a methodology question:
I have a "main" python script which runs on an infinite loop on my system, and I want to send information to it (a json data string for example) occasionally with some other python scripts that will be started later by myself or another program and will end just after sending the string.
I can't use subprocess here because my main script doesn't know when the other will run and what code they will execute.
I'm thinking of making the main script listen on a local port and making the other scripts send it the strings on that port, but is there a better way to do it ?
zeromq: http://www.zeromq.org/ - is best solution for interprocess communications imho and have a excelent binding for python: http://www.zeromq.org/bindings:python
Since the "main" script looks like a service you can enhance it with a web API. bottle is the perfect solution for this. With this additional code your python script is able to receive requests and process them:
import json
from bottle import run, post, request, response
@post('/process')
def my_process():
req_obj = json.loads(request.body.read())
# do something with req_obj
# ...
return 'All done'
run(host='localhost', port=8080, debug=True)
The client script may use the httplib to send a message to the server and read the response:
import httplib
c = httplib.HTTPConnection('localhost', 8080)
c.request('POST', '/process', '{}')
doc = c.getresponse().read()
print doc
# 'All done'
In case you are interested in implementing the client script that Mike presented in Python 3.x, you will quickly find that there is no httplib available. Fortunately, the same thing is done with the library http.client.
Otherwise it is the same:
import http.client
c = http.client.HTTPConnection('localhost', 8080)
c.request('POST', '/process', '{}')
doc = c.getresponse().read()
print(doc)
Though this is old I would figure I would post this since I had a similar question today but using a server.