Streaming, interval=0 about cppagent HOT 4 CLOSED

Curl doesn't have builtin support for mixed replace content, but many browsers do. The wikipedia link you sent is the correct reference. I have written software in C#, C, ruby, scala/java, and python to parse these messages.

The question back to you is what are you trying to do? If you can give a little more info I can provide a small util that will probably get you exactly what you want. In the current tools directory there is a module called long_pull.rb that parses the stream and blocks the data correctly. The agent dumper uses this module to dump the events in an adapter format.

The relevant code is:

require './long_pull'
require 'net/http'

client = Net::HTTP.new('agent.mtconnect.org', 80)
path = "/sample?interval=0&count=1000"
puller = LongPull.new(client)
puller.long_pull(path) do |xml|
  # do something with the block of xml, this just prints....
  puts xml
end

Will be happy to help...

Best,

• Will

from cppagent.

Hey Will,

maybe my approach is wrong. I'm reading the XML output from the Agent with Python to provide single elements with web sockets to a browser. (I tried pycurl). In addition, I'm currently displaying thinks like since when the machine is up, how many parts produced in what kind of time (productive time vs unproductive/idle time) or grab the error messages. It's a science machine that I'm monitoring so, there is no part number or so.

Further, I tried it with: http://docs.python-requests.org/en/latest/user/advanced.html#streaming-requests this works, due to its parsing line by line and I can easily filter --longnumber and Content-...

Here is my workaround:

import requests
from lxml import objectify, etree

class Client:
    def __init__(self):
        r = requests.get('http://agent.mtconnectorg/sample?interval=0', stream=True)

        self.buffer = ""
        for line in r.iter_lines():
            self.on_receive(line)

    def on_receive(self, data):
        if data.startswith("--"):
            data = ""
        if data.startswith("Content"):
            data = ""
        else:
            self.buffer += data

        if data.endswith('</MTConnectStreams>'):
            xml = self.buffer
            self.buffer = ""

            try:
                self.root = objectify.fromstring(xml)
                # objectify.deannotate(self.root, xsi_nil=True)
            except BaseException as myError:
                print(myError)
                pass

            self.header = self.root.Header
            self.streams = self.root.Streams

            print etree.tostring(self.header, pretty_print=True)

client = Client()

I'm not limited to a programming language, but I personally like python ;) The advantage for me to use these interval=0 is that the Agent acts like its pushing and I can just use easy if-states to do something. I don't need to look every second at the XML output..

Thanks,
Arne

from cppagent.

You may want to check this out:

https://github.com/mtconnect/ros_bridge/tree/master/mtconnect/mtconnect_ros_bridge/scripts

The long pull source is in

https://github.com/mtconnect/ros_bridge/tree/master/mtconnect/mtconnect_ros_bridge/scripts/src

and an example is:

https://github.com/mtconnect/ros_bridge/tree/master/mtconnect/mtconnect_ros_bridge/scripts/robot_sim.py

Here's a basic example, though I didn't pull in all the pieces. Check the module above for a complete example:

import urllib2
import httplib
import socket

def xml_get_response(data):
  # Unpack data
  url, url_port, port, conn, req = data

  # Get response from url    
  rospy.logdebug('Attempting HTTP connection on url: %s:%s\tPort:%s' % (url, url_port, port))
  conn.request("GET", req)
  response = conn.getresponse()
  if response.status != 200:
      rospy.logerror("Request failed: %s - %d" % (response.reason, response.status))
      sys.exit(0)
  else:
      rospy.logdebug('Request --> %s, Status --> %s' % (response.reason, response.status))
  return response

# ... some __init__ stuff
  self.conn = httplib.HTTPConnection('agent.mtconnect.org', 80)
  response = xml_get_response(('localhost', 5000, None, self.conn, "/cnc/sample?interval=0&count=1000"))
    lp = LongPull(response)
    lp.long_pull(callback)

def callback(self, xml)
  # do something with xml

from cppagent.

Thanks a lot! This is really nice and it was already there. I should look in more detail what you guys are posting.

Further I would suggest a faster method to enable cross-origin resource sharing than changing outgoing.headers["Access-Control-Allow-Origin"] = “*”; below https://github.com/mtconnect/cppagent/blob/master/agent/agent.cpp#L232

from cppagent.