Beanstalk is a simple, fast work queue. Its interface is generic, but was originally designed for reducing the latency of page views in high-volume web applications by running time-consuming tasks asynchronously

http://kr.github.io/beanstalkd/

pystalkd is a beanstalkd bindings targeting python3. This library is based on https://github.com/earl/beanstalkc and should be API compatible.

pip install pystalkd

or from source:

python setup.py install

from pystalkd.Beanstalkd import Connection
c = Connection("localhost", 11300) #if no argument is given default configuration is used
c.put("hey!")
job = c.reserve(0)
print(job.body) # "hey!"

One of the goals is to be API compatible with beanstalkc, so this tutorial should be valid: https://github.com/earl/beanstalkc/blob/master/TUTORIAL.mkd

The main differences, API wise are:

1. where number of seconds is expected pystalkd also accepts a timedelta object

2. you can temporarily watch and use a tube using the with keyword

from pystalkd.Beanstalkd import Connection
c = Connection("localhost", 11300)
print(c.using()) # "default"
with c.temporary_use("test"):
  print(c.using()) # "test"
print(c.using()) # "default"

print(c.watching()) # ["default"]
with c.temporary_use("test"):
  print(c.watching()) # ["default", "test"]
print(c.watching()) # ["default"]

3. you also have access to the "bytes" API. To maintain compatibility with beanstalkc the API worked only with strings but now you can use the functions ending in "_bytes" (internally this is controlled using the raw paramater) to work directly with bytes

from pystalkd.Beanstalkd import Connection
c = Connection("localhost", 11300)
from os import urandom
test_bytes = urandom(50)
job_id = c.put_bytes(test_bytes)
job = c.reserve_bytes(0)
print(job.body) # b'i\x91\xdf\xf8\x1b?zj....'

job_id2 = c.put("string")
job2 = c.reserve_bytes(0)
print(job2.body) # b'string'

Note: you can use reserve_bytes with put and get the raw string (not encoded), but the other way around might cause problems

To test with default host and port (localhost, 11300):

python3 test.py

To test on a specific host (if port is not specified 11300 is used)

python3 test.py host [port]

Copyright (C) 2008-2014 Andreas Bolka.

Copyright (C) 2015-2016 Gabriel Menezes. Licensed under the MIT.