pika queue manager works now in basic version

This commit is contained in:
Matthias Bisping 2022-03-08 13:37:40 +01:00
parent be0a4fe58f
commit 21592342d5
13 changed files with 278 additions and 1 deletions

View File

@ -1,5 +1,5 @@
service: service:
logging_level: $LOGGING_LEVEL_ROOT|DEBUG # Logging level for service logger logging_level: $LOGGING_LEVEL_ROOT|INFO # Logging level for service logger
response: response:
type: $RESPONSE_TYPE|"file" # Whether the analysis response is stored as file on storage or sent as stream type: $RESPONSE_TYPE|"file" # Whether the analysis response is stored as file on storage or sent as stream
extension: $RESPONSE_FILE_EXTENSION|".IMAGE_INFO.json.gz" # {.IMAGE_INFO.json.gz | .NER_ENTITIES.json.gz} extension: $RESPONSE_FILE_EXTENSION|".IMAGE_INFO.json.gz" # {.IMAGE_INFO.json.gz | .NER_ENTITIES.json.gz}
@ -37,6 +37,7 @@ storage:
backend: $STORAGE_BACKEND|s3 # The type of storage to use {s3, azure} backend: $STORAGE_BACKEND|s3 # The type of storage to use {s3, azure}
bucket: $STORAGE_BUCKET|"pyinfra-test-bucket" # The bucket / container to pull files specified in queue requests from bucket: $STORAGE_BUCKET|"pyinfra-test-bucket" # The bucket / container to pull files specified in queue requests from
# TODO: Caller should specify exact file name, including extension!
target_file_extension: $TARGET_FILE_EXTENSION|".ORIGIN.pdf.gz" # {.TEXT.json.gz | .ORIGIN.pdf.gz} Defines type of file to pull from storage target_file_extension: $TARGET_FILE_EXTENSION|".ORIGIN.pdf.gz" # {.TEXT.json.gz | .ORIGIN.pdf.gz} Defines type of file to pull from storage
s3: s3:

View File

View File

View File

@ -0,0 +1,8 @@
import abc
class Connection(abc.ABC):
@abc.abstractmethod
def establish(self):
pass

View File

@ -0,0 +1,7 @@
from pyampq.connection.connection import Connection
class MockConnection(Connection):
def establish(self):
pass

View File

@ -0,0 +1,4 @@
class PikaConnection:
def establish(self):
pass

View File

@ -0,0 +1,13 @@
from pyinfra.pyampq.queue_manager.queue_manager import QueueManager
class Consumer:
def __init__(self, callback, queue_manager: QueueManager):
self.queue_manager = queue_manager
self.callback = callback
def consume_and_publish(self):
yield from self.queue_manager.consume_and_publish(self.callback)
def consume(self):
yield from self.queue_manager.consume()

View File

View File

@ -0,0 +1,107 @@
import json
import logging
import pika
from pyinfra.config import CONFIG
from pyinfra.pyampq.queue_manager.queue_manager import QueueManager, QueueHandle
logger = logging.getLogger("pika")
logger.setLevel(logging.WARNING)
logger = logging.getLogger(__name__)
logger.setLevel(CONFIG.service.logging_level)
def monkey_patch_queue_handle(channel, queue) -> QueueHandle:
empty_message = (None, None, None)
def is_empty_message(message):
return message == empty_message
queue_handle = QueueHandle()
queue_handle.empty = lambda: is_empty_message(channel.basic_get(queue))
def produce_items():
while True:
message = channel.basic_get(queue)
print(message)
if is_empty_message(message):
break
method_frame, properties, body = message
channel.basic_ack(method_frame.delivery_tag)
yield json.loads(body)
queue_handle.to_list = lambda: list(produce_items())
return queue_handle
def get_connection():
credentials = pika.PlainCredentials(
username=CONFIG.rabbitmq.user,
password=CONFIG.rabbitmq.password
)
kwargs = {
"host": CONFIG.rabbitmq.host,
"port": CONFIG.rabbitmq.port,
"credentials": credentials
}
parameters = pika.ConnectionParameters(**kwargs)
connection = pika.BlockingConnection(parameters=parameters)
return connection
class PikaQueueManager(QueueManager):
def __init__(self, input_queue, output_queue):
super().__init__(input_queue, output_queue)
connection = get_connection()
self.channel = connection.channel()
self.channel.queue_declare(input_queue)
self.channel.queue_declare(output_queue)
def publish_request(self, request):
self.channel.basic_publish("", self._input_queue, json.dumps(request).encode())
def publish_response(self, message, callback):
frame, properties, body = message
response = json.dumps(callback(json.loads(body))).encode()
self.channel.basic_publish("", self._output_queue, response)
self.channel.basic_ack(frame.delivery_tag)
def pull_request(self):
return self.channel.basic_get(self._input_queue)
def consume(self):
return self.channel.consume(self._input_queue)
def consume_and_publish(self, callback):
logger.info(f"Consuming with callback {callback.__name__}")
for message in self.consume():
self.publish_response(message, callback)
def clear(self):
self.channel.queue_purge(self._input_queue)
self.channel.queue_purge(self._output_queue)
@property
def input_queue(self) -> QueueHandle:
return monkey_patch_queue_handle(self.channel, self._input_queue)
@property
def output_queue(self) -> QueueHandle:
return monkey_patch_queue_handle(self.channel, self._output_queue)

