ptoal/toallab-automation
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Adding Netbox

Browse Source
This commit is contained in:
Patrick Toal
2019-05-06 00:34:45 -04:00
parent 832502de34
commit 6e2205a046
278 changed files with 12767 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

283
roles/ansible-network.network-engine/docs/directives/parser_directives.md Normal file
View File

@@ -0,0 +1,283 @@
# CLI Parser Directives
The `command_parser` module is a module that can be used to parse the results of
text strings into Ansible facts. The primary motivation for developing the
`command_parser` module is to convert structured ASCII text output (such as
the stdout returned from network devices) into JSON data structures suitable to be
used as host facts.
The parser template file format is loosely based on the Ansible playbook directives
language. It uses the Ansible directive language to ease the transition from
writing playbooks to writing parser templates. However, parser templates developed using this
module are not written directly into the playbook, but are a separate file
called from playbooks. This is done for a variety of reasons but most notably
to keep separation between the parsing logic and playbook execution.
The `command_parser` works based on a set of directives that perform actions
on structured data with the end result being a valid JSON structure that can be
returned to the Ansible facts system.
## Parser language
The parser template format uses YAML formatting, providing an ordered list of directives
to be performed on the content (provided by the module argument). The overall
general structure of a directive is as follows:
```yaml
- name: some description name of the task to be performed
directive:
argument: value
argument_option: value
argument: value
directive_option: value
directive_option: value
```
The `command_parser` currently supports the following top-level directives:
* `pattern_match`
* `pattern_group`
* `json_template`
* `export_facts`
In addition to the directives, the following common directive options are
currently supported:
* `name`
* `block`
* `loop`
* `loop_control`
* `loop_var`
* `when`
* `register`
* `export`
* `export_as`
* `extend`
Any of the directive options are accepted but in some cases, the option may
provide no operation. For instance, when using the `export_facts`
directive, the options `register`, `export` and `export_as` are all
ignored. The module should provide warnings when an option is ignored.
The following sections provide more details about how to use the parser
directives to parse text into JSON structure.
## Directive Options
This section provides details on the various options that are available to be
configured on any directive.
### `name`
All entries in the parser template many contain a `name` directive. The
`name` directive can be used to provide an arbitrary description as to the
purpose of the parser items. The use of `name` is optional for all
directives.
The default value for `name` is `null`.
### `register`
Use the `register` option to register the results of a directive operation
temporarily into the variable name you specify
so you can retrieve it later in your parser template. You use `register` in
a parser template just as you would in an Ansible playbook.
Variables created with `register` alone are not available outside of the parser context.
Any values registered are only available within the scope of the parser activities.
If you want to provide values back to the playbook, you must also define the [export](#export) option.
Typically you will use `register` alone for parsing each individual part of the
command output, then amalgamate them into a single variable at the end of the parser template,
register that variable and set `export: yes` on it.
The default value for `register` is `null`.
<a id="export"></a>
### `export`
Use the `export` option to export any value back to the calling task as an
Ansible fact. The `export` option accepts a boolean value that defines if
the registered fact should be exported to the calling task in the playbook (or
role) scope. To export the value, simply set `export` to True.
Note this option requires the `register` value to be set in some cases and will
produce a warning message if the `register` option is not provided.
The default value for `export` is `False`.
### `export_as`
Use the `export_as` option to export a value back to the calling task as an
