def __init__(self, nethost: str, netport: int, database: str, user: str,
concurrency: int, protocol: str, **kwargs):
super().__init__(**kwargs)
if protocol != self.get_proto_name():
raise RuntimeError(f'unknown protocol {protocol!r}')
if concurrency <= 0 or concurrency > defines.HTTP_PORT_MAX_CONCURRENCY:
raise RuntimeError(
f'concurrency must be greater than 0 and '
f'less than {defines.HTTP_PORT_MAX_CONCURRENCY}')
self._compilers = asyncio.LifoQueue()
self._pgcons = asyncio.LifoQueue()
self._compilers_list = []
self._pgcons_list = []
self._nethost = nethost
self._netport = netport
self.database = database
self.user = user
self.concurrency = concurrency
self._servers = []
self._query_cache = cache.StatementsCache(
maxsize=defines.HTTP_PORT_QUERY_CACHE_SIZE)
def create_task_queues(app, loop):
app.threaded_executor = ThreadPoolExecutor(config.thread_pool_size)
app.updates_queue = asyncio.Queue(maxsize=1_000)
app.ls_queue = asyncio.LifoQueue(maxsize=100)
app.add_task(
b2browser.b2.worker(f"ls_worker", app.ls_queue, app.updates_queue,
jinja))
app.preview_queue = asyncio.LifoQueue(maxsize=10_000)
for i in range(config.thread_pool_size):
app.add_task(
b2browser.preview.worker(f"preview_worker_{i}",
app.threaded_executor, app.preview_queue,
app.updates_queue))
def __init__(self, event_loop=None, n_worker=1):
self.event_loop = event_loop
self.n_worker = n_worker
self._awaiting_worker_queue = asyncio.Queue(loop=self.event_loop,
maxsize=self.n_worker)
self._processing_worker_queue = asyncio.LifoQueue(
loop=self.event_loop, maxsize=self.n_worker)
async def crawl(self, seed=''):
counts = {}
visited = set()
count_kws = keyword_counter(self.keywords)
q = asyncio.LifoQueue()
await q.put((0, seed or self.DEFAULT_SEED))
while not q.empty():
depth, url = await q.get()
visited.add(url)
print(url)
content = await fetch(self.session, url)
parser = BeautifulSoup(content, 'lxml')
if self.is_story(url):
story = self.get_story(parser)
story = normalize_story(story.text) if story else ''
counts[url] = count_kws(story)
pprint(counts[url])
if depth < self.MAX_DEPTH:
for link in parser.find_all('a', href=True):
href = link['href']
if href not in self.visited and self.should_visit(href):
await q.put((depth + 1, href))
return counts
def __init__(self, *args, **kwargs):
self.pool_size = kwargs.pop('pool_size', 8)
self.connection_args = args
self.connection_kwargs = kwargs
self.queue = asyncio.LifoQueue() # important because released connections are put in the end
self.connection_list = [] # used to close in the end
self._async__init__() # not sure how to model this
def test_order(self):
q = asyncio.LifoQueue(loop=self.loop)
for i in [1, 3, 2]:
q.put_nowait(i)
items = [q.get_nowait() for _ in range(3)]
self.assertEqual([2, 3, 1], items)
async def run(loop):
device = None
print('looking for devices...')
while device is None:
devices = await discover()
device = next((d for d in devices if d.name == 'Lyd2'), None)
async with BleakClient(device.address, loop=loop) as client:
queue = asyncio.LifoQueue(maxsize=1)
print('disconnecting...')
await client.disconnect()
print('connecting...')
