DeprecationWarning)
+        return self.client
 
     def lookup_job(self, job_id):
         document = self.collection.find_one(job_id, ['job_state'])
@@ -59,16 +71,19 @@ class MongoDBJobStore(BaseJobStore):
         return self._get_jobs({'next_run_time': {'$lte': timestamp}})
 
     def get_next_run_time(self):
-        document = self.collection.find_one({'next_run_time': {'$ne': None}}, fields=['next_run_time'],
+        document = self.collection.find_one({'next_run_time': {'$ne': None}},
+                                            projection=['next_run_time'],
                                             sort=[('next_run_time', ASCENDING)])
         return utc_timestamp_to_datetime(document['next_run_time']) if document else None
 
     def get_all_jobs(self):
-        return self._get_jobs({})
+        jobs = self._get_jobs({})
+        self._fix_paused_jobs_sorting(jobs)
+        return jobs
 
     def add_job(self, job):
         try:
-            self.collection.insert({
+            self.collection.insert_one({
                 '_id': job.id,
                 'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
                 'job_state': Binary(pickle.dumps(job.__getstate__(), self.pickle_protocol))
@@ -81,20 +96,20 @@ class MongoDBJobStore(BaseJobStore):
             'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
             'job_state': Binary(pickle.dumps(job.__getstate__(), self.pickle_protocol))
         }
-        result = self.collection.update({'_id': job.id}, {'$set': changes})
-        if result and result['n'] == 0:
-            raise JobLookupError(id)
+        result = self.collection.update_one({'_id': job.id}, {'$set': changes})
+        if result and result.matched_count == 0:
+            raise JobLookupError(job.id)
 
     def remove_job(self, job_id):
-        result = self.collection.remove(job_id)
-        if result and result['n'] == 0:
+        result = self.collection.delete_one({'_id': job_id})
+        if result and result.deleted_count == 0:
             raise JobLookupError(job_id)
 
     def remove_all_jobs(self):
-        self.collection.remove()
+        self.collection.delete_many({})
 
     def shutdown(self):
-        self.connection.disconnect()
+        self.client.close()
 
     def _reconstitute_job(self, job_state):
         job_state = pickle.loads(job_state)
@@ -107,18 +122,20 @@ class MongoDBJobStore(BaseJobStore):
     def _get_jobs(self, conditions):
         jobs = []
         failed_job_ids = []
-        for document in self.collection.find(conditions, ['_id', 'job_state'], sort=[('next_run_time', ASCENDING)]):
+        for document in self.collection.find(conditions, ['_id', 'job_state'],
+                                             sort=[('next_run_time', ASCENDING)]):
             try:
                 jobs.append(self._reconstitute_job(document['job_state']))
-            except:
-                self._logger.exception('Unable to restore job "%s" -- removing it', document['_id'])
+            except BaseException:
+                self._logger.exception('Unable to restore job "%s" -- removing it',
+                                       document['_id'])
                 failed_job_ids.append(document['_id'])
 
         # Remove all the jobs we failed to restore
         if failed_job_ids:
-            self.collection.remove({'_id': {'$in': failed_job_ids}})
+            self.collection.delete_many({'_id': {'$in': failed_job_ids}})
 
         return jobs
 
     def __repr__(self):
-        return '<%s (client=%s)>' % (self.__class__.__name__, self.connection)
+        return '<%s (client=%s)>' % (self.__class__.__name__, self.client)

lib/apscheduler/jobstores/redis.py

@@ -1,5 +1,7 @@
+from __future__ import absolute_import
+from datetime import datetime
 
-
+from pytz import utc
 import six
 
 from apscheduler.jobstores.base import BaseJobStore, JobLookupError, ConflictingIdError
@@ -7,26 +9,28 @@ from apscheduler.util import datetime_to_utc_timestamp, utc_timestamp_to_datetim
 from apscheduler.job import Job
 
 try:
-    import pickle as pickle
+    import cPickle as pickle
 except ImportError:  # pragma: nocover
     import pickle
 
 try:
-    from redis import StrictRedis
+    from redis import Redis
 except ImportError:  # pragma: nocover
     raise ImportError('RedisJobStore requires redis installed')
 
 
 class RedisJobStore(BaseJobStore):
     """
-    Stores jobs in a Redis database. Any leftover keyword arguments are directly passed to redis's StrictRedis.
+    Stores jobs in a Redis database. Any leftover keyword arguments are directly passed to redis's
+    :class:`~redis.StrictRedis`.
 
     Plugin alias: ``redis``
 
     :param int db: the database number to store jobs in
     :param str jobs_key: key to store jobs in
     :param str run_times_key: key to store the jobs' run times in
-    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the highest available
+    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the
+        highest available
     """
 
     def __init__(self, db=0, jobs_key='apscheduler.jobs', run_times_key='apscheduler.run_times',
@@ -43,7 +47,7 @@ class RedisJobStore(BaseJobStore):
         self.pickle_protocol = pickle_protocol
         self.jobs_key = jobs_key
         self.run_times_key = run_times_key
-        self.redis = StrictRedis(db=int(db), **connect_args)
+        self.redis = Redis(db=int(db), **connect_args)
 
     def lookup_job(self, job_id):
         job_state = self.redis.hget(self.jobs_key, job_id)
@@ -65,7 +69,8 @@ class RedisJobStore(BaseJobStore):
     def get_all_jobs(self):
         job_states = self.redis.hgetall(self.jobs_key)
         jobs = self._reconstitute_jobs(six.iteritems(job_states))
-        return sorted(jobs, key=lambda job: job.next_run_time)
+        paused_sort_key = datetime(9999, 12, 31, tzinfo=utc)
+        return sorted(jobs, key=lambda job: job.next_run_time or paused_sort_key)
 
     def add_job(self, job):
         if self.redis.hexists(self.jobs_key, job.id):
@@ -73,8 +78,12 @@ class RedisJobStore(BaseJobStore):
 
         with self.redis.pipeline() as pipe:
             pipe.multi()
-            pipe.hset(self.jobs_key, job.id, pickle.dumps(job.__getstate__(), self.pickle_protocol))
-            pipe.zadd(self.run_times_key, datetime_to_utc_timestamp(job.next_run_time), job.id)
+            pipe.hset(self.jobs_key, job.id, pickle.dumps(job.__getstate__(),
+                                                          self.pickle_protocol))
+            if job.next_run_time:
+                pipe.zadd(self.run_times_key,
+                          {job.id: datetime_to_utc_timestamp(job.next_run_time)})
+
             pipe.execute()
 
     def update_job(self, job):
@@ -82,11 +91,14 @@ class RedisJobStore(BaseJobStore):
             raise JobLookupError(job.id)
 
         with self.redis.pipeline() as pipe:
-            pipe.hset(self.jobs_key, job.id, pickle.dumps(job.__getstate__(), self.pickle_protocol))
+            pipe.hset(self.jobs_key, job.id, pickle.dumps(job.__getstate__(),
+                                                          self.pickle_protocol))
             if job.next_run_time:
-                pipe.zadd(self.run_times_key, datetime_to_utc_timestamp(job.next_run_time), job.id)
+                pipe.zadd(self.run_times_key,
+                          {job.id: datetime_to_utc_timestamp(job.next_run_time)})
             else:
                 pipe.zrem(self.run_times_key, job.id)
+
             pipe.execute()
 
     def remove_job(self, job_id):
@@ -121,7 +133,7 @@ class RedisJobStore(BaseJobStore):
         for job_id, job_state in job_states:
             try:
                 jobs.append(self._reconstitute_job(job_state))
-            except:
+            except BaseException:
                 self._logger.exception('Unable to restore job "%s" -- removing it', job_id)
                 failed_job_ids.append(job_id)
 

lib/apscheduler/jobstores/rethinkdb.py

@@ -0,0 +1,155 @@
+from __future__ import absolute_import
+
+from apscheduler.jobstores.base import BaseJobStore, JobLookupError, ConflictingIdError
+from apscheduler.util import maybe_ref, datetime_to_utc_timestamp, utc_timestamp_to_datetime
+from apscheduler.job import Job
+
+try:
+    import cPickle as pickle
+except ImportError:  # pragma: nocover
+    import pickle
+
+try:
+    from rethinkdb import RethinkDB
+except ImportError:  # pragma: nocover
+    raise ImportError('RethinkDBJobStore requires rethinkdb installed')
+
+
+class RethinkDBJobStore(BaseJobStore):
+    """
+    Stores jobs in a RethinkDB database. Any leftover keyword arguments are directly passed to
+    rethinkdb's `RethinkdbClient <http://www.rethinkdb.com/api/#connect>`_.
+
+    Plugin alias: ``rethinkdb``
+
+    :param str database: database to store jobs in
+    :param str collection: collection to store jobs in
+    :param client: a :class:`rethinkdb.net.Connection` instance to use instead of providing
+        connection arguments
+    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the
+        highest available
+    """
+
+    def __init__(self, database='apscheduler', table='jobs', client=None,
+                 pickle_protocol=pickle.HIGHEST_PROTOCOL, **connect_args):
+        super(RethinkDBJobStore, self).__init__()
+
+        if not database:
+            raise ValueError('The "database" parameter must not be empty')
+        if not table:
+            raise ValueError('The "table" parameter must not be empty')
+
+        self.database = database
+        self.table_name = table
+        self.table = None
+        self.client = client
+        self.pickle_protocol = pickle_protocol
+        self.connect_args = connect_args
+        self.r = RethinkDB()
+        self.conn = None
+
+    def start(self, scheduler, alias):
+        super(RethinkDBJobStore, self).start(scheduler, alias)
+
+        if self.client:
+            self.conn = maybe_ref(self.client)
+        else:
+            self.conn = self.r.connect(db=self.database, **self.connect_args)
+
+        if self.database not in self.r.db_list().run(self.conn):
+            self.r.db_create(self.database).run(self.conn)
+
+        if self.table_name not in self.r.table_list().run(self.conn):
+            self.r.table_create(self.table_name).run(self.conn)
+
+        if 'next_run_time' not in self.r.table(self.table_name).index_list().run(self.conn):
+            self.r.table(self.table_name).index_create('next_run_time').run(self.conn)
+
+        self.table = self.r.db(self.database).table(self.table_name)
+
+    def lookup_job(self, job_id):
+        results = list(self.table.get_all(job_id).pluck('job_state').run(self.conn))
+        return self._reconstitute_job(results[0]['job_state']) if results else None
+
+    def get_due_jobs(self, now):
+        return self._get_jobs(self.r.row['next_run_time'] <= datetime_to_utc_timestamp(now))
+
+    def get_next_run_time(self):
+        results = list(
+            self.table
+            .filter(self.r.row['next_run_time'] != None)  # noqa
+            .order_by(self.r.asc('next_run_time'))
+            .map(lambda x: x['next_run_time'])
+            .limit(1)
+            .run(self.conn)
+        )
+        return utc_timestamp_to_datetime(results[0]) if results else None
+
+    def get_all_jobs(self):
+        jobs = self._get_jobs()
+        self._fix_paused_jobs_sorting(jobs)
+        return jobs
+
+    def add_job(self, job):
+        job_dict = {
+            'id': job.id,
+            'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
+            'job_state': self.r.binary(pickle.dumps(job.__getstate__(), self.pickle_protocol))
+        }
+        results = self.table.insert(job_dict).run(self.conn)
+        if results['errors'] > 0:
+            raise ConflictingIdError(job.id)
+
+    def update_job(self, job):
+        changes = {
+            'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
+            'job_state': self.r.binary(pickle.dumps(job.__getstate__(), self.pickle_protocol))
+        }
+        results = self.table.get_all(job.id).update(changes).run(self.conn)
+        skipped = False in map(lambda x: results[x] == 0, results.keys())
+        if results['skipped'] > 0 or results['errors'] > 0 or not skipped:
+            raise JobLookupError(job.id)
+
+    def remove_job(self, job_id):
+        results = self.table.get_all(job_id).delete().run(self.conn)
+        if results['deleted'] + results['skipped'] != 1:
+            raise JobLookupError(job_id)
+
+    def remove_all_jobs(self):
+        self.table.delete().run(self.conn)
+
+    def shutdown(self):
+        self.conn.close()
+
+    def _reconstitute_job(self, job_state):
+        job_state = pickle.loads(job_state)
+        job = Job.__new__(Job)
+        job.__setstate__(job_state)
+        job._scheduler = self._scheduler
+        job._jobstore_alias = self._alias
+        return job
+
+    def _get_jobs(self, predicate=None):
+        jobs = []
+        failed_job_ids = []
+        query = (self.table.filter(self.r.row['next_run_time'] != None).filter(predicate)  # noqa
+                 if predicate else self.table)
+        query = query.order_by('next_run_time', 'id').pluck('id', 'job_state')
+
+        for document in query.run(self.conn):
+            try:
+                jobs.append(self._reconstitute_job(document['job_state']))
+            except Exception:
+                self._logger.exception('Unable to restore job "%s" -- removing it', document['id'])
+                failed_job_ids.append(document['id'])
+
+        # Remove all the jobs we failed to restore
+        if failed_job_ids:
+            self.r.expr(failed_job_ids).for_each(
+                lambda job_id: self.table.get_all(job_id).delete()).run(self.conn)
+
+        return jobs
+
+    def __repr__(self):
+        connection = self.conn
+        return '<%s (connection=%s)>' % (self.__class__.__name__, connection)

lib/apscheduler/jobstores/sqlalchemy.py

@@ -1,38 +1,47 @@
-
+from __future__ import absolute_import
 
 from apscheduler.jobstores.base import BaseJobStore, JobLookupError, ConflictingIdError
 from apscheduler.util import maybe_ref, datetime_to_utc_timestamp, utc_timestamp_to_datetime
 from apscheduler.job import Job
 
 try:
-    import pickle as pickle
+    import cPickle as pickle
 except ImportError:  # pragma: nocover
     import pickle
 
 try:
-    from sqlalchemy import create_engine, Table, Column, MetaData, Unicode, Float, LargeBinary, select
+    from sqlalchemy import (
+        create_engine, Table, Column, MetaData, Unicode, Float, LargeBinary, select, and_)
     from sqlalchemy.exc import IntegrityError
+    from sqlalchemy.sql.expression import null
 except ImportError:  # pragma: nocover
     raise ImportError('SQLAlchemyJobStore requires SQLAlchemy installed')
 
 
 class SQLAlchemyJobStore(BaseJobStore):
     """
-    Stores jobs in a database table using SQLAlchemy. The table will be created if it doesn't exist in the database.
+    Stores jobs in a database table using SQLAlchemy.
+    The table will be created if it doesn't exist in the database.
 
