Using regex_search to extract a string between a string and a whitespace - regex

Question

I would like to extract the version number 1.0.1 from the following string in Ansible, I tried using regex_serach to extra the string between ide- and a whitespace, but that's what I end up getting ide-1.0.1 ; argv[]=/home/bin/ide-1.0.1 start. I tried \s instead of \\s and it doesn't work. How should I fix the regex pattern? Any help would be appreciated!
ExecStart={ path=/home/bin/ide-1.0.1 ; argv[]=/home/bin/ide-1.0.1 start }
- name: Check if ide is active
command: systemctl show ide.service --property=ExecStart
register: version_check
ignore_errors: yes
- name: Set fact
set_fact:
version: "{{ version_check.stdout | regex_search('ide-(.*)\\s'}}"
- name: Debug version
debug:{{ version }}"

Answer 1

Use regex_replace, e.g.
- set_fact:
version: "{{ _path.split('-').1 }}"
vars:
_regex: '^(.*)=(.*)=(.*);(.*)$'
_replace: '\3'
_path: "{{ version_check.stdout|regex_replace(_regex, _replace)|trim }}"
gives
version: 1.0.1
Parsing semi-structured text with Ansible is available, e.g.
- ansible.utils.cli_parse:
text: "{{ version_check.stdout }}"
parser:
name: ansible.netcommon.native
template_path: templates/property.yaml
set_fact: property
and the template
shell> cat templates/property.yaml
---
- example: 'ExecStart={ path=/home/bin/ide-1.0.1 ; argv[]=/home/bin/ide-1.0.1 start }'
getval: '(?P<property>\S+)={\s*(?P<key1>\S+)=(?P<val1>\S+)\s*;\s*(?P<key2>\S+)=(?P<val2>\S+)\s*(?P<state>\S+)\s*}'
result:
"{{ property }}":
"{{ key1 }}": "{{ val1 }}"
"{{ key2 }}": "{{ val2 }}"
state: "{{ state }}"
give
property:
ExecStart:
argv[]: /home/bin/ide-1.0.1
path: /home/bin/ide-1.0.1
state: start
Then parse the version, e.g.
- set_fact:
version: "{{ (property.ExecStart.path|basename).split('-').1 }}"
gives
version: 1.0.1