await client.connect()
print('starting notificiations')
def callback(sender, data):
print(f"{sender}: {data}")
await client.start_notify(
characteristicUuid,
lambda _, data: loop.call_soon(queue.put_nowait, data))
print('waiting')
while True:
args = await queue.get()
music.play()
print(args)
def __init__(self, nethost: str, netport: int, database: str, user: str,
concurrency: int, protocol: str, **kwargs):
super().__init__(**kwargs)
self._compiler_pool_size = 0
if protocol != self.get_proto_name():
raise RuntimeError(f'unknown protocol {protocol!r}')
if concurrency <= 0 or concurrency > defines.HTTP_PORT_MAX_CONCURRENCY:
raise RuntimeError(
f'concurrency must be greater than 0 and '
f'less than {defines.HTTP_PORT_MAX_CONCURRENCY}')
self._compilers: asyncio.LifoQueue[Any] = asyncio.LifoQueue()
self._compilers_list: List[Any] = []
self._nethost = nethost
self._netport = netport
self.database = database
self.user = user
self.concurrency = concurrency
self.last_minute_requests = windowedsum.WindowedSum()
self._http_proto_server = None
self._http_request_logger = None
self._query_cache = cache.StatementsCache(
maxsize=defines.HTTP_PORT_QUERY_CACHE_SIZE)
def setUp(self):
super(EntryWorkerTestCase, self).setUp()
self.queue = asyncio.LifoQueue()
self.loop.run_until_complete(self.init_queue())
self.buf = set() # Having add() method is enough
self.sess = SessionStub()
def __init__(self, factory: tp.Coroutine, size: int = POOL_SIZE):
self._factory = factory
self._size = size
self._queue = asyncio.LifoQueue(maxsize=self._size)
self._used = set()
for _ in range(self._size):
self._queue.put_nowait(None)
def run():
bsr_end = RabbitmqInputEndpoint('amz_bsr:input', **RABBITMQ_CONF)
output_end = RabbitmqOutputEndpoint(
['amz_bsr:input', 'amz_bsr:output', 'amz_ip_ban:input'],
**RABBITMQ_CONF)
queue = asyncio.Queue()
notify_input_end = pipeflow.QueueInputEndpoint(queue)
notify_output_end = pipeflow.QueueOutputEndpoint(queue)
queue = asyncio.LifoQueue()
inner_input_end = pipeflow.QueueInputEndpoint(queue)
inner_output_end = pipeflow.QueueOutputEndpoint(queue)
server = pipeflow.Server()
task_group = server.add_group('task', 1)
task_group.set_handle(handle_task)
task_group.add_input_endpoint('input', bsr_end)
task_group.add_input_endpoint('notify', notify_input_end)
task_group.add_output_endpoint('input_back', output_end, 'amz_bsr:input')
task_group.add_output_endpoint('inner_output', inner_output_end)
worker_group = server.add_group('work', MAX_WORKERS)
worker_group.set_handle(handle_worker)
worker_group.add_input_endpoint('inner_input', inner_input_end)
worker_group.add_output_endpoint('output',
output_end,
'amz_bsr:output',
buffer_size=MAX_WORKERS * 20)
worker_group.add_output_endpoint('inner_output', inner_output_end)
worker_group.add_output_endpoint('notify', notify_output_end)
worker_group.add_output_endpoint('ban', output_end, 'amz_ip_ban:input')
server.run()
def __init__(self, *connect_args,
min_size,
max_size,
max_queries,
max_inactive_connection_lifetime,
setup,
init,
loop,
connection_class,
**connect_kwargs):
if loop is None:
loop = asyncio.get_event_loop()
self._loop = loop
if max_size <= 0:
raise ValueError('max_size is expected to be greater than zero')
if min_size < 0:
raise ValueError(
'min_size is expected to be greater or equal to zero')
if min_size > max_size:
raise ValueError('min_size is greater than max_size')
if max_queries <= 0:
raise ValueError('max_queries is expected to be greater than zero')
if max_inactive_connection_lifetime < 0:
raise ValueError(
'max_inactive_connection_lifetime is expected to be greater '
'or equal to zero')
self._minsize = min_size
self._maxsize = max_size
self._holders = []
self._initialized = False
self._queue = asyncio.LifoQueue(maxsize=self._maxsize, loop=self._loop)
self._working_addr = None
self._working_config = None
self._working_params = None
self._connection_class = connection_class
self._closed = False
for _ in range(max_size):
ch = PoolConnectionHolder(
self,
connect_args=connect_args,
connect_kwargs=connect_kwargs,
max_queries=max_queries,
max_inactive_time=max_inactive_connection_lifetime,
setup=setup,
init=init)
self._holders.append(ch)
self._queue.put_nowait(ch)
async def test_order(self):
q = asyncio.LifoQueue()
for i in [1, 3, 2]:
await q.put(i)
items = [await q.get() for _ in range(3)]
self.assertEqual([2, 3, 1], items)
def __init__(self, root, *, max_tasks=10, loop=None):
self._loop = loop or asyncio.get_event_loop()
self._root = URL(root)
self._q = asyncio.LifoQueue(loop=self._loop)
self._max_tasks = max_tasks
self._session = aiohttp.ClientSession(loop=self._loop)
self._fetched_url_num = 0
async def main(contracts, number_of_workers, now=datetime.now()):
asyncio.get_event_loop().set_debug(True)
log.debug('main function started')
await ib.qualifyContractsAsync(*contracts)
log.debug('contracts qualified')
queue = asyncio.LifoQueue()
producers = [
asyncio.create_task(
schedule_task(**initial_schedule(c, now), queue=queue))
for c in contracts
]
workers = [
asyncio.create_task(worker(f'worker {i}', queue))
for i in range(number_of_workers)
]
await asyncio.gather(*producers, return_exceptions=True)
# wait until the queue is fully processed (implicitly awaits workers)
await queue.join()
# cancel all workers
log.debug('cancelling workers')
for w in workers:
w.cancel()
# wait until all worker tasks are cancelled
await asyncio.gather(*workers, return_exceptions=True)
async def _initialize(self):
self._queue = asyncio.LifoQueue(maxsize=self._maxsize)
for _ in range(self._maxsize):
ch = PoolConnectionHolder(
self,
max_queries=self._max_queries,
max_inactive_time=self._max_inactive_connection_lifetime,
setup=self._setup)
self._holders.append(ch)
self._queue.put_nowait(ch)
if self._minsize:
# Since we use a LIFO queue, the first items in the queue will be
# the last ones in `self._holders`. We want to pre-connect the
# first few connections in the queue, therefore we want to walk
# `self._holders` in reverse.
# Connect the first connection holder in the queue so that it
# can record `_working_addr` and `_working_opts`, which will
# speed up successive connection attempts.
first_ch = self._holders[-1] # type: PoolConnectionHolder
await first_ch.connect()
if self._minsize > 1:
connect_tasks = []
for i, ch in enumerate(reversed(self._holders[:-1])):
# `minsize - 1` because we already have first_ch
if i >= self._minsize - 1:
break
connect_tasks.append(ch.connect())
await asyncio.gather(*connect_tasks)
async def _async__init__(self):
if self._initialized:
return
if self._initializing:
raise errors.InterfaceError(
'pool is being initialized in another task')
if self._closed:
raise errors.InterfaceError('pool is closed')
self._initializing = True
self._queue = asyncio.LifoQueue(maxsize=self._maxsize)
for _ in range(self._maxsize):
ch = PoolConnectionHolder(self,
on_acquire=self._on_acquire,
on_release=self._on_release)
self._holders.append(ch)
self._queue.put_nowait(ch)
try:
await self._initialize()
return self
finally:
self._initializing = False
self._initialized = True
async def main(holder: ContractHolder):
log.debug('inside main')
contracts = holder()
log.debug('past holder')
number_of_workers = min(len(contracts), max_number_of_workers)
log.debug(f'main function started, '
f'retrieving data for {len(contracts)} instruments')
queue = asyncio.LifoQueue()
for contract in contracts:
await queue.put(contract)
workers = [
asyncio.create_task(worker(f'worker {i}', queue))
for i in range(number_of_workers)
]
await queue.join()
# cancel all workers
log.debug('cancelling workers')
for w in workers:
w.cancel()
# wait until all worker tasks are cancelled
await asyncio.gather(*workers)
def __init__(self, *connect_args,
min_size,
max_size,
on_acquire,
on_release,
on_connect,
connection_class,
**connect_kwargs):
loop = asyncio.get_event_loop()
self._loop = loop
if max_size <= 0:
raise ValueError('max_size is expected to be greater than zero')
if min_size < 0:
raise ValueError(
'min_size is expected to be greater or equal to zero')
if min_size > max_size:
raise ValueError('min_size is greater than max_size')
if not issubclass(connection_class, asyncio_con.AsyncIOConnection):
raise TypeError(
'connection_class is expected to be a subclass of '
'edgedb.AsyncIOConnection, got {!r}'.format(connection_class))
self._minsize = min_size
self._maxsize = max_size
self._holders = []
self._initialized = False
self._initializing = False
self._queue = asyncio.LifoQueue(maxsize=self._maxsize, loop=self._loop)
self._working_addr = None
self._working_config = None
self._working_params = None
self._connection_class = connection_class
self._closing = False
self._closed = False
self._generation = 0
self._on_connect = on_connect
self._connect_args = connect_args
self._connect_kwargs = connect_kwargs
for _ in range(max_size):
ch = PoolConnectionHolder(
self,
on_acquire=on_acquire,
on_release=on_release)
self._holders.append(ch)
self._queue.put_nowait(ch)
async def main():
assert len(
sys.argv
) > 1 # Supply the token to the bot as an argument to the script
async with websockets.connect(GATEWAY_URI) as websocket:
queue = asyncio.Queue(
) # Used to let the producer know what kind of message
heartbeatqueue = asyncio.LifoQueue(
) # Used for heartbeat messages, the first element that is added to this queue
await handler(websocket, queue, heartbeatqueue)
async def test_subscribe_custom_queue(self):
sp = AsyncSubPub()
q = asyncio.PriorityQueue()
q_out = await sp.subscribe('', queue=q)
self.assertIs(q_out, q)
q_out = await sp.subscribe('', queue=q)
self.assertIs(q_out, q)
q_out = await sp.subscribe('')
self.assertIsInstance(q_out, asyncio.Queue)