     Plugin alias: ``sqlalchemy``
 
-    :param str url: connection string (see `SQLAlchemy documentation
-                    <http://docs.sqlalchemy.org/en/latest/core/engines.html?highlight=create_engine#database-urls>`_
-                    on this)
-    :param engine: an SQLAlchemy Engine to use instead of creating a new one based on ``url``
+    :param str url: connection string (see
+        :ref:`SQLAlchemy documentation <sqlalchemy:database_urls>` on this)
+    :param engine: an SQLAlchemy :class:`~sqlalchemy.engine.Engine` to use instead of creating a
+        new one based on ``url``
     :param str tablename: name of the table to store jobs in
-    :param metadata: a :class:`~sqlalchemy.MetaData` instance to use instead of creating a new one
-    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the highest available
+    :param metadata: a :class:`~sqlalchemy.schema.MetaData` instance to use instead of creating a
+        new one
+    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the
+        highest available
+    :param str tableschema: name of the (existing) schema in the target database where the table
+        should be
+    :param dict engine_options: keyword arguments to :func:`~sqlalchemy.create_engine`
+        (ignored if ``engine`` is given)
     """
 
     def __init__(self, url=None, engine=None, tablename='apscheduler_jobs', metadata=None,
-                 pickle_protocol=pickle.HIGHEST_PROTOCOL):
+                 pickle_protocol=pickle.HIGHEST_PROTOCOL, tableschema=None, engine_options=None):
         super(SQLAlchemyJobStore, self).__init__()
         self.pickle_protocol = pickle_protocol
         metadata = maybe_ref(metadata) or MetaData()
@@ -40,37 +49,46 @@ class SQLAlchemyJobStore(BaseJobStore):
         if engine:
             self.engine = maybe_ref(engine)
         elif url:
-            self.engine = create_engine(url)
+            self.engine = create_engine(url, **(engine_options or {}))
         else:
             raise ValueError('Need either "engine" or "url" defined')
 
-        # 191 = max key length in MySQL for InnoDB/utf8mb4 tables, 25 = precision that translates to an 8-byte float
+        # 191 = max key length in MySQL for InnoDB/utf8mb4 tables,
+        # 25 = precision that translates to an 8-byte float
         self.jobs_t = Table(
             tablename, metadata,
-            Column('id', Unicode(191, _warn_on_bytestring=False), primary_key=True),
+            Column('id', Unicode(191), primary_key=True),
             Column('next_run_time', Float(25), index=True),
-            Column('job_state', LargeBinary, nullable=False)
+            Column('job_state', LargeBinary, nullable=False),
+            schema=tableschema
         )
 
+    def start(self, scheduler, alias):
+        super(SQLAlchemyJobStore, self).start(scheduler, alias)
         self.jobs_t.create(self.engine, True)
 
     def lookup_job(self, job_id):
-        selectable = select([self.jobs_t.c.job_state]).where(self.jobs_t.c.id == job_id)
-        job_state = self.engine.execute(selectable).scalar()
-        return self._reconstitute_job(job_state) if job_state else None
+        selectable = select(self.jobs_t.c.job_state).where(self.jobs_t.c.id == job_id)
+        with self.engine.begin() as connection:
+            job_state = connection.execute(selectable).scalar()
+            return self._reconstitute_job(job_state) if job_state else None
 
     def get_due_jobs(self, now):
         timestamp = datetime_to_utc_timestamp(now)
         return self._get_jobs(self.jobs_t.c.next_run_time <= timestamp)
 
     def get_next_run_time(self):
-        selectable = select([self.jobs_t.c.next_run_time]).where(self.jobs_t.c.next_run_time != None).\
+        selectable = select(self.jobs_t.c.next_run_time).\
+            where(self.jobs_t.c.next_run_time != null()).\
             order_by(self.jobs_t.c.next_run_time).limit(1)
-        next_run_time = self.engine.execute(selectable).scalar()
-        return utc_timestamp_to_datetime(next_run_time)
+        with self.engine.begin() as connection:
+            next_run_time = connection.execute(selectable).scalar()
+            return utc_timestamp_to_datetime(next_run_time)
 
     def get_all_jobs(self):
-        return self._get_jobs()
+        jobs = self._get_jobs()
+        self._fix_paused_jobs_sorting(jobs)
+        return jobs
 
     def add_job(self, job):
         insert = self.jobs_t.insert().values(**{
@@ -78,29 +96,33 @@ class SQLAlchemyJobStore(BaseJobStore):
             'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
             'job_state': pickle.dumps(job.__getstate__(), self.pickle_protocol)
         })
-        try:
-            self.engine.execute(insert)
-        except IntegrityError:
-            raise ConflictingIdError(job.id)
+        with self.engine.begin() as connection:
+            try:
+                connection.execute(insert)
+            except IntegrityError:
+                raise ConflictingIdError(job.id)
 
     def update_job(self, job):
         update = self.jobs_t.update().values(**{
             'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
             'job_state': pickle.dumps(job.__getstate__(), self.pickle_protocol)
         }).where(self.jobs_t.c.id == job.id)
-        result = self.engine.execute(update)
-        if result.rowcount == 0:
-            raise JobLookupError(id)
+        with self.engine.begin() as connection:
+            result = connection.execute(update)
+            if result.rowcount == 0:
+                raise JobLookupError(job.id)
 
     def remove_job(self, job_id):
         delete = self.jobs_t.delete().where(self.jobs_t.c.id == job_id)
-        result = self.engine.execute(delete)
-        if result.rowcount == 0:
-            raise JobLookupError(job_id)
+        with self.engine.begin() as connection:
+            result = connection.execute(delete)
+            if result.rowcount == 0:
+                raise JobLookupError(job_id)
 
     def remove_all_jobs(self):
         delete = self.jobs_t.delete()
-        self.engine.execute(delete)
+        with self.engine.begin() as connection:
+            connection.execute(delete)
 
     def shutdown(self):
         self.engine.dispose()
@@ -116,20 +138,22 @@ class SQLAlchemyJobStore(BaseJobStore):
 
     def _get_jobs(self, *conditions):
         jobs = []
-        selectable = select([self.jobs_t.c.id, self.jobs_t.c.job_state]).order_by(self.jobs_t.c.next_run_time)
-        selectable = selectable.where(*conditions) if conditions else selectable
+        selectable = select(self.jobs_t.c.id, self.jobs_t.c.job_state).\
+            order_by(self.jobs_t.c.next_run_time)
+        selectable = selectable.where(and_(*conditions)) if conditions else selectable
         failed_job_ids = set()
-        for row in self.engine.execute(selectable):
-            try:
-                jobs.append(self._reconstitute_job(row.job_state))
-            except:
-                self._logger.exception('Unable to restore job "%s" -- removing it', row.id)
-                failed_job_ids.add(row.id)
+        with self.engine.begin() as connection:
+            for row in connection.execute(selectable):
+                try:
+                    jobs.append(self._reconstitute_job(row.job_state))
+                except BaseException:
+                    self._logger.exception('Unable to restore job "%s" -- removing it', row.id)
+                    failed_job_ids.add(row.id)
 
-        # Remove all the jobs we failed to restore
-        if failed_job_ids:
-            delete = self.jobs_t.delete().where(self.jobs_t.c.id.in_(failed_job_ids))
-            self.engine.execute(delete)
+            # Remove all the jobs we failed to restore
+            if failed_job_ids:
+                delete = self.jobs_t.delete().where(self.jobs_t.c.id.in_(failed_job_ids))
+                connection.execute(delete)
 
         return jobs
 

lib/apscheduler/jobstores/zookeeper.py

@@ -0,0 +1,178 @@
+from __future__ import absolute_import
+
+from datetime import datetime
+
+from pytz import utc
+from kazoo.exceptions import NoNodeError, NodeExistsError
+
+from apscheduler.jobstores.base import BaseJobStore, JobLookupError, ConflictingIdError
+from apscheduler.util import maybe_ref, datetime_to_utc_timestamp, utc_timestamp_to_datetime
+from apscheduler.job import Job
+
+try:
+    import cPickle as pickle
+except ImportError:  # pragma: nocover
+    import pickle
+
+try:
+    from kazoo.client import KazooClient
+except ImportError:  # pragma: nocover
+    raise ImportError('ZooKeeperJobStore requires Kazoo installed')
+
+
+class ZooKeeperJobStore(BaseJobStore):
+    """
+    Stores jobs in a ZooKeeper tree. Any leftover keyword arguments are directly passed to
+    kazoo's `KazooClient
+    <http://kazoo.readthedocs.io/en/latest/api/client.html>`_.
+
+    Plugin alias: ``zookeeper``
+
+    :param str path: path to store jobs in
+    :param client: a :class:`~kazoo.client.KazooClient` instance to use instead of
+        providing connection arguments
+    :param int pickle_protocol: pickle protocol level to use (for serialization), defaults to the
+        highest available
+    """
+
+    def __init__(self, path='/apscheduler', client=None, close_connection_on_exit=False,
+                 pickle_protocol=pickle.HIGHEST_PROTOCOL, **connect_args):
+        super(ZooKeeperJobStore, self).__init__()
+        self.pickle_protocol = pickle_protocol
+        self.close_connection_on_exit = close_connection_on_exit
+
+        if not path:
+            raise ValueError('The "path" parameter must not be empty')
+
+        self.path = path
+
+        if client:
+            self.client = maybe_ref(client)
+        else:
+            self.client = KazooClient(**connect_args)
+        self._ensured_path = False
+
+    def _ensure_paths(self):
+        if not self._ensured_path:
+            self.client.ensure_path(self.path)
+        self._ensured_path = True
+
+    def start(self, scheduler, alias):
+        super(ZooKeeperJobStore, self).start(scheduler, alias)
+        if not self.client.connected:
+            self.client.start()
+
+    def lookup_job(self, job_id):
+        self._ensure_paths()
+        node_path = self.path + "/" + str(job_id)
+        try:
+            content, _ = self.client.get(node_path)
+            doc = pickle.loads(content)
+            job = self._reconstitute_job(doc['job_state'])
+            return job
+        except BaseException:
+            return None
+
+    def get_due_jobs(self, now):
+        timestamp = datetime_to_utc_timestamp(now)
+        jobs = [job_def['job'] for job_def in self._get_jobs()
+                if job_def['next_run_time'] is not None and job_def['next_run_time'] <= timestamp]
+        return jobs
+
+    def get_next_run_time(self):
+        next_runs = [job_def['next_run_time'] for job_def in self._get_jobs()
+                     if job_def['next_run_time'] is not None]
+        return utc_timestamp_to_datetime(min(next_runs)) if len(next_runs) > 0 else None
+
+    def get_all_jobs(self):
+        jobs = [job_def['job'] for job_def in self._get_jobs()]
+        self._fix_paused_jobs_sorting(jobs)
+        return jobs
+
+    def add_job(self, job):
+        self._ensure_paths()
+        node_path = self.path + "/" + str(job.id)
+        value = {
+            'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
+            'job_state': job.__getstate__()
+        }
+        data = pickle.dumps(value, self.pickle_protocol)
+        try:
+            self.client.create(node_path, value=data)
+        except NodeExistsError:
+            raise ConflictingIdError(job.id)
+
+    def update_job(self, job):
+        self._ensure_paths()
+        node_path = self.path + "/" + str(job.id)
+        changes = {
+            'next_run_time': datetime_to_utc_timestamp(job.next_run_time),
+            'job_state': job.__getstate__()
+        }
+        data = pickle.dumps(changes, self.pickle_protocol)
+        try:
+            self.client.set(node_path, value=data)
+        except NoNodeError:
+            raise JobLookupError(job.id)
+
+    def remove_job(self, job_id):
+        self._ensure_paths()
+        node_path = self.path + "/" + str(job_id)
+        try:
+            self.client.delete(node_path)
+        except NoNodeError:
+            raise JobLookupError(job_id)
+
+    def remove_all_jobs(self):
+        try:
+            self.client.delete(self.path, recursive=True)
+        except NoNodeError:
+            pass
+        self._ensured_path = False
+
+    def shutdown(self):
+        if self.close_connection_on_exit:
+            self.client.stop()
+            self.client.close()
+
+    def _reconstitute_job(self, job_state):
+        job_state = job_state
+        job = Job.__new__(Job)
+        job.__setstate__(job_state)
+        job._scheduler = self._scheduler
+        job._jobstore_alias = self._alias
+        return job
+
+    def _get_jobs(self):
+        self._ensure_paths()
+        jobs = []
+        failed_job_ids = []
+        all_ids = self.client.get_children(self.path)
+        for node_name in all_ids:
+            try:
+                node_path = self.path + "/" + node_name
+                content, _ = self.client.get(node_path)
+                doc = pickle.loads(content)
+                job_def = {
+                    'job_id': node_name,
+                    'next_run_time': doc['next_run_time'] if doc['next_run_time'] else None,
+                    'job_state': doc['job_state'],
+                    'job': self._reconstitute_job(doc['job_state']),
+                    'creation_time': _.ctime
+                }
+                jobs.append(job_def)
+            except BaseException:
+                self._logger.exception('Unable to restore job "%s" -- removing it' % node_name)
+                failed_job_ids.append(node_name)
+
+        # Remove all the jobs we failed to restore
+        if failed_job_ids:
+            for failed_id in failed_job_ids:
+                self.remove_job(failed_id)
+        paused_sort_key = datetime(9999, 12, 31, tzinfo=utc)
+        return sorted(jobs, key=lambda job_def: (job_def['job'].next_run_time or paused_sort_key,
+                                                 job_def['creation_time']))
+
+    def __repr__(self):
+        self._logger.exception('<%s (client=%s)>' % (self.__class__.__name__, self.client))
+        return '<%s (client=%s)>' % (self.__class__.__name__, self.client)

lib/apscheduler/schedulers/asyncio.py

@@ -1,22 +1,16 @@
-
-from functools import wraps
+from __future__ import absolute_import
+import asyncio
+from functools import wraps, partial
 
 from apscheduler.schedulers.base import BaseScheduler
 from apscheduler.util import maybe_ref
 
-try:
-    import asyncio
-except ImportError:  # pragma: nocover
-    try:
-        import trollius as asyncio
-    except ImportError:
-        raise ImportError('AsyncIOScheduler requires either Python 3.4 or the asyncio package installed')
-
 
 def run_in_event_loop(func):
     @wraps(func)
     def wrapper(self, *args, **kwargs):
-        self._eventloop.call_soon_threadsafe(func, self, *args, **kwargs)
+        wrapped = partial(func, self, *args, **kwargs)
+        self._eventloop.call_soon_threadsafe(wrapped)
     return wrapper
 
 
@@ -24,6 +18,8 @@ class AsyncIOScheduler(BaseScheduler):
     """
     A scheduler that runs on an asyncio (:pep:`3156`) event loop.
 
