pyang.md

Pyang

Pyangbind is simply a parser for all openconfig models. Pyang will turn actual YANG models into python objects ie lists, tupples and dictionaries.

Lets take a look at the bgp model.

sh-3.2# docker exec -it mgt1 bash
pyang -f tree /models/bgp/*
root@b6efac7b7b0f:/# cd /tmp/scripts/
module: openconfig-bgp
  +--rw bgp
     +--rw global
     |  +--rw config
     |  |  +--rw as           oc-inet:as-number
     |  |  +--rw router-id?   oc-yang:dotted-quad
     |  +--ro state
     |  |  +--ro as           oc-inet:as-number
     |  |  +--ro router-id?   oc-yang:dotted-quad
     |  +--rw default-route-distance

the bgp_example.py is a good way to reflect the model and how it interacts with the python code.

ocbgp = ocbind.openconfig_bgp(path_helper=ph)
ocbgp.bgp.global_.config.as_ = 1
ocbgp.bgp.global_.config.router_id = "1.1.1.1"
struct =  json.dumps(ocbgp.get(filter=True), indent=4)

Lets reflect upon those and compare then to the module above it.

ocbgp.bgp.global_.config.as_ = 1 #This has to be a oc-inet which is a integer

ocbgp.bgp.global_.config.router_id = "1.1.1.1" #This has to be a dotted-quad which is a ipv4 address

If either are changed to something that is not the same as the unit should be pyang will bomb and not work properly.

Lets look at a more complicated example bgp_config.py which will take a python dictionary and turn it into code.

root@b6efac7b7b0f:/tmp/scripts# python bgp_config.py
{
    "bgp": {
        "global": {
            "config": {
                "as": 65002,
                "router-id": "2.2.2.2"
            }
        },
        "neighbors": {
            "neighbor": {
                "12.12.12.1": {
                    "config": {
                        "neighbor-address": "12.12.12.1",
                        "peer-as": 65001
                    },
                    "neighbor-address": "12.12.12.1"
                }
            }
        }
    }
}
{
    "bgp": {
        "global": {
            "config": {
                "as": 65001,
                "router-id": "1.1.1.1"
            }
        },
        "neighbors": {
            "neighbor": {
                "12.12.12.2": {
                    "config": {
                        "neighbor-address": "12.12.12.2",
                        "peer-as": 65002
                    },
                    "neighbor-address": "12.12.12.2"
                }
            }
        }
    }
}

Either way, the key point here is that pynag can parse through and add yang models to code. The result would be importing any module into python code and then sending it directly to the device.