Added on the filter usecase for Ansible

This commit is contained in:
2021-09-24 12:12:43 +02:00
parent d02cefcdd0
commit d64472cd62

View File

@@ -20,6 +20,8 @@ ansible all -l localhost -m systemd -a "name=nginx state=restarted" -u test --be
`sample.result_list` is a list of lists containing stdout output. We want to filter some numbers from this output. To do this we first filter the correct lines by using the `json_query` filter. If you need to match a specific substring, you could use `select("match", "<regex>")` in combination with the `json_query`. Next we `map` the `regex_search` filter on all lines to extract the data we need. As the `map` filter is a Python generator object, we need to convert it to a list.
The regex `(?<=\()\d+` uses a feature called *positive lookbehind*: `(?<=\()`. It matches a group before the main expression with including it in the result. In this case it matches the opening parantheses `(`. This feature may not be supported in your regex implementation, but in Ansible/Python it is supported. To make the regex even more specific you could add a *positive lookahead* (`(?=\))`) after the `\d+` to match the closing parantheses `)`.
```yaml
---
@@ -37,5 +39,5 @@ ansible all -l localhost -m systemd -a "name=nginx state=restarted" -u test --be
tasks:
- debug:
msg: '{{ sample.results | json_query("[*].stdout_lines[1]") | map("regex_search" ,"(?<=\()\d*") | list }}'
msg: '{{ sample.results | json_query("[*].stdout_lines[1]") | map("regex_search" ,"(?<=\()\d+") | list }}'
```