+    The default executor can run jobs based on native coroutines (``async def``).
+
     Extra options:
 
     ============== =============================================================
@@ -34,9 +30,11 @@ class AsyncIOScheduler(BaseScheduler):
     _eventloop = None
     _timeout = None
 
-    def start(self):
-        super(AsyncIOScheduler, self).start()
-        self.wakeup()
+    def start(self, paused=False):
+        if not self._eventloop:
+            self._eventloop = asyncio.get_event_loop()
+
+        super(AsyncIOScheduler, self).start(paused)
 
     @run_in_event_loop
     def shutdown(self, wait=True):
@@ -44,7 +42,7 @@ class AsyncIOScheduler(BaseScheduler):
         self._stop_timer()
 
     def _configure(self, config):
-        self._eventloop = maybe_ref(config.pop('event_loop', None)) or asyncio.get_event_loop()
+        self._eventloop = maybe_ref(config.pop('event_loop', None))
         super(AsyncIOScheduler, self)._configure(config)
 
     def _start_timer(self, wait_seconds):

lib/apscheduler/schedulers/background.py

@@ -1,3 +1,4 @@
+from __future__ import absolute_import
 
 from threading import Thread, Event
 
@@ -13,11 +14,12 @@ class BackgroundScheduler(BlockingScheduler):
 
     Extra options:
 
-    ========== ============================================================================================
-    ``daemon`` Set the ``daemon`` option in the background thread (defaults to ``True``,
-               see `the documentation <https://docs.python.org/3.4/library/threading.html#thread-objects>`_
+    ========== =============================================================================
+    ``daemon`` Set the ``daemon`` option in the background thread (defaults to ``True``, see
+               `the documentation
+               <https://docs.python.org/3.4/library/threading.html#thread-objects>`_
                for further details)
-    ========== ============================================================================================
+    ========== =============================================================================
     """
 
     _thread = None
@@ -26,14 +28,16 @@ class BackgroundScheduler(BlockingScheduler):
         self._daemon = asbool(config.pop('daemon', True))
         super(BackgroundScheduler, self)._configure(config)
 
-    def start(self):
-        BaseScheduler.start(self)
-        self._event = Event()
+    def start(self, *args, **kwargs):
+        if self._event is None or self._event.is_set():
+            self._event = Event()
+
+        BaseScheduler.start(self, *args, **kwargs)
         self._thread = Thread(target=self._main_loop, name='APScheduler')
         self._thread.daemon = self._daemon
         self._thread.start()
 
-    def shutdown(self, wait=True):
-        super(BackgroundScheduler, self).shutdown(wait)
+    def shutdown(self, *args, **kwargs):
+        super(BackgroundScheduler, self).shutdown(*args, **kwargs)
         self._thread.join()
         del self._thread

lib/apscheduler/schedulers/blocking.py

@@ -1,21 +1,23 @@
+from __future__ import absolute_import
 
 from threading import Event
 
-from apscheduler.schedulers.base import BaseScheduler
+from apscheduler.schedulers.base import BaseScheduler, STATE_STOPPED
+from apscheduler.util import TIMEOUT_MAX
 
 
 class BlockingScheduler(BaseScheduler):
     """
-    A scheduler that runs in the foreground (:meth:`~apscheduler.schedulers.base.BaseScheduler.start` will block).
+    A scheduler that runs in the foreground
+    (:meth:`~apscheduler.schedulers.base.BaseScheduler.start` will block).
     """
-
-    MAX_WAIT_TIME = 4294967  # Maximum value accepted by Event.wait() on Windows
-
     _event = None
 
-    def start(self):
-        super(BlockingScheduler, self).start()
-        self._event = Event()
+    def start(self, *args, **kwargs):
+        if self._event is None or self._event.is_set():
+            self._event = Event()
+
+        super(BlockingScheduler, self).start(*args, **kwargs)
         self._main_loop()
 
     def shutdown(self, wait=True):
@@ -23,10 +25,11 @@ class BlockingScheduler(BaseScheduler):
         self._event.set()
 
     def _main_loop(self):
-        while self.running:
-            wait_seconds = self._process_jobs()
-            self._event.wait(wait_seconds if wait_seconds is not None else self.MAX_WAIT_TIME)
+        wait_seconds = TIMEOUT_MAX
+        while self.state != STATE_STOPPED:
+            self._event.wait(wait_seconds)
             self._event.clear()
+            wait_seconds = self._process_jobs()
 
     def wakeup(self):
         self._event.set()

lib/apscheduler/schedulers/gevent.py

@@ -1,4 +1,4 @@
-
+from __future__ import absolute_import
 
 from apscheduler.schedulers.blocking import BlockingScheduler
 from apscheduler.schedulers.base import BaseScheduler
@@ -16,14 +16,14 @@ class GeventScheduler(BlockingScheduler):
 
     _greenlet = None
 
-    def start(self):
-        BaseScheduler.start(self)
+    def start(self, *args, **kwargs):
         self._event = Event()
+        BaseScheduler.start(self, *args, **kwargs)
         self._greenlet = gevent.spawn(self._main_loop)
         return self._greenlet
 
-    def shutdown(self, wait=True):
-        super(GeventScheduler, self).shutdown(wait)
+    def shutdown(self, *args, **kwargs):
+        super(GeventScheduler, self).shutdown(*args, **kwargs)
         self._greenlet.join()
         del self._greenlet
 

lib/apscheduler/schedulers/qt.py

@@ -1,17 +1,24 @@
-
+from __future__ import absolute_import
 
 from apscheduler.schedulers.base import BaseScheduler
 
 try:
     from PyQt5.QtCore import QObject, QTimer
-except ImportError:  # pragma: nocover
+except (ImportError, RuntimeError):  # pragma: nocover
     try:
         from PyQt4.QtCore import QObject, QTimer
     except ImportError:
         try:
-            from PySide.QtCore import QObject, QTimer  # flake8: noqa
+            from PySide6.QtCore import QObject, QTimer  # noqa
         except ImportError:
-            raise ImportError('QtScheduler requires either PyQt5, PyQt4 or PySide installed')
+            try:
+                from PySide2.QtCore import QObject, QTimer  # noqa
+            except ImportError:
+                try:
+                    from PySide.QtCore import QObject, QTimer  # noqa
+                except ImportError:
+                    raise ImportError('QtScheduler requires either PyQt5, PyQt4, PySide6, PySide2 '
+                                      'or PySide installed')
 
 
 class QtScheduler(BaseScheduler):
@@ -19,18 +26,15 @@ class QtScheduler(BaseScheduler):
 
     _timer = None
 
-    def start(self):
-        super(QtScheduler, self).start()
-        self.wakeup()
-
-    def shutdown(self, wait=True):
-        super(QtScheduler, self).shutdown(wait)
+    def shutdown(self, *args, **kwargs):
+        super(QtScheduler, self).shutdown(*args, **kwargs)
         self._stop_timer()
 
     def _start_timer(self, wait_seconds):
         self._stop_timer()
         if wait_seconds is not None:
-            self._timer = QTimer.singleShot(wait_seconds * 1000, self._process_jobs)
+            wait_time = min(int(wait_seconds * 1000), 2147483647)
+            self._timer = QTimer.singleShot(wait_time, self._process_jobs)
 
     def _stop_timer(self):
         if self._timer:

lib/apscheduler/schedulers/tornado.py

@@ -1,3 +1,4 @@
+from __future__ import absolute_import
 
 from datetime import timedelta
 from functools import wraps
@@ -22,6 +23,8 @@ class TornadoScheduler(BaseScheduler):
     """
     A scheduler that runs on a Tornado IOLoop.
 
+    The default executor can run jobs based on native coroutines (``async def``).
+
     =========== ===============================================================
     ``io_loop`` Tornado IOLoop instance to use (defaults to the global IO loop)
     =========== ===============================================================
@@ -30,10 +33,6 @@ class TornadoScheduler(BaseScheduler):
     _ioloop = None
     _timeout = None
 
-    def start(self):
-        super(TornadoScheduler, self).start()
-        self.wakeup()
-
     @run_in_ioloop
     def shutdown(self, wait=True):
         super(TornadoScheduler, self).shutdown(wait)
@@ -53,6 +52,10 @@ class TornadoScheduler(BaseScheduler):
             self._ioloop.remove_timeout(self._timeout)
             del self._timeout
 
+    def _create_default_executor(self):
+        from apscheduler.executors.tornado import TornadoExecutor
+        return TornadoExecutor()
+
     @run_in_ioloop
     def wakeup(self):
         self._stop_timer()

lib/apscheduler/schedulers/twisted.py

@@ -1,3 +1,4 @@
+from __future__ import absolute_import
 
 from functools import wraps
 
@@ -35,10 +36,6 @@ class TwistedScheduler(BaseScheduler):
         self._reactor = maybe_ref(config.pop('reactor', default_reactor))
         super(TwistedScheduler, self)._configure(config)
 
-    def start(self):
-        super(TwistedScheduler, self).start()
-        self.wakeup()
-
     @run_in_reactor
     def shutdown(self, wait=True):
         super(TwistedScheduler, self).shutdown(wait)

lib/apscheduler/triggers/base.py

@@ -1,4 +1,6 @@
 from abc import ABCMeta, abstractmethod
+from datetime import timedelta
+import random
 
 import six
 
@@ -6,11 +8,30 @@ import six
 class BaseTrigger(six.with_metaclass(ABCMeta)):
     """Abstract base class that defines the interface that every trigger must implement."""
 
+    __slots__ = ()
+
     @abstractmethod
     def get_next_fire_time(self, previous_fire_time, now):
         """
-        Returns the next datetime to fire on, If no such datetime can be calculated, returns ``None``.
+        Returns the next datetime to fire on, If no such datetime can be calculated, returns
+        ``None``.
 
         :param datetime.datetime previous_fire_time: the previous time the trigger was fired
         :param datetime.datetime now: current datetime
         """
+
+    def _apply_jitter(self, next_fire_time, jitter, now):
+        """
+        Randomize ``next_fire_time`` by adding a random value (the jitter).
+
+        :param datetime.datetime|None next_fire_time: next fire time without jitter applied. If
+            ``None``, returns ``None``.
+        :param int|None jitter: maximum number of seconds to add to ``next_fire_time``
+            (if ``None`` or ``0``, returns ``next_fire_time``)
+        :param datetime.datetime now: current datetime
+        :return datetime.datetime|None: next fire time with a jitter.
+        """
+        if next_fire_time is None or not jitter:
+            return next_fire_time
+
+        return next_fire_time + timedelta(seconds=random.uniform(0, jitter))

lib/apscheduler/triggers/combining.py

@@ -0,0 +1,95 @@
+from apscheduler.triggers.base import BaseTrigger
+from apscheduler.util import obj_to_ref, ref_to_obj
+
+
+class BaseCombiningTrigger(BaseTrigger):
+    __slots__ = ('triggers', 'jitter')
+
+    def __init__(self, triggers, jitter=None):
+        self.triggers = triggers
+        self.jitter = jitter
+
+    def __getstate__(self):
+        return {
+            'version': 1,
+            'triggers': [(obj_to_ref(trigger.__class__), trigger.__getstate__())
+                         for trigger in self.triggers],
+            'jitter': self.jitter
+        }
+
+    def __setstate__(self, state):
+        if state.get('version', 1) > 1:
+            raise ValueError(
+                'Got serialized data for version %s of %s, but only versions up to 1 can be '
+                'handled' % (state['version'], self.__class__.__name__))
+
+        self.jitter = state['jitter']
+        self.triggers = []
+        for clsref, state in state['triggers']:
+            cls = ref_to_obj(clsref)
+            trigger = cls.__new__(cls)
+            trigger.__setstate__(state)
+            self.triggers.append(trigger)
+
+    def __repr__(self):
+        return '<{}({}{})>'.format(self.__class__.__name__, self.triggers,
+                                   ', jitter={}'.format(self.jitter) if self.jitter else '')
+
+
+class AndTrigger(BaseCombiningTrigger):
+    """
+    Always returns the earliest next fire time that all the given triggers can agree on.
+    The trigger is considered to be finished when any of the given triggers has finished its
+    schedule.
+
+    Trigger alias: ``and``
+
+    :param list triggers: triggers to combine
+    :param int|None jitter: delay the job execution by ``jitter`` seconds at most
+    """
+
+    __slots__ = ()
+
+    def get_next_fire_time(self, previous_fire_time, now):
+        while True:
+            fire_times = [trigger.get_next_fire_time(previous_fire_time, now)
+                          for trigger in self.triggers]
+            if None in fire_times:
+                return None
+            elif min(fire_times) == max(fire_times):
+                return self._apply_jitter(fire_times[0], self.jitter, now)
+            else:
+                now = max(fire_times)
+
+    def __str__(self):
+        return 'and[{}]'.format(', '.join(str(trigger) for trigger in self.triggers))
+
+
+class OrTrigger(BaseCombiningTrigger):
+    """
+    Always returns the earliest next fire time produced by any of the given triggers.
+    The trigger is considered finished when all the given triggers have finished their schedules.
+
+    Trigger alias: ``or``
+
+    :param list triggers: triggers to combine
+    :param int|None jitter: delay the job execution by ``jitter`` seconds at most
+
+    .. note:: Triggers that depends on the previous fire time, such as the interval trigger, may
+        seem to behave strangely since they are always passed the previous fire time produced by
+        any of the given triggers.
+    """
+
+    __slots__ = ()
+
+    def get_next_fire_time(self, previous_fire_time, now):
+        fire_times = [trigger.get_next_fire_time(previous_fire_time, now)
+                      for trigger in self.triggers]
+        fire_times = [fire_time for fire_time in fire_times if fire_time is not None]
+        if fire_times:
+            return self._apply_jitter(min(fire_times), self.jitter, now)
+        else:
+            return None
+
+    def __str__(self):
+        return 'or[{}]'.format(', '.join(str(trigger) for trigger in self.triggers))

lib/apscheduler/triggers/cron/__init__.py

@@ -4,17 +4,20 @@ from tzlocal import get_localzone
 import six
 
 from apscheduler.triggers.base import BaseTrigger
-from apscheduler.triggers.cron.fields import BaseField, WeekField, DayOfMonthField, DayOfWeekField, DEFAULT_VALUES
-from apscheduler.util import datetime_ceil, convert_to_datetime, datetime_repr, astimezone
+from apscheduler.triggers.cron.fields import (
+    BaseField, MonthField, WeekField, DayOfMonthField, DayOfWeekField, DEFAULT_VALUES)
+from apscheduler.util import (
+    datetime_ceil, convert_to_datetime, datetime_repr, astimezone, localize, normalize)
 
 
 class CronTrigger(BaseTrigger):
     """
-    Triggers when current time matches all specified time constraints, similarly to how the UNIX cron scheduler works.
+    Triggers when current time matches all specified time constraints,
+    similarly to how the UNIX cron scheduler works.
 
     :param int|str year: 4-digit year
     :param int|str month: month (1-12)
-    :param int|str day: day of the (1-31)
+    :param int|str day: day of month (1-31)
     :param int|str week: ISO week (1-53)
     :param int|str day_of_week: number or name of weekday (0-6 or mon,tue,wed,thu,fri,sat,sun)
     :param int|str hour: hour (0-23)
@@ -22,8 +25,9 @@ class CronTrigger(BaseTrigger):
     :param int|str second: second (0-59)
     :param datetime|str start_date: earliest possible date/time to trigger on (inclusive)
     :param datetime|str end_date: latest possible date/time to trigger on (inclusive)
-    :param datetime.tzinfo|str timezone: time zone to use for the date/time calculations
-                                         (defaults to scheduler timezone)
+    :param datetime.tzinfo|str timezone: time zone to use for the date/time calculations (defaults
+        to scheduler timezone)
+    :param int|None jitter: delay the job execution by ``jitter`` seconds at most
 
     .. note:: The first weekday is always **monday**.
     """
@@ -31,7 +35,7 @@ class CronTrigger(BaseTrigger):
     FIELD_NAMES = ('year', 'month', 'day', 'week', 'day_of_week', 'hour', 'minute', 'second')
     FIELDS_MAP = {
         'year': BaseField,
-        'month': BaseField,
+        'month': MonthField,
         'week': WeekField,
         'day': DayOfMonthField,
         'day_of_week': DayOfWeekField,
@@ -40,15 +44,16 @@ class CronTrigger(BaseTrigger):
         'second': BaseField
     }
 
-    __slots__ = 'timezone', 'start_date', 'end_date', 'fields'
+    __slots__ = 'timezone', 'start_date', 'end_date', 'fields', 'jitter'
 
-    def __init__(self, year=None, month=None, day=None, week=None, day_of_week=None, hour=None, minute=None,
-                 second=None, start_date=None, end_date=None, timezone=None):
+    def __init__(self, year=None, month=None, day=None, week=None, day_of_week=None, hour=None,
+                 minute=None, second=None, start_date=None, end_date=None, timezone=None,
+                 jitter=None):
         if timezone:
             self.timezone = astimezone(timezone)
-        elif start_date and start_date.tzinfo:
+        elif isinstance(start_date, datetime) and start_date.tzinfo:
             self.timezone = start_date.tzinfo
-        elif end_date and end_date.tzinfo:
+        elif isinstance(end_date, datetime) and end_date.tzinfo:
             self.timezone = end_date.tzinfo
         else:
             self.timezone = get_localzone()
@@ -56,6 +61,8 @@ class CronTrigger(BaseTrigger):
         self.start_date = convert_to_datetime(start_date, self.timezone, 'start_date')
         self.end_date = convert_to_datetime(end_date, self.timezone, 'end_date')
 
+        self.jitter = jitter
+
         values = dict((key, value) for (key, value) in six.iteritems(locals())
                       if key in self.FIELD_NAMES and value is not None)
         self.fields = []
@@ -76,13 +83,35 @@ class CronTrigger(BaseTrigger):
             field = field_class(field_name, exprs, is_default)
             self.fields.append(field)
 
+    @classmethod
+    def from_crontab(cls, expr, timezone=None):
+        """
+        Create a :class:`~CronTrigger` from a standard crontab expression.
+
+        See https://en.wikipedia.org/wiki/Cron for more information on the format accepted here.
+
+        :param expr: minute, hour, day of month, month, day of week
+        :param datetime.tzinfo|str timezone: time zone to use for the date/time calculations (
+            defaults to scheduler timezone)
+        :return: a :class:`~CronTrigger` instance
+
+        """
+        values = expr.split()
+        if len(values) != 5:
+            raise ValueError('Wrong number of fields; got {}, expected 5'.format(len(values)))
+
+        return cls(minute=values[0], hour=values[1], day=values[2], month=values[3],
+                   day_of_week=values[4], timezone=timezone)
+
     def _increment_field_value(self, dateval, fieldnum):
         """
-        Increments the designated field and resets all less significant fields to their minimum values.
+        Increments the designated field and resets all less significant fields to their minimum
+        values.
 
