mirror of
https://github.com/sprockets/sprockets.mixins.avro-publisher.git
synced 2024-11-29 11:19:51 +00:00
148 lines
5.1 KiB
Python
148 lines
5.1 KiB
Python
"""The AvroPublishingMixin wraps RabbitMQ use into a request handler, with
|
|
methods to speed the development of publishing RabbitMQ messages serialized
|
|
as Avro datums.
|
|
|
|
RabbitMQ is configured using two environment variables: ``AMQP_URL`` and
|
|
``AMQP_TIMEOUT``.
|
|
|
|
``AMQP_URL`` is the AMQP url to connect to, defaults to
|
|
``amqp://guest:guest@localhost:5672/%2f``.
|
|
|
|
``AMQP_TIMEOUT`` is the number of seconds to wait until timing out when
|
|
connecting to RabbitMQ.
|
|
|
|
To configure the URL format for the avro schema, add a
|
|
Tornado application setting called ``avro_schema_uri_format``. The format
|
|
should be similar to the following:
|
|
|
|
``http://my-schema-repository/avro/%(name)s.avsc``
|
|
|
|
"""
|
|
import io
|
|
import logging
|
|
import sys
|
|
|
|
from sprockets.mixins import amqp
|
|
from tornado import gen
|
|
from tornado import httpclient
|
|
import avro.io
|
|
import avro.schema
|
|
|
|
version_info = (0, 1, 0)
|
|
__version__ = '.'.join(str(v) for v in version_info)
|
|
|
|
LOGGER = logging.getLogger(__name__)
|
|
|
|
PYTHON3 = True if sys.version_info > (3, 0, 0) else False
|
|
|
|
|
|
class AvroPublishingMixin(amqp.PublishingMixin):
|
|
"""The request handler will connect to RabbitMQ on the first request,
|
|
blocking until the connection and channel are established. If RabbitMQ
|
|
closes it's connection to the app at any point, a connection attempt will
|
|
be made on the next request.
|
|
|
|
This class implements a pattern for the use of a single AMQP connection
|
|
to RabbitMQ.
|
|
|
|
Expects the :envvar:`AMQP_URL` environment variable to construct
|
|
:class:`pika.connection.URLParameters`.
|
|
|
|
"""
|
|
DATUM_MIME_TYPE = 'application/vnd.apache.avro.datum'
|
|
DEFAULT_SCHEMA_URI_FORMAT = 'http://localhost/avro/%(name)s.avsc'
|
|
|
|
def initialize(self):
|
|
"""Initialize the RequestHandler ensuring there is an a dict for
|
|
caching avro schemas.
|
|
|
|
"""
|
|
super(AvroPublishingMixin, self).initialize()
|
|
if 'avro_schema_uri_format' not in self.application.settings:
|
|
LOGGER.warning('avro_schema_uri_format is not set, using default')
|
|
if not hasattr(self.application, 'avro'):
|
|
self.application.avro = {}
|
|
|
|
@gen.coroutine
|
|
def amqp_publish(self, exchange, routing_key, body, properties):
|
|
"""Publish the message to RabbitMQ
|
|
|
|
:param str exchange: The exchange to publish to
|
|
:param str routing_key: The routing key to publish with
|
|
:param dict body: The message body
|
|
:param dict properties: The message properties
|
|
|
|
"""
|
|
if ('content_type' in properties and
|
|
properties['content_type']) == self.DATUM_MIME_TYPE:
|
|
body = yield self._avro_serialize(properties['type'], body)
|
|
yield self.application.amqp.publish(exchange, routing_key, body,
|
|
properties)
|
|
|
|
@gen.coroutine
|
|
def _avro_fetch_schema(self, schema_name):
|
|
"""Fetch the avro schema file from the remote HTTP endpoint
|
|
|
|
:param str schema_name: The schema name
|
|
:rtype: str
|
|
|
|
"""
|
|
http_client = httpclient.AsyncHTTPClient()
|
|
url = self._avro_schema_url(schema_name)
|
|
LOGGER.info('Loading schema for %s from %s', schema_name, url)
|
|
try:
|
|
response = yield http_client.fetch(url)
|
|
except httpclient.HTTPError as error:
|
|
LOGGER.error('Could not fetch Avro schema for %s (%s)', schema_name,
|
|
error)
|
|
raise ValueError('Error fetching avro schema')
|
|
raise gen.Return(response.body)
|
|
|
|
@gen.coroutine
|
|
def _avro_schema(self, schema_name):
|
|
"""Fetch the Avro schema file from cache or the filesystem.
|
|
|
|
:param str schema_name: The avro schema name
|
|
:rtype: str
|
|
|
|
"""
|
|
if schema_name not in self.application.avro:
|
|
schema = yield self._avro_fetch_schema(schema_name)
|
|
if PYTHON3:
|
|
schema = str(schema, 'utf-8')
|
|
self.application.avro[schema_name] = avro.schema.Parse(schema)
|
|
else:
|
|
self.application.avro[schema_name] = avro.schema.parse(schema)
|
|
raise gen.Return(self.application.avro[schema_name])
|
|
|
|
def _avro_schema_url(self, schema_name):
|
|
"""Return the Avro schema URL for the specified schema name.
|
|
|
|
:param str schema_name: The avro schema name
|
|
:rtype: str
|
|
|
|
"""
|
|
if 'avro_schema_uri_format' in self.application.settings:
|
|
schema_format = self.application.settings['avro_schema_uri_format']
|
|
else:
|
|
schema_format = self.DEFAULT_SCHEMA_URI_FORMAT
|
|
return schema_format % {'name': schema_name}
|
|
|
|
@gen.coroutine
|
|
def _avro_serialize(self, schema_name, data):
|
|
"""Serialize a data structure into an Avro datum
|
|
|
|
:param str schema_name: The Avro schema name
|
|
:param dict data: The value to turn into an Avro datum
|
|
:rtype: str
|
|
|
|
"""
|
|
schema = yield self._avro_schema(schema_name)
|
|
bytes_io = io.BytesIO()
|
|
encoder = avro.io.BinaryEncoder(bytes_io)
|
|
writer = avro.io.DatumWriter(schema)
|
|
try:
|
|
writer.write(data, encoder)
|
|
except avro.io.AvroTypeException as error:
|
|
raise ValueError(error)
|
|
raise gen.Return(bytes_io.getvalue())
|