q_out = await sp.subscribe('', queue=asyncio.LifoQueue())
self.assertIsInstance(q_out, asyncio.LifoQueue)
async def main(contracts, number_of_workers, now=datetime.now()):
if logmode == "logging.INFO":
FileHandler(
f'{__file__[:-3]}_{datetime.today().strftime("%Y-%m-%d_%H-%M")}',
format_string=
'[{record.time:%y-%m-%d %H:%M:%S.%f%z}] {record.level_name}: {record.lineno}: {record.message}',
bubble=True,
delay=True).push_application()
#logger=createLogger(__name__,args.log,logging.INFO)
else:
FileHandler(
f'{__file__[:-3]}_{datetime.today().strftime("%Y-%m-%d_%H-%M")}',
format_string=
'[{record.time:%y-%m-%d %H:%M:%S.%f%z}] {record.level_name}: {record.lineno}: {record.message}',
bubble=True,
delay=True).push_application()
#logger=createLogger(__name__,args.log,logging.DEBUG)
logger = logging.getLogger(__name__)
# logger=createLogger(__name__,"scan.log",logging.DEBUG)
asyncio.get_event_loop().set_debug(True)
log.debug('main function started')
for i in contracts:
log.debug(contracts)
#log.debug('main function started'+contracts)
#await ib.qualifyContractsAsync(*contracts)
log.debug('contracts qualified')
queue = asyncio.LifoQueue()
producers = [
asyncio.create_task(
schedule_task(**initial_scheduleBF(c, backdays), queue=queue))
for c in contracts
]
workers = [
asyncio.create_task(worker(f'worker {i}', queue))
for i in range(number_of_workers)
]
await asyncio.gather(*producers, return_exceptions=True)
# wait until the queue is fully processed (implicitly awaits workers)
await queue.join()
# cancel all workers
log.debug('cancelling workers')
for w in workers:
w.cancel()
# wait until all worker tasks are cancelled
await asyncio.gather(*workers, return_exceptions=True)
async def s3_dir_dir(url, depth, dst_q, s3):
"""
Find directories certain depth from the base, push them to the `dst_q`
```
s3://bucket/a
|- b1
|- c1/...
|- c2/...
|- some_file.txt
|- b2
|- c3/...
```
Given a bucket structure above, calling this function with
- url s3://bucket/a/
- depth=1 will produce
- s3://bucket/a/b1/
- s3://bucket/a/b2/
- depth=2 will produce
- s3://bucket/a/b1/c1/
- s3://bucket/a/b1/c2/
- s3://bucket/a/b2/c3/
Any files are ignored.
"""
if not url.endswith('/'):
url = url + '/'
pp = s3.get_paginator('list_objects_v2')
async def step(bucket, prefix, depth, work_q, dst_q):
async for o in pp.paginate(Bucket=bucket, Prefix=prefix,
Delimiter='/'):
for d in o.get('CommonPrefixes', []):
d = d.get('Prefix')
if depth > 1:
await work_q.put((d, depth - 1))
else:
d = 's3://{}/{}'.format(bucket, d)
await dst_q.put(d)
bucket, prefix = s3_url_parse(url)
work_q = asyncio.LifoQueue()
work_q.put_nowait((prefix, depth))
while work_q.qsize() > 0:
_dir, depth = work_q.get_nowait()
await step(bucket, _dir, depth, work_q, dst_q)
def _get_stack(message_type):
if message_type in message_stacks:
return message_stacks[message_type]
else:
try:
ms_lock.acquire()
if message_type in message_stacks:
return message_stacks[message_type]
else:
selected_stack = asyncio.LifoQueue()
message_stacks[message_type] = selected_stack
return selected_stack
finally:
ms_lock.release()
async def using_queues():
q = asyncio.Queue()
q.put_nowait('Hello')
await q.get()
await q.put('world')
q.get_nowait()
pq = asyncio.PriorityQueue()
stack = asyncio.LifoQueue()
def _queue(self) -> asyncio.Queue[_T]:
# Delay creation of the queue until it is first used, to avoid
# binding it to a possibly wrong event loop.
# By delaying the creation of the pool we accommodate the common
# usage pattern of instantiating the engine at module level, where a
# different event loop is in present compared to when the application
# is actually run.
queue: asyncio.Queue[_T]
if self.use_lifo:
queue = asyncio.LifoQueue(maxsize=self.maxsize)
else:
queue = asyncio.Queue(maxsize=self.maxsize)
return queue
def __init__(self, asset_data: AssetData, **passargs):
"""Create a Downloader object.