         :type dateval: datetime
         :type fieldnum: int
-        :return: a tuple containing the new date, and the number of the field that was actually incremented
+        :return: a tuple containing the new date, and the number of the field that was actually
+            incremented
         :rtype: tuple
         """
 
@@ -115,7 +144,7 @@ class CronTrigger(BaseTrigger):
                     i += 1
 
         difference = datetime(**values) - dateval.replace(tzinfo=None)
-        return self.timezone.normalize(dateval + difference), fieldnum
+        return normalize(dateval + difference), fieldnum
 
     def _set_field_value(self, dateval, fieldnum, new_value):
         values = {}
@@ -128,12 +157,13 @@ class CronTrigger(BaseTrigger):
                 else:
                     values[field.name] = new_value
 
-        difference = datetime(**values) - dateval.replace(tzinfo=None)
-        return self.timezone.normalize(dateval + difference)
+        return localize(datetime(**values), self.timezone)
 
     def get_next_fire_time(self, previous_fire_time, now):
         if previous_fire_time:
-            start_date = max(now, previous_fire_time + timedelta(microseconds=1))
+            start_date = min(now, previous_fire_time + timedelta(microseconds=1))
+            if start_date == previous_fire_time:
+                start_date += timedelta(microseconds=1)
         else:
             start_date = max(now, self.start_date) if self.start_date else now
 
@@ -163,7 +193,34 @@ class CronTrigger(BaseTrigger):
                 return None
 
         if fieldnum >= 0:
-            return next_date
+            next_date = self._apply_jitter(next_date, self.jitter, now)
+            return min(next_date, self.end_date) if self.end_date else next_date
+
+    def __getstate__(self):
+        return {
+            'version': 2,
+            'timezone': self.timezone,
+            'start_date': self.start_date,
+            'end_date': self.end_date,
+            'fields': self.fields,
+            'jitter': self.jitter,
+        }
+
+    def __setstate__(self, state):
+        # This is for compatibility with APScheduler 3.0.x
+        if isinstance(state, tuple):
+            state = state[1]
+
+        if state.get('version', 1) > 2:
+            raise ValueError(
+                'Got serialized data for version %s of %s, but only versions up to 2 can be '
+                'handled' % (state['version'], self.__class__.__name__))
+
+        self.timezone = state['timezone']
+        self.start_date = state['start_date']
+        self.end_date = state['end_date']
+        self.fields = state['fields']
+        self.jitter = state.get('jitter')
 
     def __str__(self):
         options = ["%s='%s'" % (f.name, f) for f in self.fields if not f.is_default]
@@ -172,5 +229,11 @@ class CronTrigger(BaseTrigger):
     def __repr__(self):
         options = ["%s='%s'" % (f.name, f) for f in self.fields if not f.is_default]
         if self.start_date:
-            options.append("start_date='%s'" % datetime_repr(self.start_date))
-        return '<%s (%s)>' % (self.__class__.__name__, ', '.join(options))
+            options.append("start_date=%r" % datetime_repr(self.start_date))
+        if self.end_date:
+            options.append("end_date=%r" % datetime_repr(self.end_date))
+        if self.jitter:
+            options.append('jitter=%s' % self.jitter)
+
+        return "<%s (%s, timezone='%s')>" % (
+            self.__class__.__name__, ', '.join(options), self.timezone)

lib/apscheduler/triggers/cron/expressions.py

@@ -1,17 +1,16 @@
-"""
-This module contains the expressions applicable for CronTrigger's fields.
-"""
+"""This module contains the expressions applicable for CronTrigger's fields."""
 
 from calendar import monthrange
 import re
 
 from apscheduler.util import asint
 
-__all__ = ('AllExpression', 'RangeExpression', 'WeekdayRangeExpression', 'WeekdayPositionExpression',
-           'LastDayOfMonthExpression')
+__all__ = ('AllExpression', 'RangeExpression', 'WeekdayRangeExpression',
+           'WeekdayPositionExpression', 'LastDayOfMonthExpression')
 
 
 WEEKDAYS = ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun']
+MONTHS = ['jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug', 'sep', 'oct', 'nov', 'dec']
 
 
 class AllExpression(object):
@@ -22,6 +21,14 @@ class AllExpression(object):
         if self.step == 0:
             raise ValueError('Increment must be higher than 0')
 
+    def validate_range(self, field_name):
+        from apscheduler.triggers.cron.fields import MIN_VALUES, MAX_VALUES
+
+        value_range = MAX_VALUES[field_name] - MIN_VALUES[field_name]
+        if self.step and self.step > value_range:
+            raise ValueError('the step value ({}) is higher than the total range of the '
+                             'expression ({})'.format(self.step, value_range))
+
     def get_next_value(self, date, field):
         start = field.get_value(date)
         minval = field.get_min(date)
@@ -37,6 +44,9 @@ class AllExpression(object):
         if next <= maxval:
             return next
 
+    def __eq__(self, other):
+        return isinstance(other, self.__class__) and self.step == other.step
+
     def __str__(self):
         if self.step:
             return '*/%d' % self.step
@@ -51,7 +61,7 @@ class RangeExpression(AllExpression):
         r'(?P<first>\d+)(?:-(?P<last>\d+))?(?:/(?P<step>\d+))?$')
 
     def __init__(self, first, last=None, step=None):
-        AllExpression.__init__(self, step)
+        super(RangeExpression, self).__init__(step)
         first = asint(first)
         last = asint(last)
         if last is None and step is None:
@@ -61,25 +71,41 @@ class RangeExpression(AllExpression):
         self.first = first
         self.last = last
 
+    def validate_range(self, field_name):
+        from apscheduler.triggers.cron.fields import MIN_VALUES, MAX_VALUES
+
+        super(RangeExpression, self).validate_range(field_name)
+        if self.first < MIN_VALUES[field_name]:
+            raise ValueError('the first value ({}) is lower than the minimum value ({})'
+                             .format(self.first, MIN_VALUES[field_name]))
+        if self.last is not None and self.last > MAX_VALUES[field_name]:
+            raise ValueError('the last value ({}) is higher than the maximum value ({})'
+                             .format(self.last, MAX_VALUES[field_name]))
+        value_range = (self.last or MAX_VALUES[field_name]) - self.first
+        if self.step and self.step > value_range:
+            raise ValueError('the step value ({}) is higher than the total range of the '
+                             'expression ({})'.format(self.step, value_range))
+
     def get_next_value(self, date, field):
-        start = field.get_value(date)
+        startval = field.get_value(date)
         minval = field.get_min(date)
         maxval = field.get_max(date)
 
         # Apply range limits
         minval = max(minval, self.first)
-        if self.last is not None:
-            maxval = min(maxval, self.last)
-        start = max(start, minval)
+        maxval = min(maxval, self.last) if self.last is not None else maxval
+        nextval = max(minval, startval)
 
-        if not self.step:
-            next = start
-        else:
-            distance_to_next = (self.step - (start - minval)) % self.step
-            next = start + distance_to_next
+        # Apply the step if defined
+        if self.step:
+            distance_to_next = (self.step - (nextval - minval)) % self.step
+            nextval += distance_to_next
 
-        if next <= maxval:
-            return next
+        return nextval if nextval <= maxval else None
+
+    def __eq__(self, other):
+        return (isinstance(other, self.__class__) and self.first == other.first and
+                self.last == other.last)
 
     def __str__(self):
         if self.last != self.first and self.last is not None:
@@ -100,6 +126,37 @@ class RangeExpression(AllExpression):
         return "%s(%s)" % (self.__class__.__name__, ', '.join(args))
 
 
+class MonthRangeExpression(RangeExpression):
+    value_re = re.compile(r'(?P<first>[a-z]+)(?:-(?P<last>[a-z]+))?', re.IGNORECASE)
+
+    def __init__(self, first, last=None):
+        try:
+            first_num = MONTHS.index(first.lower()) + 1
+        except ValueError:
+            raise ValueError('Invalid month name "%s"' % first)
+
+        if last:
+            try:
+                last_num = MONTHS.index(last.lower()) + 1
+            except ValueError:
+                raise ValueError('Invalid month name "%s"' % last)
+        else:
+            last_num = None
+
+        super(MonthRangeExpression, self).__init__(first_num, last_num)
+
+    def __str__(self):
+        if self.last != self.first and self.last is not None:
+            return '%s-%s' % (MONTHS[self.first - 1], MONTHS[self.last - 1])
+        return MONTHS[self.first - 1]
+
+    def __repr__(self):
+        args = ["'%s'" % MONTHS[self.first]]
+        if self.last != self.first and self.last is not None:
+            args.append("'%s'" % MONTHS[self.last - 1])
+        return "%s(%s)" % (self.__class__.__name__, ', '.join(args))
+
+
 class WeekdayRangeExpression(RangeExpression):
     value_re = re.compile(r'(?P<first>[a-z]+)(?:-(?P<last>[a-z]+))?', re.IGNORECASE)
 
@@ -117,7 +174,7 @@ class WeekdayRangeExpression(RangeExpression):
         else:
             last_num = None
 
-        RangeExpression.__init__(self, first_num, last_num)
+        super(WeekdayRangeExpression, self).__init__(first_num, last_num)
 
     def __str__(self):
         if self.last != self.first and self.last is not None:
@@ -133,9 +190,11 @@ class WeekdayRangeExpression(RangeExpression):
 
 class WeekdayPositionExpression(AllExpression):
     options = ['1st', '2nd', '3rd', '4th', '5th', 'last']
-    value_re = re.compile(r'(?P<option_name>%s) +(?P<weekday_name>(?:\d+|\w+))' % '|'.join(options), re.IGNORECASE)
+    value_re = re.compile(r'(?P<option_name>%s) +(?P<weekday_name>(?:\d+|\w+))' %
+                          '|'.join(options), re.IGNORECASE)
 
     def __init__(self, option_name, weekday_name):
+        super(WeekdayPositionExpression, self).__init__(None)
         try:
             self.option_num = self.options.index(option_name.lower())
         except ValueError:
@@ -147,8 +206,7 @@ class WeekdayPositionExpression(AllExpression):
             raise ValueError('Invalid weekday name "%s"' % weekday_name)
 
     def get_next_value(self, date, field):
-        # Figure out the weekday of the month's first day and the number
-        # of days in that month
+        # Figure out the weekday of the month's first day and the number of days in that month
         first_day_wday, last_day = monthrange(date.year, date.month)
 
         # Calculate which day of the month is the first of the target weekdays
@@ -160,23 +218,28 @@ class WeekdayPositionExpression(AllExpression):
         if self.option_num < 5:
             target_day = first_hit_day + self.option_num * 7
         else:
-            target_day = first_hit_day + ((last_day - first_hit_day) / 7) * 7
+            target_day = first_hit_day + ((last_day - first_hit_day) // 7) * 7
 
         if target_day <= last_day and target_day >= date.day:
             return target_day
 
+    def __eq__(self, other):
+        return (super(WeekdayPositionExpression, self).__eq__(other) and
+                self.option_num == other.option_num and self.weekday == other.weekday)
+
     def __str__(self):
         return '%s %s' % (self.options[self.option_num], WEEKDAYS[self.weekday])
 
     def __repr__(self):
-        return "%s('%s', '%s')" % (self.__class__.__name__, self.options[self.option_num], WEEKDAYS[self.weekday])
+        return "%s('%s', '%s')" % (self.__class__.__name__, self.options[self.option_num],
+                                   WEEKDAYS[self.weekday])
 
 
 class LastDayOfMonthExpression(AllExpression):
     value_re = re.compile(r'last', re.IGNORECASE)
 
     def __init__(self):
-        pass
+        super(LastDayOfMonthExpression, self).__init__(None)
 
     def get_next_value(self, date, field):
         return monthrange(date.year, date.month)[1]

lib/apscheduler/triggers/cron/fields.py

@@ -1,22 +1,26 @@
-"""
-Fields represent CronTrigger options which map to :class:`~datetime.datetime`
-fields.
-"""
+"""Fields represent CronTrigger options which map to :class:`~datetime.datetime` fields."""
 
 from calendar import monthrange
+import re
+
+import six
 
 from apscheduler.triggers.cron.expressions import (
-    AllExpression, RangeExpression, WeekdayPositionExpression, LastDayOfMonthExpression, WeekdayRangeExpression)
+    AllExpression, RangeExpression, WeekdayPositionExpression, LastDayOfMonthExpression,
+    WeekdayRangeExpression, MonthRangeExpression)
 
 
-__all__ = ('MIN_VALUES', 'MAX_VALUES', 'DEFAULT_VALUES', 'BaseField', 'WeekField', 'DayOfMonthField', 'DayOfWeekField')
+__all__ = ('MIN_VALUES', 'MAX_VALUES', 'DEFAULT_VALUES', 'BaseField', 'WeekField',
+           'DayOfMonthField', 'DayOfWeekField')
 
 
-MIN_VALUES = {'year': 1970, 'month': 1, 'day': 1, 'week': 1, 'day_of_week': 0, 'hour': 0, 'minute': 0, 'second': 0}
-MAX_VALUES = {'year': 2 ** 63, 'month': 12, 'day:': 31, 'week': 53, 'day_of_week': 6, 'hour': 23, 'minute': 59,
-              'second': 59}
-DEFAULT_VALUES = {'year': '*', 'month': 1, 'day': 1, 'week': '*', 'day_of_week': '*', 'hour': 0, 'minute': 0,
-                  'second': 0}
+MIN_VALUES = {'year': 1970, 'month': 1, 'day': 1, 'week': 1, 'day_of_week': 0, 'hour': 0,
+              'minute': 0, 'second': 0}
+MAX_VALUES = {'year': 9999, 'month': 12, 'day': 31, 'week': 53, 'day_of_week': 6, 'hour': 23,
+              'minute': 59, 'second': 59}
+DEFAULT_VALUES = {'year': '*', 'month': 1, 'day': 1, 'week': '*', 'day_of_week': '*', 'hour': 0,
+                  'minute': 0, 'second': 0}
+SEPARATOR = re.compile(' *, *')
 
 
 class BaseField(object):
@@ -50,23 +54,29 @@ class BaseField(object):
         self.expressions = []
 