View File

@ -0,0 +1,49 @@
import abc
class QueueHandle:
def empty(self) -> bool:
raise NotImplemented()
def to_list(self) -> list:
raise NotImplemented()
class QueueManager(abc.ABC):
def __init__(self, input_queue, output_queue):
self._input_queue = input_queue
self._output_queue = output_queue
@abc.abstractmethod
def publish_request(self, request):
pass
@abc.abstractmethod
def publish_response(self, response, callback):
pass
@abc.abstractmethod
def pull_request(self):
pass
@abc.abstractmethod
def consume(self):
pass
@abc.abstractmethod
def clear(self):
pass
@abc.abstractmethod
def input_queue(self) -> QueueHandle:
pass
@abc.abstractmethod
def output_queue(self) -> QueueHandle:
pass
@abc.abstractmethod
def consume_and_publish(self, callback):
pass

View File

@ -0,0 +1,42 @@
from pyinfra.pyampq.queue_manager.queue_manager import QueueManager, QueueHandle
from pyinfra.test.queue_mock import QueueMock
def monkey_patch_queue_handle(queue) -> QueueHandle:
queue_handle = QueueHandle()
queue_handle.empty = lambda: not queue
queue_handle.to_list = lambda: list(queue)
return queue_handle
class QueueManagerMock(QueueManager):
def __init__(self, input_queue, output_queue):
super().__init__(QueueMock(), QueueMock())
def publish_request(self, request):
self._input_queue.append(request)
def publish_response(self, response):
self._output_queue.append(response)
def pull_request(self):
return self._input_queue.pop()
def consume(self, callback):
while self._input_queue:
request = self._input_queue.popleft()
response = callback(request)
self._output_queue.append(response)
def clear(self):
self._input_queue.clear()
self._output_queue.clear()
@property
def input_queue(self) -> QueueHandle:
return monkey_patch_queue_handle(self._input_queue)
@property
def output_queue(self) -> QueueHandle:
return monkey_patch_queue_handle(self._output_queue)

View File

@ -0,0 +1,5 @@
from collections import deque
class QueueMock(deque):
pass

View File

@ -0,0 +1,41 @@
import pytest
from pyinfra.pyampq.consumer import Consumer
from pyinfra.pyampq.queue_manager.pika_queue_manager import PikaQueueManager
from pyinfra.pyampq.queue_manager.queue_manager import QueueManager
from pyinfra.test.queue_manager_mock import QueueManagerMock
@pytest.fixture
def callback():
return lambda x: x**2
@pytest.fixture(scope="session")
def queue_manager() -> QueueManager:
# return QueueManagerMock("input", "output")
return PikaQueueManager("input", "output")
@pytest.fixture
def consumer(queue_manager, callback):
return Consumer(callback, queue_manager)
# def test_consuming_empty_input_queue_does_not_put_anything_on_output_queue(consumer, queue_manager):
# queue_manager.clear()
# consumer.consume()
# assert queue_manager.output_queue.empty()
def test_consuming_nonempty_input_queue_puts_messages_on_output_queue_in_fifo_order(consumer, queue_manager, callback):
queue_manager.clear()
queue_manager.publish_request(1)
queue_manager.publish_request(2)
requests = consumer.consume()
for i, r in enumerate(requests):
queue_manager.publish_response(r, callback)
if i == 1:
break
assert queue_manager.output_queue.to_list() == [1, 4]