Parameters:
asset_data: Asset Data
passargs: Keyword arguments
"""
self.asset_data = asset_data
self.passargs = passargs
self.finished = False
self._task: Optional[asyncio.Task] = None
self._queue: asyncio.Queue = asyncio.LifoQueue()
self._progress = 0
self._stop_event = threading.Event()
async def _run_push_pipeline(self, object_service_root: str, object_service_headers: dict,
objects: List[PushObject], progress_update_fn: Callable,
multipart_chunk_size: int, upload_chunk_size: int = 4194304,
num_workers: int = 4) -> None:
"""Method to run the async upload pipeline
Args:
object_service_root: The root URL to use for all objects, including the namespace and dataset name
object_service_headers: The headers to use when requesting signed urls, including auth info
objects: A list of PushObjects to push
progress_update_fn: A callable with arg "completed_bytes" (int) indicating how many bytes have been
uploaded in since last called
multipart_chunk_size: Size in bytes for break a file apart for multi-part uploading
upload_chunk_size: Size in bytes for streaming IO chunks
num_workers: the number of consumer workers to start
Returns:
"""
# We use a LifoQueue to ensure S3 uploads start as soon as they are ready to help ensure pre-signed urls do
# not timeout before they can be used if there are a lot of files.
queue: asyncio.LifoQueue = asyncio.LifoQueue()
async with aiohttp.ClientSession() as session:
# Start workers
workers = []
for i in range(num_workers):
task = asyncio.ensure_future(self._push_object_consumer(queue, session, progress_update_fn))
workers.append(task)
# Populate the work queue
await self._push_object_producer(queue,
object_service_root,
object_service_headers,
multipart_chunk_size,
upload_chunk_size,
objects)
# wait until the consumer has processed all items
await queue.join()
# the workers are still awaiting for work so close them
for worker in workers:
worker.cancel()
async def _run_pull_pipeline(self,
object_service_root: str,
object_service_headers: dict,
objects: List[PullObject],
status_update_fn: Callable,
download_chunk_size: int = 4194304,
num_workers: int = 4):
"""Method to run the async download pipeline
Args:
object_service_root: The root URL to use for all objects, including the namespace and dataset name
object_service_headers: The headers to use when requesting signed urls, including auth info
objects: A list of PushObjects to push
status_update_fn: the update function for providing feedback
download_chunk_size: Size in bytes for streaming IO chunks
num_workers: the number of consumer workers to start
Returns:
"""
# We use a LifoQueue to ensure S3 uploads start as soon as they are ready to help ensure pre-signed urls do
# not timeout before they can be used if there are a lot of files.
queue: asyncio.LifoQueue = asyncio.LifoQueue()
async with aiohttp.ClientSession() as session:
workers = []
for i in range(num_workers):
task = asyncio.ensure_future(
self._pull_object_consumer(queue, session,
status_update_fn))
workers.append(task)
# Populate the work queue
await self._pull_object_producer(queue, object_service_root,
object_service_headers,
download_chunk_size, objects)
# wait until the consumer has processed all items
await queue.join()
# the workers are still awaiting for work so close them
for worker in workers:
worker.cancel()
def __init__(self, pool_min_size: int, pool_max_size: int,
pool_max_queries: int, pool_loop, connect_url: str,
**connect_kwargs):
if pool_loop is None:
pool_loop = asyncio.get_event_loop()
self._loop = pool_loop
if pool_max_size <= 0:
raise ValueError('max_size is expected to be greater than zero')
if pool_min_size < 0:
raise ValueError(
'min_size is expected to be greater or equal to zero')
if pool_min_size > pool_max_size:
raise ValueError('min_size is greater than max_size')
if pool_max_queries < 0:
raise ValueError(
'max_queries is expected to be greater than or equal zero')
self._minsize = pool_min_size
self._maxsize = pool_max_size
self._holders = []
self._initialized = False
self._queue = asyncio.LifoQueue(loop=self._loop)
self._closing = False
self._closed = False
self._connect_url = connect_url
self._connect_kwargs = connect_kwargs
for _ in range(pool_max_size):
ch = PoolConnectionHolder(self, max_queries=pool_max_queries)
self._holders.append(ch)
self._queue.put_nowait(ch)