         # Split a comma-separated expression list, if any
-        exprs = str(exprs).strip()
-        if ',' in exprs:
-            for expr in exprs.split(','):
-                self.compile_expression(expr)
-        else:
-            self.compile_expression(exprs)
+        for expr in SEPARATOR.split(str(exprs).strip()):
+            self.compile_expression(expr)
 
     def compile_expression(self, expr):
         for compiler in self.COMPILERS:
             match = compiler.value_re.match(expr)
             if match:
                 compiled_expr = compiler(**match.groupdict())
+
+                try:
+                    compiled_expr.validate_range(self.name)
+                except ValueError as e:
+                    exc = ValueError('Error validating expression {!r}: {}'.format(expr, e))
+                    six.raise_from(exc, None)
+
                 self.expressions.append(compiled_expr)
                 return
 
         raise ValueError('Unrecognized expression "%s" for field "%s"' % (expr, self.name))
 
+    def __eq__(self, other):
+        return isinstance(self, self.__class__) and self.expressions == other.expressions
+
     def __str__(self):
         expr_strings = (str(e) for e in self.expressions)
         return ','.join(expr_strings)
@@ -95,3 +105,7 @@ class DayOfWeekField(BaseField):
 
     def get_value(self, dateval):
         return dateval.weekday()
+
+
+class MonthField(BaseField):
+    COMPILERS = BaseField.COMPILERS + [MonthRangeExpression]

lib/apscheduler/triggers/date.py

@@ -14,15 +14,36 @@ class DateTrigger(BaseTrigger):
     :param datetime.tzinfo|str timezone: time zone for ``run_date`` if it doesn't have one already
     """
 
-    __slots__ = 'timezone', 'run_date'
+    __slots__ = 'run_date'
 
     def __init__(self, run_date=None, timezone=None):
         timezone = astimezone(timezone) or get_localzone()
-        self.run_date = convert_to_datetime(run_date or datetime.now(), timezone, 'run_date')
+        if run_date is not None:
+            self.run_date = convert_to_datetime(run_date, timezone, 'run_date')
+        else:
+            self.run_date = datetime.now(timezone)
 
     def get_next_fire_time(self, previous_fire_time, now):
         return self.run_date if previous_fire_time is None else None
 
+    def __getstate__(self):
+        return {
+            'version': 1,
+            'run_date': self.run_date
+        }
+
+    def __setstate__(self, state):
+        # This is for compatibility with APScheduler 3.0.x
+        if isinstance(state, tuple):
+            state = state[1]
+
+        if state.get('version', 1) > 1:
+            raise ValueError(
+                'Got serialized data for version %s of %s, but only version 1 can be handled' %
+                (state['version'], self.__class__.__name__))
+
+        self.run_date = state['run_date']
+
     def __str__(self):
         return 'date[%s]' % datetime_repr(self.run_date)
 

lib/apscheduler/triggers/interval.py

@@ -4,13 +4,15 @@ from math import ceil
 from tzlocal import get_localzone
 
 from apscheduler.triggers.base import BaseTrigger
-from apscheduler.util import convert_to_datetime, timedelta_seconds, datetime_repr, astimezone
+from apscheduler.util import (
+    convert_to_datetime, normalize, timedelta_seconds, datetime_repr,
+    astimezone)
 
 
 class IntervalTrigger(BaseTrigger):
     """
-    Triggers on specified intervals, starting on ``start_date`` if specified, ``datetime.now()`` + interval
-    otherwise.
+    Triggers on specified intervals, starting on ``start_date`` if specified, ``datetime.now()`` +
+    interval otherwise.
 
     :param int weeks: number of weeks to wait
     :param int days: number of days to wait
@@ -20,12 +22,15 @@ class IntervalTrigger(BaseTrigger):
     :param datetime|str start_date: starting point for the interval calculation
     :param datetime|str end_date: latest possible date/time to trigger on
     :param datetime.tzinfo|str timezone: time zone to use for the date/time calculations
+    :param int|None jitter: delay the job execution by ``jitter`` seconds at most
     """
 
-    __slots__ = 'timezone', 'start_date', 'end_date', 'interval'
+    __slots__ = 'timezone', 'start_date', 'end_date', 'interval', 'interval_length', 'jitter'
 
-    def __init__(self, weeks=0, days=0, hours=0, minutes=0, seconds=0, start_date=None, end_date=None, timezone=None):
-        self.interval = timedelta(weeks=weeks, days=days, hours=hours, minutes=minutes, seconds=seconds)
+    def __init__(self, weeks=0, days=0, hours=0, minutes=0, seconds=0, start_date=None,
+                 end_date=None, timezone=None, jitter=None):
+        self.interval = timedelta(weeks=weeks, days=days, hours=hours, minutes=minutes,
+                                  seconds=seconds)
         self.interval_length = timedelta_seconds(self.interval)
         if self.interval_length == 0:
             self.interval = timedelta(seconds=1)
@@ -33,9 +38,9 @@ class IntervalTrigger(BaseTrigger):
 
         if timezone:
             self.timezone = astimezone(timezone)
-        elif start_date and start_date.tzinfo:
+        elif isinstance(start_date, datetime) and start_date.tzinfo:
             self.timezone = start_date.tzinfo
-        elif end_date and end_date.tzinfo:
+        elif isinstance(end_date, datetime) and end_date.tzinfo:
             self.timezone = end_date.tzinfo
         else:
             self.timezone = get_localzone()
@@ -44,6 +49,8 @@ class IntervalTrigger(BaseTrigger):
         self.start_date = convert_to_datetime(start_date, self.timezone, 'start_date')
         self.end_date = convert_to_datetime(end_date, self.timezone, 'end_date')
 
+        self.jitter = jitter
+
     def get_next_fire_time(self, previous_fire_time, now):
         if previous_fire_time:
             next_fire_time = previous_fire_time + self.interval
@@ -54,12 +61,48 @@ class IntervalTrigger(BaseTrigger):
             next_interval_num = int(ceil(timediff_seconds / self.interval_length))
             next_fire_time = self.start_date + self.interval * next_interval_num
 
+        if self.jitter is not None:
+            next_fire_time = self._apply_jitter(next_fire_time, self.jitter, now)
+
         if not self.end_date or next_fire_time <= self.end_date:
-            return self.timezone.normalize(next_fire_time)
+            return normalize(next_fire_time)
+
+    def __getstate__(self):
+        return {
+            'version': 2,
+            'timezone': self.timezone,
+            'start_date': self.start_date,
+            'end_date': self.end_date,
+            'interval': self.interval,
+            'jitter': self.jitter,
+        }
+
+    def __setstate__(self, state):
+        # This is for compatibility with APScheduler 3.0.x
+        if isinstance(state, tuple):
+            state = state[1]
+
+        if state.get('version', 1) > 2:
+            raise ValueError(
+                'Got serialized data for version %s of %s, but only versions up to 2 can be '
+                'handled' % (state['version'], self.__class__.__name__))
+
+        self.timezone = state['timezone']
+        self.start_date = state['start_date']
+        self.end_date = state['end_date']
+        self.interval = state['interval']
+        self.interval_length = timedelta_seconds(self.interval)
+        self.jitter = state.get('jitter')
 
     def __str__(self):
         return 'interval[%s]' % str(self.interval)
 
     def __repr__(self):
-        return "<%s (interval=%r, start_date='%s')>" % (self.__class__.__name__, self.interval,
-                                                        datetime_repr(self.start_date))
+        options = ['interval=%r' % self.interval, 'start_date=%r' % datetime_repr(self.start_date)]
+        if self.end_date:
+            options.append("end_date=%r" % datetime_repr(self.end_date))
+        if self.jitter:
+            options.append('jitter=%s' % self.jitter)
+
+        return "<%s (%s, timezone='%s')>" % (
+            self.__class__.__name__, ', '.join(options), self.timezone)

lib/apscheduler/util.py

@@ -1,29 +1,36 @@
 """This module contains several handy functions primarily meant for internal use."""
 
+from __future__ import division
 
+from asyncio import iscoroutinefunction
 from datetime import date, datetime, time, timedelta, tzinfo
-from inspect import isfunction, ismethod, getargspec
 from calendar import timegm
+from functools import partial
+from inspect import isclass, ismethod
 import re
+import sys
 
-from pytz import timezone, utc
+from pytz import timezone, utc, FixedOffset
 import six
 
 try:
     from inspect import signature
 except ImportError:  # pragma: nocover
-    try:
-        from funcsigs import signature
-    except ImportError:
-        signature = None
+    from funcsigs import signature
+
+try:
+    from threading import TIMEOUT_MAX
+except ImportError:
+    TIMEOUT_MAX = 4294967  # Maximum value accepted by Event.wait() on Windows
 
 __all__ = ('asint', 'asbool', 'astimezone', 'convert_to_datetime', 'datetime_to_utc_timestamp',
-           'utc_timestamp_to_datetime', 'timedelta_seconds', 'datetime_ceil', 'get_callable_name', 'obj_to_ref',
-           'ref_to_obj', 'maybe_ref', 'repr_escape', 'check_callable_args')
+           'utc_timestamp_to_datetime', 'timedelta_seconds', 'datetime_ceil', 'get_callable_name',
+           'obj_to_ref', 'ref_to_obj', 'maybe_ref', 'repr_escape', 'check_callable_args',
+           'normalize', 'localize', 'TIMEOUT_MAX')
 
 
 class _Undefined(object):
-    def __bool__(self):
+    def __nonzero__(self):
         return False
 
     def __bool__(self):
@@ -32,17 +39,18 @@ class _Undefined(object):
     def __repr__(self):
         return '<undefined>'
 
+
 undefined = _Undefined()  #: a unique object that only signifies that no value is defined
 
 
 def asint(text):
     """
-    Safely converts a string to an integer, returning None if the string is None.
+    Safely converts a string to an integer, returning ``None`` if the string is ``None``.
 
     :type text: str
     :rtype: int
-    """
 
+    """
     if text is not None:
         return int(text)
 
@@ -52,8 +60,8 @@ def asbool(obj):
     Interprets an object as a boolean value.
 
     :rtype: bool
-    """
 
+    """
     if isinstance(obj, str):
         obj = obj.strip().lower()
         if obj in ('true', 'yes', 'on', 'y', 't', '1'):
@@ -69,15 +77,17 @@ def astimezone(obj):
     Interprets an object as a timezone.
 
     :rtype: tzinfo
-    """
 
+    """
     if isinstance(obj, six.string_types):
         return timezone(obj)
     if isinstance(obj, tzinfo):
-        if not hasattr(obj, 'localize') or not hasattr(obj, 'normalize'):
-            raise TypeError('Only timezones from the pytz library are supported')
-        if obj.zone == 'local':
-            raise ValueError('Unable to determine the name of the local timezone -- use an explicit timezone instead')
+        if obj.tzname(None) == 'local':
+            raise ValueError(
+                'Unable to determine the name of the local timezone -- you must explicitly '
+                'specify the name of the local timezone. Please refrain from using timezones like '
+                'EST to prevent problems with daylight saving time. Instead, use a locale based '
+                'timezone name (such as Europe/Helsinki).')
         return obj
     if obj is not None:
         raise TypeError('Expected tzinfo, got %s instead' % obj.__class__.__name__)
@@ -85,27 +95,30 @@ def astimezone(obj):
 
 _DATE_REGEX = re.compile(
     r'(?P<year>\d{4})-(?P<month>\d{1,2})-(?P<day>\d{1,2})'
-    r'(?: (?P<hour>\d{1,2}):(?P<minute>\d{1,2}):(?P<second>\d{1,2})'
-    r'(?:\.(?P<microsecond>\d{1,6}))?)?')
+    r'(?:[ T](?P<hour>\d{1,2}):(?P<minute>\d{1,2}):(?P<second>\d{1,2})'
+    r'(?:\.(?P<microsecond>\d{1,6}))?'
+    r'(?P<timezone>Z|[+-]\d\d:\d\d)?)?$')
 
 
 def convert_to_datetime(input, tz, arg_name):
     """
     Converts the given object to a timezone aware datetime object.
+
     If a timezone aware datetime object is passed, it is returned unmodified.
     If a native datetime object is passed, it is given the specified timezone.
     If the input is a string, it is parsed as a datetime with the given timezone.
 
-    Date strings are accepted in three different forms: date only (Y-m-d),
-    date with time (Y-m-d H:M:S) or with date+time with microseconds
-    (Y-m-d H:M:S.micro).
+    Date strings are accepted in three different forms: date only (Y-m-d), date with time
+    (Y-m-d H:M:S) or with date+time with microseconds (Y-m-d H:M:S.micro). Additionally you can
+    override the time zone by giving a specific offset in the format specified by ISO 8601:
+    Z (UTC), +HH:MM or -HH:MM.
 
     :param str|datetime input: the datetime or string to convert to a timezone aware datetime
     :param datetime.tzinfo tz: timezone to interpret ``input`` in
     :param str arg_name: the name of the argument (used in an error message)
     :rtype: datetime
-    """
 
+    """
     if input is None:
         return
     elif isinstance(input, datetime):
@@ -116,8 +129,17 @@ def convert_to_datetime(input, tz, arg_name):
         m = _DATE_REGEX.match(input)
         if not m:
             raise ValueError('Invalid date string')
-        values = [(k, int(v or 0)) for k, v in list(m.groupdict().items())]
-        values = dict(values)
+
+        values = m.groupdict()
+        tzname = values.pop('timezone')
+        if tzname == 'Z':
+            tz = utc
+        elif tzname:
+            hours, minutes = (int(x) for x in tzname[1:].split(':'))
+            sign = 1 if tzname[0] == '+' else -1
+            tz = FixedOffset(sign * (hours * 60 + minutes))
+
+        values = {k: int(v or 0) for k, v in values.items()}
         datetime_ = datetime(**values)
     else:
         raise TypeError('Unsupported type for %s: %s' % (arg_name, input.__class__.__name__))
@@ -125,14 +147,12 @@ def convert_to_datetime(input, tz, arg_name):
     if datetime_.tzinfo is not None:
         return datetime_
     if tz is None:
-        raise ValueError('The "tz" argument must be specified if %s has no timezone information' % arg_name)
+        raise ValueError(
+            'The "tz" argument must be specified if %s has no timezone information' % arg_name)
     if isinstance(tz, six.string_types):
         tz = timezone(tz)
 
-    try:
-        return tz.localize(datetime_, is_dst=None)
-    except AttributeError:
-        raise TypeError('Only pytz timezones are supported (need the localize() and normalize() methods)')
+    return localize(datetime_, tz)
 
 
 def datetime_to_utc_timestamp(timeval):
@@ -141,8 +161,8 @@ def datetime_to_utc_timestamp(timeval):
 
     :type timeval: datetime
     :rtype: float
-    """
 
+    """
     if timeval is not None:
         return timegm(timeval.utctimetuple()) + timeval.microsecond / 1000000
 
@@ -153,8 +173,8 @@ def utc_timestamp_to_datetime(timestamp):
 
     :type timestamp: float
     :rtype: datetime
-    """
 
+    """
     if timestamp is not None:
         return datetime.fromtimestamp(timestamp, utc)
 
@@ -165,8 +185,8 @@ def timedelta_seconds(delta):
 
     :type delta: timedelta
     :rtype: float
-    """
 
+    """
     return delta.days * 24 * 60 * 60 + delta.seconds + \
         delta.microseconds / 1000000.0
 
@@ -176,8 +196,8 @@ def datetime_ceil(dateval):
     Rounds the given datetime object upwards.
 
