Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #70770 > unrolled thread

Designing a network in Python

Back to article view | Back to comp.lang.python

Contents

  Designing a network in Python varun7rs@gmail.com - 2014-04-30 02:30 -0700
    RE: Designing a network in Python "Joseph L. Casale" <jcasale@activenetwerx.com> - 2014-04-30 12:28 +0000
      Re: Designing a network in Python varun7rs@gmail.com - 2014-04-30 07:57 -0700
        Re: Designing a network in Python Mark H Harris <harrismh777@gmail.com> - 2014-04-30 10:03 -0500
        RE: Designing a network in Python "Joseph L. Casale" <jcasale@activenetwerx.com> - 2014-04-30 18:38 +0000
          Re: Designing a network in Python varun7rs@gmail.com - 2014-05-02 02:33 -0700

#70770 — Designing a network in Python

Hello Friends

I would like to design a network given the topology and the source I use is 
http://sndlib.zib.de/home.action

I have managed to read most of the important data in the xml onto lists. Now, I have two lists, Source and Destination and I'd like to create bi-directional links between them. My code is as below. And moreover, I'd like to assign some kind of a bandwidth capacity to the links and similarly, storage and processing capacity to the nodes. So, I generated some random integers for them. I don't know how to proceed further to design the topology. I'd prefer to do the topology design in another .py file probably. I'd be glad if you guys could lend me a helping hand.

import sys
import xml.dom.minidom as dom 
import string
#from xml.dom.minidom import getDOMimplementation
from xml.dom import minidom
from xml.dom.minidom import parse
import os
import random



class PHY_NODES:
    def __init__(self, nodeID, x, y, capacity_proc, capacity_stor, capacity_switch):
        self.nodeID = nodeID
        self.x = x
        self.y = y
        self.linklist = []
        self.capacity_proc = capacity_proc
        self.capacity_stor = capacity_stor
        self.capacity_switch = capacity_switch
        
class PHY_LINKS:
    def __init__(self, linkID, source, destination, capacity_bdw):
        self.linkID = linkID
        self.source = source
        self.destination = destination
        self.capacity_bdw = capacity_bdw
        
# Reading Nodes

Read_Data = minidom.parse("germany50.xml")
nodelist = Read_Data.getElementsByTagName("node")
corenodes = []
for node in nodelist :
	corenodes.append(node.getAttribute("id"))
	nodeid = node.getAttribute("id")
	xCoordinates = node.getElementsByTagName("x") [0]
	yCoordinates = node.getElementsByTagName("y") [0]
	proc = random.randint(20, 40)
	stor = random.randint(40, 80)
	switch = random.randint(60, 90)
	nodeobj = PHY_NODES(nodeid, xCoordinates, yCoordinates, proc, stor, switch)



# Reading Links 

linklist = Read_Data.getElementsByTagName("link")

links = []
sources = []
destinations = []
capacity_link = []
for link in linklist :
	linkid = link.getAttribute("id")
	Source = link.getElementsByTagName("source") [0]
	Destination = link.getElementsByTagName("target") [0]
	Capacity = link.getElementsByTagName("capacity") [0]
	linkobj = PHY_LINKS(linkid, Source, Destination, Capacity)
	sources.append(Source.firstChild.data)
	destinations.append(Destination.firstChild.data)
	capacity_link.append(Capacity.firstChild.data)

[toc] | [next] | [standalone]

#70775

> I have managed to read most of the important data in the xml onto lists.
> Now, I have two lists, Source and Destination and I'd like to create bi-directional
> links between them.

> And moreover, I'd like to assign some kind of a bandwidth capacity to the links and
> similarly, storage and processing capacity to the nodes.

I don't know anything about your use case or data really, but from the above I
already know I would shove this all into a database, adding tables with relations
and mining data based on relations becomes trivial. More importantly, it also then
becomes easily extensible without refactoring as your requirements change.

Just my opinion,
jlc

[toc] | [prev] | [next] | [standalone]

#70780

I don't know how to do that stuff in python. Basically, I'm trying to pull certain data from the xml file like the node-name, source, destination and the capacity. Since, I am done with that part, I now want to have a link between source and destination and assign capacity to it.

eg., [a,b,c,d,e] [l,m,n,o,p] [5,2,3,4,5]

I need something like a - l: 5, b - m: 2,....I don't know how to describe this exactly but I hope this would give you a rough idea. After this, I put the stuff back into an xml but this time only there are more parameters and looks more detailed....

Like nodeID = Aachen...stor_capacity = 100, proc_capacity = 200, switch_capacity = 190...etc 

[toc] | [prev] | [next] | [standalone]

#70781

On 4/30/14 9:57 AM, varun7rs@gmail.com wrote:
> I don't know how to do that stuff in python.

    Always a good time to learn.


Let the database do the work for you;  try not to re-invent the 
relational database wheel.  Access the database via python-sql:

https://pypi.python.org/pypi/python-sql/


marcus

[toc] | [prev] | [next] | [standalone]

#70788

> I don't know how to do that stuff in python. Basically, I'm trying to pull certain data from the
> xml file like the node-name, source, destination and the capacity. Since, I am done with that
> part, I now want to have a link between source and destination and assign capacity to it.

I dont mind writing you an SQLite schema and accessor class, can you define your data in a tabular
format and mail it to me offline, we add relationships etc as we go.

Hopefully it inspires you to adopt this approach in the future as it often proves powerful.

jlc

[toc] | [prev] | [next] | [standalone]

#70859

On Wednesday, 30 April 2014 20:38:07 UTC+2, Joseph L. Casale  wrote:
> > I don't know how to do that stuff in python. Basically, I'm trying to pull certain data from the
> > xml file like the node-name, source, destination and the capacity. Since, I am done with that
> > part, I now want to have a link between source and destination and assign capacity to it.
> 
> I dont mind writing you an SQLite schema and accessor class, can you define your data in a tabular
> format and mail it to me offline, we add relationships etc as we go.
> 
> Hopefully it inspires you to adopt this approach in the future as it often proves powerful.
> 
> jlc

Thanks a lot for your help. But, how do I mail you? I can't find your mail id here

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web