Ansible fact in a specific format. The `export_as` option defines the structure of the exported data.
Accepted values for `export_as`:
* `dict`
* `hash`
* `object`
* `list`
* `elements` that defines the structure
**Note** this option requires the `register` value to be set and `export: True`.
Variables can also be used with `export_as`.
How to use variable with `export_as` is as follows:
Variable should be defined in vars or defaults or in playbook.
```yaml
vars:
export_type: "list"
```
Parser file needs to have the variable set to `export_as`.
```
export_as: "{{ export_type }}"
```
### `extend`
Use the `extend` option to extend a current fact hierarchy with the new
registered fact. This will case the facts to be merged and returned as a
single tree. If the fact doesn't previously exist, this will create the entire
structure.
The default value for `extend` is `null`.
### loop
Use the `loop` option to loop over a directive in order to process values.
With the `loop` option, the parser will iterate over the directive and
provide each of the values provided by the loop content to the directive for
processing.
Access to the individual items is the same as it would be for Ansible
playbooks. When iterating over a list of items, you can access the individual
item using the `{{ item }}` variable. When looping over a hash, you can
access `{{ item.key }}` and `{{ item.value }}`.
### `loop_control`
Use the `loop_control` option to specify the name of the variable to be
used for the loop instead of default loop variable `item`.
When looping over a hash, you can access `{{ foo.key }}` and `{{ foo.value }}` where `foo`
is `loop_var`.
The general structure of `loop_control` is as follows:
```yaml
- name: User defined variable
pattern_match:
regex: "^(\\S+)"
content: "{{ foo }}"
loop: "{{ context }}"
loop_control:
loop_var: foo
```
### `when`
Use the `when` option to place a condition on the directive to
decided if it is executed or not. The `when` option operates the same as
it would in an Ansible playbook.
For example, if you only want to perform the match statement
when the value of `ansible_network_os` is set to `ios`, you can apply
the `when` conditional like this:
```yaml
- name: conditionally matched var
pattern_match:
regex: "hostname (.+)"
when: ansible_network_os == 'ios'
```
## Directives
The directives perform actions on the content using regular expressions to
extract various values. Each directive provides some additional arguments that
can be used to perform its operation.
### `pattern_match`
Use the `pattern_match` directive to extract one or more values from
the structured ASCII text based on regular expressions.
The following arguments are supported for this directive:
* `regex`
* `content`
* `match_all`
* `match_greedy`
* `match_until` : Sets a ending boundary for `match_greedy`.
The `regex` argument templates the value given to it so variables and filters can be used.
Example :
```yaml
- name: Use a variable and a filter
pattern_match:
regex: "{{ inventory_hostname | lower }} (.+)"
```
### `pattern_group`
Use the `pattern_group` directive to group multiple
`pattern_match` results together.
The following arguments are supported for this directive:
* `json_template`
* `set_vars`
* `export_facts`
### `json_template`
Use the `json_template` directive to create a JSON data structure based on a
template. This directive will allow you to template out a multi-level JSON
blob.
The following arguments are supported for this directive:
* `template`
**Note**
Native jinja2 datatype (eg. 'int', 'float' etc.) rendering is supported with Ansible version >= 2.7
and jinja2 library version >= 2.10. To enable native jinja2 config add below configuration in active
ansible configuration file.
```
[defaults]
jinja2_native= True
```
Usage example:
```yaml
- set_fact:
count: "1"
- name: print count
debug:
msg: "{{ count|int }}"
```
With jinja2_native configuration enabled the output of above example task will have
```
"msg": 1
```
and with jinja2_native configuration disabled (default) output of above example task will have
```
"msg": "1"
```
### `set_vars`
Use the `set_vars` directive to set variables to the values like key / value pairs
and return a dictionary.
### `export_facts`
Use the `export_facts` directive to take an arbitrary set of key / value pairs
and expose (return) them back to the playbook global namespace. Any key /
value pairs that are provided in this directive become available on the host.