     :type dateval: datetime
-    """
 
+    """
     if dateval.microsecond > 0:
         return dateval + timedelta(seconds=1, microseconds=-dateval.microsecond)
     return dateval
@@ -192,8 +212,8 @@ def get_callable_name(func):
     Returns the best available display name for the given function/callable.
 
     :rtype: str
-    """
 
+    """
     # the easy case (on Python 3.3+)
     if hasattr(func, '__qualname__'):
         return func.__qualname__
@@ -201,7 +221,7 @@ def get_callable_name(func):
     # class methods, bound and unbound methods
     f_self = getattr(func, '__self__', None) or getattr(func, 'im_self', None)
     if f_self and hasattr(func, '__name__'):
-        f_class = f_self if isinstance(f_self, type) else f_self.__class__
+        f_class = f_self if isclass(f_self) else f_self.__class__
     else:
         f_class = getattr(func, 'im_class', None)
 
@@ -222,20 +242,35 @@ def get_callable_name(func):
 
 def obj_to_ref(obj):
     """
-    Returns the path to the given object.
+    Returns the path to the given callable.
 
     :rtype: str
+    :raises TypeError: if the given object is not callable
+    :raises ValueError: if the given object is a :class:`~functools.partial`, lambda or a nested
+        function
+
     """
+    if isinstance(obj, partial):
+        raise ValueError('Cannot create a reference to a partial()')
 
-    try:
-        ref = '%s:%s' % (obj.__module__, get_callable_name(obj))
-        obj2 = ref_to_obj(ref)
-        if obj != obj2:
-            raise ValueError
-    except Exception:
-        raise ValueError('Cannot determine the reference to %r' % obj)
+    name = get_callable_name(obj)
+    if '<lambda>' in name:
+        raise ValueError('Cannot create a reference to a lambda')
+    if '<locals>' in name:
+        raise ValueError('Cannot create a reference to a nested function')
 
-    return ref
+    if ismethod(obj):
+        if hasattr(obj, 'im_self') and obj.im_self:
+            # bound method
+            module = obj.im_self.__module__
+        elif hasattr(obj, 'im_class') and obj.im_class:
+            # unbound method
+            module = obj.im_class.__module__
+        else:
+            module = obj.__module__
+    else:
+        module = obj.__module__
+    return '%s:%s' % (module, name)
 
 
 def ref_to_obj(ref):
@@ -243,8 +278,8 @@ def ref_to_obj(ref):
     Returns the object pointed to by ``ref``.
 
     :type ref: str
-    """
 
+    """
     if not isinstance(ref, six.string_types):
         raise TypeError('References must be strings')
     if ':' not in ref:
@@ -252,12 +287,12 @@ def ref_to_obj(ref):
 
     modulename, rest = ref.split(':', 1)
     try:
-        obj = __import__(modulename)
+        obj = __import__(modulename, fromlist=[rest])
     except ImportError:
         raise LookupError('Error resolving reference %s: could not import module' % ref)
 
     try:
-        for name in modulename.split('.')[1:] + rest.split('.'):
+        for name in rest.split('.'):
             obj = getattr(obj, name)
         return obj
     except Exception:
@@ -268,8 +303,8 @@ def maybe_ref(ref):
     """
     Returns the object that the given reference points to, if it is indeed a reference.
     If it is not a reference, the object is returned as-is.
-    """
 
+    """
     if not isinstance(ref, str):
         return ref
     return ref_to_obj(ref)
@@ -281,7 +316,8 @@ if six.PY2:
             return string.encode('ascii', 'backslashreplace')
         return string
 else:
-    repr_escape = lambda string: string
+    def repr_escape(string):
+        return string
 
 
 def check_callable_args(func, args, kwargs):
@@ -290,70 +326,54 @@ def check_callable_args(func, args, kwargs):
 
     :type args: tuple
     :type kwargs: dict
-    """
 
+    """
     pos_kwargs_conflicts = []  # parameters that have a match in both args and kwargs
     positional_only_kwargs = []  # positional-only parameters that have a match in kwargs
     unsatisfied_args = []  # parameters in signature that don't have a match in args or kwargs
     unsatisfied_kwargs = []  # keyword-only arguments that don't have a match in kwargs
     unmatched_args = list(args)  # args that didn't match any of the parameters in the signature
-    unmatched_kwargs = list(kwargs)  # kwargs that didn't match any of the parameters in the signature
-    has_varargs = has_var_kwargs = False  # indicates if the signature defines *args and **kwargs respectively
+    # kwargs that didn't match any of the parameters in the signature
+    unmatched_kwargs = list(kwargs)
+    # indicates if the signature defines *args and **kwargs respectively
+    has_varargs = has_var_kwargs = False
 
-    if signature:
-        try:
+    try:
+        if sys.version_info >= (3, 5):
+            sig = signature(func, follow_wrapped=False)
+        else:
             sig = signature(func)
-        except ValueError:
-            return  # signature() doesn't work against every kind of callable
+    except ValueError:
+        # signature() doesn't work against every kind of callable
+        return
 
-        for param in six.itervalues(sig.parameters):
-            if param.kind == param.POSITIONAL_OR_KEYWORD:
-                if param.name in unmatched_kwargs and unmatched_args:
-                    pos_kwargs_conflicts.append(param.name)
-                elif unmatched_args:
-                    del unmatched_args[0]
-                elif param.name in unmatched_kwargs:
-                    unmatched_kwargs.remove(param.name)
-                elif param.default is param.empty:
-                    unsatisfied_args.append(param.name)
-            elif param.kind == param.POSITIONAL_ONLY:
-                if unmatched_args:
-                    del unmatched_args[0]
-                elif param.name in unmatched_kwargs:
-                    unmatched_kwargs.remove(param.name)
-                    positional_only_kwargs.append(param.name)
-                elif param.default is param.empty:
-                    unsatisfied_args.append(param.name)
-            elif param.kind == param.KEYWORD_ONLY:
-                if param.name in unmatched_kwargs:
-                    unmatched_kwargs.remove(param.name)
-                elif param.default is param.empty:
-                    unsatisfied_kwargs.append(param.name)
-            elif param.kind == param.VAR_POSITIONAL:
-                has_varargs = True
-            elif param.kind == param.VAR_KEYWORD:
-                has_var_kwargs = True
-    else:
-        if not isfunction(func) and not ismethod(func) and hasattr(func, '__call__'):
-            func = func.__call__
-
-        try:
-            argspec = getargspec(func)
-        except TypeError:
-            return  # getargspec() doesn't work certain callables
-
-        argspec_args = argspec.args if not ismethod(func) else argspec.args[1:]
-        has_varargs = bool(argspec.varargs)
-        has_var_kwargs = bool(argspec.keywords)
-        for arg, default in six.moves.zip_longest(argspec_args, argspec.defaults or (), fillvalue=undefined):
-            if arg in unmatched_kwargs and unmatched_args:
-                pos_kwargs_conflicts.append(arg)
+    for param in six.itervalues(sig.parameters):
+        if param.kind == param.POSITIONAL_OR_KEYWORD:
+            if param.name in unmatched_kwargs and unmatched_args:
+                pos_kwargs_conflicts.append(param.name)
             elif unmatched_args:
                 del unmatched_args[0]
-            elif arg in unmatched_kwargs:
-                unmatched_kwargs.remove(arg)
-            elif default is undefined:
-                unsatisfied_args.append(arg)
+            elif param.name in unmatched_kwargs:
+                unmatched_kwargs.remove(param.name)
+            elif param.default is param.empty:
+                unsatisfied_args.append(param.name)
+        elif param.kind == param.POSITIONAL_ONLY:
+            if unmatched_args:
+                del unmatched_args[0]
+            elif param.name in unmatched_kwargs:
+                unmatched_kwargs.remove(param.name)
+                positional_only_kwargs.append(param.name)
+            elif param.default is param.empty:
+                unsatisfied_args.append(param.name)
+        elif param.kind == param.KEYWORD_ONLY:
+            if param.name in unmatched_kwargs:
+                unmatched_kwargs.remove(param.name)
+            elif param.default is param.empty:
+                unsatisfied_kwargs.append(param.name)
+        elif param.kind == param.VAR_POSITIONAL:
+            has_varargs = True
+        elif param.kind == param.VAR_KEYWORD:
+            has_var_kwargs = True
 
     # Make sure there are no conflicts between args and kwargs
     if pos_kwargs_conflicts:
@@ -365,21 +385,46 @@ def check_callable_args(func, args, kwargs):
         raise ValueError('The following arguments cannot be given as keyword arguments: %s' %
                          ', '.join(positional_only_kwargs))
 
-    # Check that the number of positional arguments minus the number of matched kwargs matches the argspec
+    # Check that the number of positional arguments minus the number of matched kwargs matches the
+    # argspec
     if unsatisfied_args:
-        raise ValueError('The following arguments have not been supplied: %s' % ', '.join(unsatisfied_args))
+        raise ValueError('The following arguments have not been supplied: %s' %
+                         ', '.join(unsatisfied_args))
 
     # Check that all keyword-only arguments have been supplied
     if unsatisfied_kwargs:
-        raise ValueError('The following keyword-only arguments have not been supplied in kwargs: %s' %
-                         ', '.join(unsatisfied_kwargs))
+        raise ValueError(
+            'The following keyword-only arguments have not been supplied in kwargs: %s' %
+            ', '.join(unsatisfied_kwargs))
 
     # Check that the callable can accept the given number of positional arguments
     if not has_varargs and unmatched_args:
-        raise ValueError('The list of positional arguments is longer than the target callable can handle '
-                         '(allowed: %d, given in args: %d)' % (len(args) - len(unmatched_args), len(args)))
+        raise ValueError(
+            'The list of positional arguments is longer than the target callable can handle '
+            '(allowed: %d, given in args: %d)' % (len(args) - len(unmatched_args), len(args)))
 
     # Check that the callable can accept the given keyword arguments
     if not has_var_kwargs and unmatched_kwargs:
-        raise ValueError('The target callable does not accept the following keyword arguments: %s' %
-                         ', '.join(unmatched_kwargs))
+        raise ValueError(
+            'The target callable does not accept the following keyword arguments: %s' %
+            ', '.join(unmatched_kwargs))
+
+
+def iscoroutinefunction_partial(f):
+    while isinstance(f, partial):
+        f = f.func
+
+    # The asyncio version of iscoroutinefunction includes testing for @coroutine
+    # decorations vs. the inspect version which does not.
+    return iscoroutinefunction(f)
+
+
+def normalize(dt):
+    return datetime.fromtimestamp(dt.timestamp(), dt.tzinfo)
+
+
+def localize(dt, tzinfo):
+    if hasattr(tzinfo, 'localize'):
+        return tzinfo.localize(dt)
+
+    return normalize(dt.replace(tzinfo=tzinfo))

