Data Validation Plugins

Data validation plugins responsible for validating loaded data to make sure it adheres to model or schema requirements.

By default TTR uses yangson plugin for data validation, alternatively TTR object validator attribute can be used to specify plugin to use.

Validation step is optional, if no models provided, data not validated. However, if required data models can be used to make sure that correct data provided prior to performing rendering step.

There are two types of validation plugins:

  1. Plugins to load models

  2. Plugins to validate the actual data

Yangson Models Loader

Reference name yangson

This plugin loads YANG models into yangson DataModel objects.

YANG models must sit within their own directories, each such a directory used to create JSON library for Yangson to load models from.

Directory name, main YANG model file name and module name must be the same, directory name used as a reference name for the model.

For example, this is directory tree with YANG models inside:

|-- Models
    |-- interface
        |-- ietf-inet-types@2013-07-15.yang
        |-- interface.yang
    |-- vrf
        |-- vrf.yang

Above directory structure translated to two models named interface and vrf, these names can be used to reference models in data for validation, e.g.:

- interface: Gi1/1
  description: Customer A
  vid: 100
  ip: 10.0.0.1
  mask: 255.255.255.0
  vrf: cust_a
  device: R1
  template: interfaces.cisco_ios
  model: interface # YANG model name to validate this data item

For reference, YANG model Models/interface/interface.yang file content is:

module interface {
    yang-version "1.1";

    namespace "http://ttr/test-1";

    import ietf-inet-types {
    prefix inet;
    }

    typedef ipmask {
    type string {
        pattern '([0-9]{1,3}.){3}[0-9]{1,3}';
    }
    description
        "Pattern to match strings like 255.255.255.0 or 255.0.0.0";
    }

    prefix "ttr";

    leaf interface {
        mandatory true;
        type string;
    }
    leaf template {
        mandatory true;
        type string;
    }
    leaf device {
        mandatory true;
        type string;
    }
    leaf description{
        type string;
    }
    leaf vid {
        type int32;
    }
    leaf ip {
        type inet:ipv4-address;
    }
    leaf mask {
        type ipmask;
    }
    leaf vrf {
        type string;
    }
}
ttr.plugins.models.yangson_model_loader.load(models_dict, models_dir, **kwargs)

Creates JSON-encoded YANG library data [RFC7895] and instantiates data model object out of it.

Parameters

  • models_dir – (str) OS path to directory with YANG models modules subdirectories, each subdirectory models loaded to form single DataModel and added to models_dict under directory name key.

  • models_dict – (dict) dictionary to store loaded model object at

  • kwargs – (dict) any additional arguments ignored

  • return – None

Yangson Data Validation

Reference name yangson

This plugin relies on Yangson library for data instance validation using YANG models.

ttr.plugins.validate.validate_yangson.validate(data, model_content, model_name, validation_scope='all', content_type='all', on_fail='raise')

Validate data for compliance with YANG modules.

Parameters

  • data – (dict) dictionary data to validate

  • model_content – (obj) Fully instantiated Yangson DataModel object

  • model_name – (str) name of the model

  • content_type – (str) optional, content type as per https://yangson.labs.nic.cz/enumerations.html supported - all, config, nonconfig

  • validation_scope – (str) optional, validation scope as per https://yangson.labs.nic.cz/enumerations.html supported - all, semantics, syntax

  • on_fail – (str) action to do if validation fails - raise (default) or log

Returns:

  • True if validation succeeded

  • False if validation failed and on_fail is “log”

  • Raises RuntimeError exception if validation failed and on_fail is “raise”