276
roles/ansible-network.network-engine/docs/directives/template_directives.md Normal file
View File

@@ -0,0 +1,276 @@
# CLI Template Directives
**Note** `network_template` lookup plugin is deprecated in v2.7.3 and will be removed
in version v2.7.7 i.e, four releases from the deprecation version.
The `network_template` module supports a number of keyword based objectives that
handle how to process the template. Templates are broken up into a series
of blocks that process lines. Blocks are logical groups that have a common
set of properties in common.
Blocks can also include other template files and are processed in the same
manner as lines. See includes below for a description on how to use the
include directive.
The template module works by processing the lines directives in sequential
order. The module will attempt to template each line in the lines directive
and, if successful, add the line to the final output. Values used for
variable substitution come from the host facts. If the line could not
be successfully templated, the line is skipped and a warning message is
displayed that the line could not be templated.
There are additional directives that can be combined to support looping over
lists and hashes as well as applying conditional statements to blocks, lines
and includes.
## `name`
Entries in the template may contain a `name` field. The `name` field
is used to provide a description of the entry. It is also used to provide
feedback when processing the template to indicate when an entry is
skipped or fails.
## `lines`
The `lines` directive provides an ordered list of statements to attempt
to template. Each entry in the `lines` directive will be evaluated for
variable substitution. If the entry can be successfully templated, then the
output will be added to the final set of entries. If the entry cannot be
successfully templated, then the entry is ignored (skipped) and a warning
message is provided. If the entry in the `lines` directive contains
only static text (no variables), then the line will always be processed.
The `lines` directive also supports standard Jinja2 filters as well as any
Ansible specific Jinja2 filters. For example, lets assume we want to add a
default value if a more specific value was not assigned by a fact.
```yaml
- name: render the system hostname
lines:
- "hostname {{ hostname | default(inventory_hostname_short }}"
```
## `block`
A group of `lines` directives can be combined into a `block`
directive. These `block` directives are used to apply a common set of
values to one or more `lines` or `includes` entries.
For instance, a `block` directive that contains one or more `lines`
entries could be use the same set of `loop` values or have a
common `when` conditional statement applied to them.
## `include`
Sometimes it is advantageous to break up templates into separate files and
combine them. The `include` directive will instruct the current template
to load another template file and process it.
The `include` directive also supports variable substitution for the
provided file name and can be processed with the `loop` and `when`
directives.
## `when`
The `when` directive allows for conditional statements to be applied to
a set of `lines`, a `block` and/or the `include` directive. The
`when` statement is evaluated prior to processing the statements and if
the condition is true, the statements will attempt to be templated. If the
statement is false, the statements are skipped and a message returned.
## `loop`
Depending on the input facts, sometimes it is necessary to iterate over a
set of statements. The `loop` directive allows the same set of statements
to be processed in such a manner. The `loop` directive takes, as input,
the name of a fact that is either a list or a hash and iterates over the
statements for each entry.
When the provided fact is a list of items, the value will be assigned to a
variable called `item` and can be referenced by the statements.
When the provided fact is a hash of items, the hash key will be assigned to
the `item.key` variable and the hash value will be assigned to the
`item.value` variable. Both can then be referenced by the statements.
## `loop_control`
The `loop_control` directive allows the template to configure aspects
related to how loops are process. This directive provides a set of suboptions
to configure how loops are processed.
### `loop_var`
The `loop_var` directive allows the template to override the default
variable name `item`. This is useful when handling nested loops such
that both inner and outer loops values can be accessed.
When setting the `loop_var` to some string, the string will replace
`item` as the variable name used to access the values.
For example, lets assume instead of using item, we want to use a different
variable name such as entry:
```yaml
- name: render entries
lines:
- "hostname {{ entry.hostname }}"
- "domain-name {{ entry.domain_name }}"
loop: "{{ system }}"
loop_control:
loop_var: entry
```
## `join`
When building template statements that include optional values for
configuration, the `join` directive can be useful. The `join`
directive instructs the template to combine the templated lines together
into a single string to insert into the configuration.
For example, lets assume there is a need to add the following statement to
the configuration:
```
ip domain-name ansible.com vrf management
ip domain-name redhat.com
```
To support templating the above lines, the facts might include the domain-name
and the vrf name values. Here is the example facts:
```yaml
---
system:
- domain_name: ansible.com
vrf: management
- domain_name redhat.com
```
And the template statement would be the following:
```yaml
- name: render domain-name
lines:
- "ip domain-name {{ item.domain_name }}"
- "vrf {{ item.vrf }}"
loop: "{{ system }}"
join: yes
```
When this entry is processed, the first iteration will successfully template
both lines and add `ip domain-name ansible.com vrf management` to the
output.
When the second entry is processed, the first line will be successfully
templated but since there is no management key, the second line will return a
null value. The final line added to the configuration will be ` ip
domain-name redhat.com`.
If the `join` directive had been omitted, then the final set of
configuration statements would be as follows:
```
ip domain-name ansible.com
vrf management
ip domain-name redhat.com
```
## `join_delimiter`
When the `join` delimiter is used, the templated values are combined into a
single string that is added to the final output. The lines are joined using a
space. The delimiting character used when processing the `join` can be
modified using `join_delimiter` directive.
Here is an example of using the this directive. Take the following entry:
```yaml
- name: render domain-name
lines:
- "ip domain-name {{ item.domain_name }}"
- "vrf {{ item.vrf }}"
loop: "{{ system }}"
join: yes
join_delimiter: ,
```
When the preceding statements are processed, the final output would be
(assuming all variables are provided):
```
ip domain-name ansible.com,vrf management
ip domain-name redhat.com
```
## `indent`
The `indent` directive is used to add one or more leading spaces to the
final templated statement. It can be used to recreated a structured
configuration file.
Take the following template entry as an example:
```yaml
- name: render the interface context
lines: "interface Ethernet0/1"
- name: render the interface configuration
lines:
- "ip address 192.168.10.1/24"
- "no shutdown"
- "description this is an example"
indent: 3
- name: render the interface context
lines: "!"
```
Then the statements above are processed, the output will look like the
following:
```
interface Ethernet0/1
ip address 192.168.10.1/24
no shutdown
description this is an example
!
```
## `required`
The `required` directive specifies that all of the statements must be
templated otherwise a failure is generated. Essentially it is a way to
make certain that the variables are defined.
For example, take the following:
```yaml
- name: render router ospf context
lines:
- "router ospf {{ process_id }}"
required: yes
```
When the above is processed, if the variable `process_id` is not present,
then the statement cannot be templated. Since the `required` directive
is set to true, the statement will cause the template to generate a failure
message.
## `missing_key`
By default, when statements are processed and a variable is undefined, the
module will simply display a warning message to the screen. In some cases, it
is desired to either suppress warning messages on a missing key or to force the
module to fail on a missing key.
To change the default behaviour, use the `missing_key` directive. This
directive accepts one of three choices:
* `ignore`
* `warn` (default)
* `fail`
The value of this directive will instruct the template how to handle any
condition where the desired variable is undefined.