lib/biplist/__init__.py

@@ -1,803 +0,0 @@
-"""biplist -- a library for reading and writing binary property list files.
-
-Binary Property List (plist) files provide a faster and smaller serialization
-format for property lists on OS X. This is a library for generating binary
-plists which can be read by OS X, iOS, or other clients.
-
-The API models the plistlib API, and will call through to plistlib when
-XML serialization or deserialization is required.
-
-To generate plists with UID values, wrap the values with the Uid object. The
-value must be an int.
-
-To generate plists with NSData/CFData values, wrap the values with the
-Data object. The value must be a string.
-
-Date values can only be datetime.datetime objects.
-
-The exceptions InvalidPlistException and NotBinaryPlistException may be 
-thrown to indicate that the data cannot be serialized or deserialized as
-a binary plist.
-
-Plist generation example:
-    
-    from biplist import *
-    from datetime import datetime
-    plist = {'aKey':'aValue',
-             '0':1.322,
-             'now':datetime.now(),
-             'list':[1,2,3],
-             'tuple':('a','b','c')
-             }
-    try:
-        writePlist(plist, "example.plist")
-    except (InvalidPlistException, NotBinaryPlistException), e:
-        print "Something bad happened:", e
-
-Plist parsing example:
-
-    from biplist import *
-    try:
-        plist = readPlist("example.plist")
-        print plist
-    except (InvalidPlistException, NotBinaryPlistException), e:
-        print "Not a plist:", e
-"""
-
-import sys
-from collections import namedtuple
-import datetime
-import io
-import math
-import plistlib
-from struct import pack, unpack
-from struct import error as struct_error
-import sys
-import time
-
-try:
-    str
-    unicodeEmpty = r''
-except NameError:
-    str = str
-    unicodeEmpty = ''
-try:
-    int
-except NameError:
-    long = int
-try:
-    {}.iteritems
-    iteritems = lambda x: iter(x.items())
-except AttributeError:
-    iteritems = lambda x: list(x.items())
-
-__all__ = [
-    'Uid', 'Data', 'readPlist', 'writePlist', 'readPlistFromString',
-    'writePlistToString', 'InvalidPlistException', 'NotBinaryPlistException'
-]
-
-# Apple uses Jan 1, 2001 as a base for all plist date/times.
-apple_reference_date = datetime.datetime.utcfromtimestamp(978307200)
-
-class Uid(int):
-    """Wrapper around integers for representing UID values. This
-       is used in keyed archiving."""
-    def __repr__(self):
-        return "Uid(%d)" % self
-
-class Data(bytes):
-    """Wrapper around str types for representing Data values."""
-    pass
-
-class InvalidPlistException(Exception):
-    """Raised when the plist is incorrectly formatted."""
-    pass
-
-class NotBinaryPlistException(Exception):
-    """Raised when a binary plist was expected but not encountered."""
-    pass
-
-def readPlist(pathOrFile):
-    """Raises NotBinaryPlistException, InvalidPlistException"""
-    didOpen = False
-    result = None
-    if isinstance(pathOrFile, (bytes, str)):
-        pathOrFile = open(pathOrFile, 'rb')
-        didOpen = True
-    try:
-        reader = PlistReader(pathOrFile)
-        result = reader.parse()
-    except NotBinaryPlistException as e:
-        try:
-            pathOrFile.seek(0)
-            result = None
-            if hasattr(plistlib, 'loads'):
-                contents = None
-                if isinstance(pathOrFile, (bytes, str)):
-                    with open(pathOrFile, 'rb') as f:
-                        contents = f.read()
-                else:
-                    contents = pathOrFile.read()
-                result = plistlib.loads(contents)
-            else:
-                result = plistlib.readPlist(pathOrFile)
-            result = wrapDataObject(result, for_binary=True)
-        except Exception as e:
-            raise InvalidPlistException(e)
-    finally:
-        if didOpen:
-            pathOrFile.close()
-    return result
-
-def wrapDataObject(o, for_binary=False):
-    if isinstance(o, Data) and not for_binary:
-        v = sys.version_info
-        if not (v[0] >= 3 and v[1] >= 4):
-            o = plistlib.Data(o)
-    elif isinstance(o, (bytes, plistlib.Data)) and for_binary:
-        if hasattr(o, 'data'):
-            o = Data(o.data)
-    elif isinstance(o, tuple):
-        o = wrapDataObject(list(o), for_binary)
-        o = tuple(o)
-    elif isinstance(o, list):
-        for i in range(len(o)):
-            o[i] = wrapDataObject(o[i], for_binary)
-    elif isinstance(o, dict):
-        for k in o:
-            o[k] = wrapDataObject(o[k], for_binary)
-    return o
-
-def writePlist(rootObject, pathOrFile, binary=True):
-    if not binary:
-        rootObject = wrapDataObject(rootObject, binary)
-        if hasattr(plistlib, "dump"):
-            if isinstance(pathOrFile, (bytes, str)):
-                with open(pathOrFile, 'wb') as f:
-                    return plistlib.dump(rootObject, f)
-            else:
-                return plistlib.dump(rootObject, pathOrFile)
-        else:
-            return plistlib.writePlist(rootObject, pathOrFile)
-    else:
-        didOpen = False
-        if isinstance(pathOrFile, (bytes, str)):
-            pathOrFile = open(pathOrFile, 'wb')
-            didOpen = True
-        writer = PlistWriter(pathOrFile)
-        result = writer.writeRoot(rootObject)
-        if didOpen:
-            pathOrFile.close()
-        return result
-
-def readPlistFromString(data):
-    return readPlist(io.BytesIO(data))
-
-def writePlistToString(rootObject, binary=True):
-    if not binary:
-        rootObject = wrapDataObject(rootObject, binary)
-        if hasattr(plistlib, "dumps"):
-            return plistlib.dumps(rootObject)
-        elif hasattr(plistlib, "writePlistToBytes"):
-            return plistlib.writePlistToBytes(rootObject)
-        else:
-            return plistlib.writePlistToString(rootObject)
-    else:
-        ioObject = io.BytesIO()
-        writer = PlistWriter(ioObject)
-        writer.writeRoot(rootObject)
-        return ioObject.getvalue()
-
-def is_stream_binary_plist(stream):
-    stream.seek(0)
-    header = stream.read(7)
-    if header == b'bplist0':
-        return True
-    else:
-        return False
-
-PlistTrailer = namedtuple('PlistTrailer', 'offsetSize, objectRefSize, offsetCount, topLevelObjectNumber, offsetTableOffset')
-PlistByteCounts = namedtuple('PlistByteCounts', 'nullBytes, boolBytes, intBytes, realBytes, dateBytes, dataBytes, stringBytes, uidBytes, arrayBytes, setBytes, dictBytes')
-
-class PlistReader(object):
-    file = None
-    contents = ''
-    offsets = None
-    trailer = None
-    currentOffset = 0
-    
-    def __init__(self, fileOrStream):
-        """Raises NotBinaryPlistException."""
-        self.reset()
-        self.file = fileOrStream
-    
-    def parse(self):
-        return self.readRoot()
-    
-    def reset(self):
-        self.trailer = None
-        self.contents = ''
-        self.offsets = []
-        self.currentOffset = 0
-    
-    def readRoot(self):
-        result = None
-        self.reset()
-        # Get the header, make sure it's a valid file.
-        if not is_stream_binary_plist(self.file):
-            raise NotBinaryPlistException()
-        self.file.seek(0)
-        self.contents = self.file.read()
-        if len(self.contents) < 32:
-            raise InvalidPlistException("File is too short.")
-        trailerContents = self.contents[-32:]
-        try:
-            self.trailer = PlistTrailer._make(unpack("!xxxxxxBBQQQ", trailerContents))
-            offset_size = self.trailer.offsetSize * self.trailer.offsetCount
-            offset = self.trailer.offsetTableOffset
-            offset_contents = self.contents[offset:offset+offset_size]
-            offset_i = 0
-            while offset_i < self.trailer.offsetCount:
-                begin = self.trailer.offsetSize*offset_i
-                tmp_contents = offset_contents[begin:begin+self.trailer.offsetSize]
-                tmp_sized = self.getSizedInteger(tmp_contents, self.trailer.offsetSize)
-                self.offsets.append(tmp_sized)
-                offset_i += 1
-            self.setCurrentOffsetToObjectNumber(self.trailer.topLevelObjectNumber)
-            result = self.readObject()
-        except TypeError as e:
-            raise InvalidPlistException(e)
-        return result
-    
-    def setCurrentOffsetToObjectNumber(self, objectNumber):
-        self.currentOffset = self.offsets[objectNumber]
-    
-    def readObject(self):
-        result = None
-        tmp_byte = self.contents[self.currentOffset:self.currentOffset+1]
-        marker_byte = unpack("!B", tmp_byte)[0]
-        format = (marker_byte >> 4) & 0x0f
-        extra = marker_byte & 0x0f
-        self.currentOffset += 1
-        
-        def proc_extra(extra):
-            if extra == 0b1111:
-                #self.currentOffset += 1
-                extra = self.readObject()
-            return extra
-        
-        # bool, null, or fill byte
-        if format == 0b0000:
-            if extra == 0b0000:
-                result = None
-            elif extra == 0b1000:
-                result = False
-            elif extra == 0b1001:
-                result = True
-            elif extra == 0b1111:
-                pass # fill byte
-            else:
-                raise InvalidPlistException("Invalid object found at offset: %d" % (self.currentOffset - 1))
-        # int
-        elif format == 0b0001:
-            extra = proc_extra(extra)
-            result = self.readInteger(pow(2, extra))
-        # real
-        elif format == 0b0010:
-            extra = proc_extra(extra)
-            result = self.readReal(extra)
-        # date
-        elif format == 0b0011 and extra == 0b0011:
-            result = self.readDate()
-        # data
-        elif format == 0b0100:
-            extra = proc_extra(extra)
-            result = self.readData(extra)
-        # ascii string
-        elif format == 0b0101:
-            extra = proc_extra(extra)
-            result = self.readAsciiString(extra)
-        # Unicode string
-        elif format == 0b0110:
-            extra = proc_extra(extra)
-            result = self.readUnicode(extra)
-        # uid
-        elif format == 0b1000:
-            result = self.readUid(extra)
-        # array
-        elif format == 0b1010:
-            extra = proc_extra(extra)
-            result = self.readArray(extra)
-        # set
-        elif format == 0b1100:
-            extra = proc_extra(extra)
-            result = set(self.readArray(extra))
-        # dict
-        elif format == 0b1101:
-            extra = proc_extra(extra)
-            result = self.readDict(extra)
-        else:    
-            raise InvalidPlistException("Invalid object found: {format: %s, extra: %s}" % (bin(format), bin(extra)))
-        return result
-    
-    def readInteger(self, byteSize):
-        result = 0
-        original_offset = self.currentOffset
-        data = self.contents[self.currentOffset:self.currentOffset + byteSize]
-        result = self.getSizedInteger(data, byteSize, as_number=True)
-        self.currentOffset = original_offset + byteSize
-        return result
-    
-    def readReal(self, length):
-        result = 0.0
-        to_read = pow(2, length)
-        data = self.contents[self.currentOffset:self.currentOffset+to_read]
-        if length == 2: # 4 bytes
-            result = unpack('>f', data)[0]
-        elif length == 3: # 8 bytes
-            result = unpack('>d', data)[0]
-        else:
-            raise InvalidPlistException("Unknown real of length %d bytes" % to_read)
-        return result
-    
-    def readRefs(self, count):    
-        refs = []
-        i = 0
-        while i < count:
-            fragment = self.contents[self.currentOffset:self.currentOffset+self.trailer.objectRefSize]
-            ref = self.getSizedInteger(fragment, len(fragment))
-            refs.append(ref)
-            self.currentOffset += self.trailer.objectRefSize
-            i += 1
-        return refs
-    
-    def readArray(self, count):
-        result = []
-        values = self.readRefs(count)
-        i = 0
-        while i < len(values):
-            self.setCurrentOffsetToObjectNumber(values[i])
-            value = self.readObject()
-            result.append(value)
-            i += 1
-        return result
-    
-    def readDict(self, count):
-        result = {}
-        keys = self.readRefs(count)
-        values = self.readRefs(count)
-        i = 0
-        while i < len(keys):
-            self.setCurrentOffsetToObjectNumber(keys[i])
-            key = self.readObject()
-            self.setCurrentOffsetToObjectNumber(values[i])
-            value = self.readObject()
-            result[key] = value
-            i += 1
-        return result
-    
-    def readAsciiString(self, length):
-        result = unpack("!%ds" % length, self.contents[self.currentOffset:self.currentOffset+length])[0]
-        self.currentOffset += length
-        return result
-    
-    def readUnicode(self, length):
-        actual_length = length*2
-        data = self.contents[self.currentOffset:self.currentOffset+actual_length]
-        # unpack not needed?!! data = unpack(">%ds" % (actual_length), data)[0]
-        self.currentOffset += actual_length
-        return data.decode('utf_16_be')
-    
-    def readDate(self):
-        result = unpack(">d", self.contents[self.currentOffset:self.currentOffset+8])[0]
-        # Use timedelta to workaround time_t size limitation on 32-bit python.
-        result = datetime.timedelta(seconds=result) + apple_reference_date
-        self.currentOffset += 8
-        return result
-    
-    def readData(self, length):
-        result = self.contents[self.currentOffset:self.currentOffset+length]
-        self.currentOffset += length
-        return Data(result)
-    
-    def readUid(self, length):
-        return Uid(self.readInteger(length+1))
-    
-    def getSizedInteger(self, data, byteSize, as_number=False):
-        """Numbers of 8 bytes are signed integers when they refer to numbers, but unsigned otherwise."""
-        result = 0
-        # 1, 2, and 4 byte integers are unsigned
-        if byteSize == 1:
-            result = unpack('>B', data)[0]
-        elif byteSize == 2:
-            result = unpack('>H', data)[0]
-        elif byteSize == 4:
-            result = unpack('>L', data)[0]
-        elif byteSize == 8:
-            if as_number:
-                result = unpack('>q', data)[0]
-            else:
-                result = unpack('>Q', data)[0]
-        elif byteSize <= 16:
-            # Handle odd-sized or integers larger than 8 bytes
-            # Don't naively go over 16 bytes, in order to prevent infinite loops.
-            result = 0
-            if hasattr(int, 'from_bytes'):
-                result = int.from_bytes(data, 'big')
-            else:
-                for byte in data:
-                    result = (result << 8) | unpack('>B', byte)[0]
-        else:
-            raise InvalidPlistException("Encountered integer longer than 16 bytes.")
-        return result
-
-class HashableWrapper(object):
-    def __init__(self, value):
-        self.value = value
-    def __repr__(self):
-        return "<HashableWrapper: %s>" % [self.value]
-
-class BoolWrapper(object):
-    def __init__(self, value):
-        self.value = value
-    def __repr__(self):
-        return "<BoolWrapper: %s>" % self.value
-
-class FloatWrapper(object):
-    _instances = {}
-    def __new__(klass, value):
-        # Ensure FloatWrapper(x) for a given float x is always the same object
-        wrapper = klass._instances.get(value)
-        if wrapper is None:
-            wrapper = object.__new__(klass)
-            wrapper.value = value
-            klass._instances[value] = wrapper
-        return wrapper
-    def __repr__(self):
-        return "<FloatWrapper: %s>" % self.value
-
-class PlistWriter(object):
-    header = b'bplist00bybiplist1.0'
-    file = None
-    byteCounts = None
-    trailer = None
-    computedUniques = None
-    writtenReferences = None
-    referencePositions = None
-    wrappedTrue = None
-    wrappedFalse = None
-    
-    def __init__(self, file):
-        self.reset()
-        self.file = file
-        self.wrappedTrue = BoolWrapper(True)
-        self.wrappedFalse = BoolWrapper(False)
-
-    def reset(self):
-        self.byteCounts = PlistByteCounts(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
-        self.trailer = PlistTrailer(0, 0, 0, 0, 0)
-        
-        # A set of all the uniques which have been computed.
-        self.computedUniques = set()
-        # A list of all the uniques which have been written.
-        self.writtenReferences = {}
-        # A dict of the positions of the written uniques.
-        self.referencePositions = {}
-        
-    def positionOfObjectReference(self, obj):
-        """If the given object has been written already, return its
-           position in the offset table. Otherwise, return None."""
-        return self.writtenReferences.get(obj)
-        
-    def writeRoot(self, root):
-        """
-        Strategy is:
-        - write header
-        - wrap root object so everything is hashable
-        - compute size of objects which will be written
-          - need to do this in order to know how large the object refs
-            will be in the list/dict/set reference lists
-        - write objects
-          - keep objects in writtenReferences
-          - keep positions of object references in referencePositions
-          - write object references with the length computed previously
-        - computer object reference length
-        - write object reference positions
-        - write trailer
-        """
-        output = self.header
-        wrapped_root = self.wrapRoot(root)
-        should_reference_root = True#not isinstance(wrapped_root, HashableWrapper)
-        self.computeOffsets(wrapped_root, asReference=should_reference_root, isRoot=True)
-        self.trailer = self.trailer._replace(**{'objectRefSize':self.intSize(len(self.computedUniques))})
-        (_, output) = self.writeObjectReference(wrapped_root, output)
-        output = self.writeObject(wrapped_root, output, setReferencePosition=True)
-        
-        # output size at this point is an upper bound on how big the
-        # object reference offsets need to be.
-        self.trailer = self.trailer._replace(**{
-            'offsetSize':self.intSize(len(output)),
-            'offsetCount':len(self.computedUniques),
-            'offsetTableOffset':len(output),
-            'topLevelObjectNumber':0
-            })
-        
-        output = self.writeOffsetTable(output)
-        output += pack('!xxxxxxBBQQQ', *self.trailer)
-        self.file.write(output)
-
-    def wrapRoot(self, root):
-        if isinstance(root, bool):
-            if root is True:
-                return self.wrappedTrue
-            else:
-                return self.wrappedFalse
-        elif isinstance(root, float):
-            return FloatWrapper(root)
-        elif isinstance(root, set):
-            n = set()
-            for value in root:
-                n.add(self.wrapRoot(value))
-            return HashableWrapper(n)
-        elif isinstance(root, dict):
-            n = {}
-            for key, value in iteritems(root):
-                n[self.wrapRoot(key)] = self.wrapRoot(value)
-            return HashableWrapper(n)
-        elif isinstance(root, list):
-            n = []
-            for value in root:
-                n.append(self.wrapRoot(value))
-            return HashableWrapper(n)
-        elif isinstance(root, tuple):
-            n = tuple([self.wrapRoot(value) for value in root])
-            return HashableWrapper(n)
-        else:
-            return root
-
-    def incrementByteCount(self, field, incr=1):
-        self.byteCounts = self.byteCounts._replace(**{field:self.byteCounts.__getattribute__(field) + incr})
-
-    def computeOffsets(self, obj, asReference=False, isRoot=False):
-        def check_key(key):
-            if key is None:
-                raise InvalidPlistException('Dictionary keys cannot be null in plists.')
-            elif isinstance(key, Data):
-                raise InvalidPlistException('Data cannot be dictionary keys in plists.')
-            elif not isinstance(key, (bytes, str)):
-                raise InvalidPlistException('Keys must be strings.')
-        
-        def proc_size(size):
-            if size > 0b1110:
-                size += self.intSize(size)
-            return size
-        # If this should be a reference, then we keep a record of it in the
-        # uniques table.
-        if asReference:
-            if obj in self.computedUniques:
-                return
-            else:
-                self.computedUniques.add(obj)
-        
-        if obj is None:
-            self.incrementByteCount('nullBytes')
-        elif isinstance(obj, BoolWrapper):
-            self.incrementByteCount('boolBytes')
-        elif isinstance(obj, Uid):
-            size = self.intSize(obj)
-            self.incrementByteCount('uidBytes', incr=1+size)
-        elif isinstance(obj, int):
-            size = self.intSize(obj)
-            self.incrementByteCount('intBytes', incr=1+size)
-        elif isinstance(obj, FloatWrapper):
-            size = self.realSize(obj)
-            self.incrementByteCount('realBytes', incr=1+size)
-        elif isinstance(obj, datetime.datetime):    
-            self.incrementByteCount('dateBytes', incr=2)
-        elif isinstance(obj, Data):
-            size = proc_size(len(obj))
-            self.incrementByteCount('dataBytes', incr=1+size)
-        elif isinstance(obj, (str, bytes)):
-            size = proc_size(len(obj))
-            self.incrementByteCount('stringBytes', incr=1+size)
-        elif isinstance(obj, HashableWrapper):
-            obj = obj.value
-            if isinstance(obj, set):
-                size = proc_size(len(obj))
-                self.incrementByteCount('setBytes', incr=1+size)
-                for value in obj:
-                    self.computeOffsets(value, asReference=True)
-            elif isinstance(obj, (list, tuple)):
-                size = proc_size(len(obj))
-                self.incrementByteCount('arrayBytes', incr=1+size)
-                for value in obj:
-                    asRef = True
-                    self.computeOffsets(value, asReference=True)
-            elif isinstance(obj, dict):
-                size = proc_size(len(obj))
-                self.incrementByteCount('dictBytes', incr=1+size)
-                for key, value in iteritems(obj):
-                    check_key(key)
-                    self.computeOffsets(key, asReference=True)
-                    self.computeOffsets(value, asReference=True)
-        else:
-            raise InvalidPlistException("Unknown object type.")
-
-    def writeObjectReference(self, obj, output):
-        """Tries to write an object reference, adding it to the references
-           table. Does not write the actual object bytes or set the reference
-           position. Returns a tuple of whether the object was a new reference
-           (True if it was, False if it already was in the reference table)
-           and the new output.
-        """
-        position = self.positionOfObjectReference(obj)
-        if position is None:
-            self.writtenReferences[obj] = len(self.writtenReferences)
-            output += self.binaryInt(len(self.writtenReferences) - 1, byteSize=self.trailer.objectRefSize)
-            return (True, output)
-        else:
-            output += self.binaryInt(position, byteSize=self.trailer.objectRefSize)
-            return (False, output)
-
-    def writeObject(self, obj, output, setReferencePosition=False):
-        """Serializes the given object to the output. Returns output.
-           If setReferencePosition is True, will set the position the
-           object was written.
-        """
-        def proc_variable_length(format, length):
-            result = b''
-            if length > 0b1110:
-                result += pack('!B', (format << 4) | 0b1111)
-                result = self.writeObject(length, result)
-            else:
-                result += pack('!B', (format << 4) | length)
-            return result
-        
-        if isinstance(obj, str) and obj == unicodeEmpty:
-            # The Apple Plist decoder can't decode a zero length Unicode string.
-            obj = b''
-       
-        if setReferencePosition:
-            self.referencePositions[obj] = len(output)
-        
-        if obj is None:
-            output += pack('!B', 0b00000000)
-        elif isinstance(obj, BoolWrapper):
-            if obj.value is False:
-                output += pack('!B', 0b00001000)
-            else:
-                output += pack('!B', 0b00001001)
-        elif isinstance(obj, Uid):
-            size = self.intSize(obj)
-            output += pack('!B', (0b1000 << 4) | size - 1)
-            output += self.binaryInt(obj)
-        elif isinstance(obj, int):
-            byteSize = self.intSize(obj)
-            root = math.log(byteSize, 2)
-            output += pack('!B', (0b0001 << 4) | int(root))
-            output += self.binaryInt(obj, as_number=True)
-        elif isinstance(obj, FloatWrapper):
-            # just use doubles
-            output += pack('!B', (0b0010 << 4) | 3)
-            output += self.binaryReal(obj)
-        elif isinstance(obj, datetime.datetime):
-            timestamp = (obj - apple_reference_date).total_seconds()
-            output += pack('!B', 0b00110011)
-            output += pack('!d', float(timestamp))
-        elif isinstance(obj, Data):
-            output += proc_variable_length(0b0100, len(obj))
-            output += obj
-        elif isinstance(obj, str):
-            byteData = obj.encode('utf_16_be')
-            output += proc_variable_length(0b0110, len(byteData)//2)
-            output += byteData
-        elif isinstance(obj, bytes):
-            output += proc_variable_length(0b0101, len(obj))
-            output += obj
-        elif isinstance(obj, HashableWrapper):
-            obj = obj.value
-            if isinstance(obj, (set, list, tuple)):
-                if isinstance(obj, set):
-                    output += proc_variable_length(0b1100, len(obj))
-                else:
-                    output += proc_variable_length(0b1010, len(obj))
-            
-                objectsToWrite = []
-                for objRef in obj:
-                    (isNew, output) = self.writeObjectReference(objRef, output)
-                    if isNew:
-                        objectsToWrite.append(objRef)
-                for objRef in objectsToWrite:
-                    output = self.writeObject(objRef, output, setReferencePosition=True)
-            elif isinstance(obj, dict):
-                output += proc_variable_length(0b1101, len(obj))
-                keys = []
-                values = []
-                objectsToWrite = []
-                for key, value in iteritems(obj):
-                    keys.append(key)
-                    values.append(value)
-                for key in keys:
-                    (isNew, output) = self.writeObjectReference(key, output)
-                    if isNew:
-                        objectsToWrite.append(key)
-                for value in values:
-                    (isNew, output) = self.writeObjectReference(value, output)
-                    if isNew:
-                        objectsToWrite.append(value)
-                for objRef in objectsToWrite:
-                    output = self.writeObject(objRef, output, setReferencePosition=True)
-        return output
-    
-    def writeOffsetTable(self, output):
-        """Writes all of the object reference offsets."""
-        all_positions = []
-        writtenReferences = list(self.writtenReferences.items())
-        writtenReferences.sort(key=lambda x: x[1])
-        for obj,order in writtenReferences:
-            # Porting note: Elsewhere we deliberately replace empty unicdoe strings
-            # with empty binary strings, but the empty unicode string
-            # goes into writtenReferences.  This isn't an issue in Py2
-            # because u'' and b'' have the same hash; but it is in
-            # Py3, where they don't.
-            if bytes != str and obj == unicodeEmpty:
-                obj = b''
-            position = self.referencePositions.get(obj)
-            if position is None:
-                raise InvalidPlistException("Error while writing offsets table. Object not found. %s" % obj)
-            output += self.binaryInt(position, self.trailer.offsetSize)
-            all_positions.append(position)
-        return output
-    
-    def binaryReal(self, obj):
-        # just use doubles
-        result = pack('>d', obj.value)
-        return result
-    
-    def binaryInt(self, obj, byteSize=None, as_number=False):
-        result = b''
-        if byteSize is None:
-            byteSize = self.intSize(obj)
-        if byteSize == 1:
-            result += pack('>B', obj)
-        elif byteSize == 2:
-            result += pack('>H', obj)
-        elif byteSize == 4:
-            result += pack('>L', obj)
-        elif byteSize == 8:
-            if as_number:
-                result += pack('>q', obj)
-            else:
-                result += pack('>Q', obj)
-        elif byteSize <= 16:
-            try:
-                result = pack('>Q', 0) + pack('>Q', obj)
-            except struct_error as e:
-                raise InvalidPlistException("Unable to pack integer %d: %s" % (obj, e))
-        else:
-            raise InvalidPlistException("Core Foundation can't handle integers with size greater than 16 bytes.")
-        return result
-    
-    def intSize(self, obj):
-        """Returns the number of bytes necessary to store the given integer."""
-        # SIGNED
-        if obj < 0: # Signed integer, always 8 bytes
-            return 8
-        # UNSIGNED
-        elif obj <= 0xFF: # 1 byte
-            return 1
-        elif obj <= 0xFFFF: # 2 bytes
-            return 2
-        elif obj <= 0xFFFFFFFF: # 4 bytes
-            return 4
-        # SIGNED
-        # 0x7FFFFFFFFFFFFFFF is the max.
-        elif obj <= 0x7FFFFFFFFFFFFFFF: # 8 bytes signed
-            return 8
-        elif obj <= 0xffffffffffffffff: # 8 bytes unsigned
-            return 16
-        else:
-            raise InvalidPlistException("Core Foundation can't handle integers with size greater than 8 bytes.")
-    
-    def realSize(self, obj):
-        return 8

