New in version 2019.2.0.
This is an Execution Module providing helpers for various NAPALM formulas, e.g., napalm-interfaces-formula, napalm-bgp-formula, napalm-ntp-formula etc., meant to provide various helper functions to make the templates more readable.
Return the list of all the possible paths in a container, down to the
config
container.
This function can be used to verify that the model
is a Python object
correctly structured and respecting the OpenConfig hierarchy.
The OpenConfig-structured object to inspect.
:
The key delimiter. In particular cases, it is indicated to use //
as :
might be already used in various cases, e.g., IPv6 addresses,
interface name (e.g., Juniper QFX series), etc.
CLI Example:
salt '*' napalm_formula.container_path "{'interfaces': {'interface': {'Ethernet1': {'config': {'name': 'Ethernet1'}}}}}"
The example above would return a list with the following element:
interfaces:interface:Ethernet1:config
which is the only possible path
in that hierarchy.
Other output examples:
- interfaces:interface:Ethernet1:config
- interfaces:interface:Ethernet1:subinterfaces:subinterface:0:config
- interfaces:interface:Ethernet2:config
Apply the defaults to a Python dictionary having the structure as described in the OpenConfig standards.
The OpenConfig model to apply the defaults to.
The dictionary of defaults. This argument must equally be structured with respect to the OpenConfig standards.
For ease of use, the keys of these support glob matching, therefore we don't have to provide the defaults for each entity but only for the entity type. See an example below.
//
The key delimiter to use. Generally, //
should cover all the possible
cases, and you don't need to override this value.
False
Whether should merge the model into the defaults, or the defaults
into the model. Default: False
(merge the model into the defaults,
i.e., any defaults would be overridden by the values from the model
).
CLI Example:
salt '*' napalm_formula.defaults "{'interfaces': {'interface': {'Ethernet1': {'config': {'name': 'Ethernet1'}}}}}" "{'interfaces': {'interface': {'*': {'config': {'enabled': True}}}}}"
As one can notice in the example above, the *
corresponds to the
interface name, therefore, the defaults will be applied on all the
interfaces.
Recursive version of the default dict.update
Merges upd recursively into dest
If recursive_update=False, will use the classic dict.update, or fall back
on a manual merge (helpful for non-dict types like FunctionWrapper
).
If merge_lists=True
, will aggregate list object types instead of replace.
The list in upd
is added to the list in dest
, so the resulting list
is dest[key] + upd[key]
. This behaviour is only activated when
recursive_update=True
. By default merge_lists=False
.
Render a field found under the field
level of the hierarchy in the
dictionary
object.
This is useful to render a field in a Jinja template without worrying that
the hierarchy might not exist. For example if we do the following in Jinja:
{{ interfaces.interface.Ethernet5.config.description }}
for the
following object:
{'interfaces': {'interface': {'Ethernet1': {'config': {'enabled': True}}}}}
it would error, as the Ethernet5
key does not exist.
With this helper, we can skip this and avoid existence checks. This must be
however used with care.
The dictionary to traverse.
The key name or part to traverse in the dictionary
.
None
The text to prepend in front of the text. Usually, we need to have the name of the field too when generating the configuration.
None
Text to append at the end.
False
Whether should wrap the text around quotes.
CLI Example:
salt '*' napalm_formula.render_field "{'enabled': True}" enabled
# This would return the value of the ``enabled`` leaf key
salt '*' napalm_formula.render_field "{'enabled': True}" description
# This would not error
Jinja usage example:
{%- set config = {'enabled': True, 'description': 'Interface description'} %}
{{ salt.napalm_formula.render_field(config, 'description', quotes=True) }}
The example above would be rendered on Arista / Cisco as:
description "Interface description"
While on Junos (the semicolon is important to be added, otherwise the configuration won't be accepted by Junos):
description "Interface description";
This function works similarly to
render_field
but for a
list of fields from the same dictionary, rendering, indenting and
distributing them on separate lines.
The dictionary to traverse.
A list of field names or paths in the dictionary.
0
The indentation to use, prepended to the rendered field.
\n
The separator to use between fields.
CLI Example:
salt '*' napalm_formula.render_fields "{'mtu': 68, 'description': 'Interface description'}" mtu description
Jinja usage example:
{%- set config={'mtu': 68, 'description': 'Interface description'} %}
{{ salt.napalm_formula.render_fields(config, 'mtu', 'description', quotes=True) }}
The Jinja example above would generate the following configuration:
mtu "68"
description "Interface description"
Set a value under the dictionary hierarchy identified under the key. The target 'foo/bar/baz' returns the dictionary hierarchy {'foo': {'bar': {'baz': {}}}}.
Note
Currently this doesn't work with integers, i.e. cannot build lists dynamically.
CLI Example:
salt '*' formula.setval foo:baz:bar True
Traverse a dict or list using a colon-delimited (or otherwise delimited,
using the delimiter
param) target string. The target foo:bar:0
will
return data['foo']['bar'][0]
if this value exists, and will otherwise
return the dict in the default argument.
Function will automatically determine the target type.
The target foo:bar:0
will return data['foo']['bar'][0] if data like
{'foo':{'bar':['baz']}}
, if data like {'foo':{'bar':{'0':'baz'}}}
then return data['foo']['bar']['0']
CLI Example:
salt '*' napalm_formula.traverse "{'foo': {'bar': {'baz': True}}}" foo:baz:bar