lib/cherrypy/_cpdispatch.py

@@ -206,12 +206,8 @@ except ImportError:
     def test_callable_spec(callable, args, kwargs):  # noqa: F811
         return None
 else:
-    getargspec = inspect.getargspec
-    # Python 3 requires using getfullargspec if
-    # keyword-only arguments are present
-    if hasattr(inspect, 'getfullargspec'):
-        def getargspec(callable):
-            return inspect.getfullargspec(callable)[:4]
+    def getargspec(callable):
+        return inspect.getfullargspec(callable)[:4]
 
 
 class LateParamPageHandler(PageHandler):

lib/cherrypy/_cperror.py

@@ -466,7 +466,7 @@ _HTTPErrorTemplate = '''<!DOCTYPE html PUBLIC
         <pre id="traceback">%(traceback)s</pre>
     <div id="powered_by">
       <span>
-        Powered by <a href="http://www.cherrypy.org">CherryPy %(version)s</a>
+        Powered by <a href="http://www.cherrypy.dev">CherryPy %(version)s</a>
       </span>
     </div>
     </body>
@@ -532,7 +532,8 @@ def get_error_page(status, **kwargs):
                     return result
             else:
                 # Load the template from this path.
-                template = io.open(error_page, newline='').read()
+                with io.open(error_page, newline='') as f:
+                    template = f.read()
         except Exception:
             e = _format_exception(*_exc_info())[-1]
             m = kwargs['message']

lib/cherrypy/_cpmodpy.py

@@ -339,11 +339,8 @@ LoadModule python_module modules/mod_python.so
                                      }
 
         mpconf = os.path.join(os.path.dirname(__file__), 'cpmodpy.conf')
-        f = open(mpconf, 'wb')
-        try:
+        with open(mpconf, 'wb') as f:
             f.write(conf_data)
-        finally:
-            f.close()
 
         response = read_process(self.apache_path, '-k start -f %s' % mpconf)
         self.ready = True

lib/cherrypy/_cprequest.py

@@ -169,7 +169,7 @@ def request_namespace(k, v):
 def response_namespace(k, v):
     """Attach response attributes declared in config."""
     # Provides config entries to set default response headers
-    # http://cherrypy.org/ticket/889
+    # http://cherrypy.dev/ticket/889
     if k[:8] == 'headers.':
         cherrypy.serving.response.headers[k.split('.', 1)[1]] = v
     else:
@@ -252,7 +252,7 @@ class Request(object):
     The query component of the Request-URI, a string of information to be
     interpreted by the resource. The query portion of a URI follows the
     path component, and is separated by a '?'. For example, the URI
-    'http://www.cherrypy.org/wiki?a=3&b=4' has the query component,
+    'http://www.cherrypy.dev/wiki?a=3&b=4' has the query component,
     'a=3&b=4'."""
 
     query_string_encoding = 'utf8'
@@ -742,6 +742,9 @@ class Request(object):
             if self.protocol >= (1, 1):
                 msg = "HTTP/1.1 requires a 'Host' request header."
                 raise cherrypy.HTTPError(400, msg)
+        else:
+            headers['Host'] = httputil.SanitizedHost(dict.get(headers, 'Host'))
+
         host = dict.get(headers, 'Host')
         if not host:
             host = self.local.name or self.local.ip

lib/cherrypy/lib/auth_digest.py

@@ -101,13 +101,12 @@ def get_ha1_file_htdigest(filename):
     """
     def get_ha1(realm, username):
         result = None
-        f = open(filename, 'r')
-        for line in f:
-            u, r, ha1 = line.rstrip().split(':')
-            if u == username and r == realm:
-                result = ha1
-                break
-        f.close()
+        with open(filename, 'r') as f:
+            for line in f:
+                u, r, ha1 = line.rstrip().split(':')
+                if u == username and r == realm:
+                    result = ha1
+                    break
         return result
 
     return get_ha1

lib/cherrypy/lib/covercp.py

@@ -334,9 +334,10 @@ class CoverStats(object):
         yield '</body></html>'
 
     def annotated_file(self, filename, statements, excluded, missing):
-        source = open(filename, 'r')
+        with open(filename, 'r') as source:
+            lines = source.readlines()
         buffer = []
-        for lineno, line in enumerate(source.readlines()):
+        for lineno, line in enumerate(lines):
             lineno += 1
             line = line.strip('\n\r')
             empty_the_buffer = True

lib/cherrypy/lib/httputil.py

@@ -516,3 +516,33 @@ class Host(object):
 
     def __repr__(self):
         return 'httputil.Host(%r, %r, %r)' % (self.ip, self.port, self.name)
+
+
+class SanitizedHost(str):
+    r"""
+    Wraps a raw host header received from the network in
+    a sanitized version that elides dangerous characters.
+
+    >>> SanitizedHost('foo\nbar')
+    'foobar'
+    >>> SanitizedHost('foo\nbar').raw
+    'foo\nbar'
+
+    A SanitizedInstance is only returned if sanitization was performed.
+
+    >>> isinstance(SanitizedHost('foobar'), SanitizedHost)
+    False
+    """
+    dangerous = re.compile(r'[\n\r]')
+
+    def __new__(cls, raw):
+        sanitized = cls._sanitize(raw)
+        if sanitized == raw:
+            return raw
+        instance = super().__new__(cls, sanitized)
+        instance.raw = raw
+        return instance
+
+    @classmethod
+    def _sanitize(cls, raw):
+        return cls.dangerous.sub('', raw)

lib/cherrypy/lib/reprconf.py

@@ -163,11 +163,8 @@ class Parser(configparser.ConfigParser):
             #     fp = open(filename)
             # except IOError:
             #     continue
-            fp = open(filename)
-            try:
+            with open(filename) as fp:
                 self._read(fp, filename)
-            finally:
-                fp.close()
 
     def as_dict(self, raw=False, vars=None):
         """Convert an INI file to a dictionary"""

lib/cherrypy/lib/sessions.py

@@ -516,11 +516,8 @@ class FileSession(Session):
         if path is None:
             path = self._get_file_path()
         try:
-            f = open(path, 'rb')
-            try:
+            with open(path, 'rb') as f:
                 return pickle.load(f)
-            finally:
-                f.close()
         except (IOError, EOFError):
             e = sys.exc_info()[1]
             if self.debug:
@@ -531,11 +528,8 @@ class FileSession(Session):
     def _save(self, expiration_time):
         assert self.locked, ('The session was saved without being locked.  '
                              "Check your tools' priority levels.")
-        f = open(self._get_file_path(), 'wb')
-        try:
+        with open(self._get_file_path(), 'wb') as f:
             pickle.dump((self._data, expiration_time), f, self.pickle_protocol)
-        finally:
-            f.close()
 
     def _delete(self):
         assert self.locked, ('The session deletion without being locked.  '

lib/cherrypy/process/plugins.py

@@ -436,7 +436,8 @@ class PIDFile(SimplePlugin):
         if self.finalized:
             self.bus.log('PID %r already written to %r.' % (pid, self.pidfile))
         else:
-            open(self.pidfile, 'wb').write(ntob('%s\n' % pid, 'utf8'))
+            with open(self.pidfile, 'wb') as f:
+                f.write(ntob('%s\n' % pid, 'utf8'))
             self.bus.log('PID %r written to %r.' % (pid, self.pidfile))
             self.finalized = True
     start.priority = 70