unable to extract version out of registered output

Archives Ansible Project
1

Hello Team,

Code

name: debug grepout

debug:

msg: “this is the version {{ vername.files.path }}”

var: files_found_path

tags: vern

Output:

files_found_path": [

“/opt/confluent-7.2.5/kafka/share/java/kafka/kafka_2.13-7.2.5-ccs.jar”

]

Above code is working but after that it failing with below code, i need 7.2.5 before -css in above output but failing to get that tried files_found_path[0].split(‘-‘) as well

  • name: get version

set_fact:

versionn: “{{ files_found_path.split(‘-‘)[0] }}”

TASK [get version] ***************************************************************************************************************************************************************************

fatal: [host1: FAILED! => {“msg”: “The task includes an option with an undefined variable. The error was: ‘list object’ has no attribute ‘split’. ‘list object’ has no attribute ‘split’\n\nThe error appears to be in ‘/Users/sameer_modak/ansibledemo/testing.yml’: line 54, column 5, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n - name: get version\n ^ here\n”}

fatal: [host2]: FAILED! => {“msg”: “The task includes an option with an undefined variable. The error was: ‘list object’ has no attribute ‘split’. ‘list object’ has no attribute ‘split’\n\nThe error appears to be in ‘/Users/sameer_modak/ansibledemo/testing.yml’: line 54, column 5, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n - name: get version\n ^ here\n”}

fatal: [host3]: FAILED! => {“msg”: "The task includes an option with an undefined variable. The error was: ‘list object’ has no attribute ‘split’. ‘list object’ has no a

Thanks in advance.

2

files_found_path is an array/list. So you need files_found_path[0].split

3

i did try that as well but no luck as i have already mentioned in my previous conversation.

below is the complete code, output i want is 7.2.5 from jar line. thats it. But i m missing something may taking long route when simple solution is there

  • name: take command grep output

shell:

cmd: “ps -ef | grep -i kafka | grep -i server.prop | awk -F’-cp’ ‘{print $2}’ | awk -F’:’ ‘{print $1}’”

register: grepout

tags: verchk1

  • name: setfact and pring

set_fact:

pathd: “{{ (grepout.stdout.split(‘\n’)[0] | replace(‘bin/…/’, ‘’) | replace(‘*’,‘’) | replace(’ /‘,’/')) }}”

pathd: “{% if ‘kafka’ in (grepout.stdout.split(‘\n’)[0] | replace(‘bin/…/’, ‘’) | replace(‘*’,‘’) | replace(’ /‘,’/')) %}‘hi’{%else}‘bye’{% endif %}}”

- pathd1: “{% if ‘kafka’ in pathd %}pathd{% else %}‘’{% endif %}”

  • name: setfact for 2nd var

set_fact:

pathd1: “{% if ‘kafka’ in pathd %}{{pathd}}{% else %}‘’{% endif %}”

  • name: pathd ls

find:

paths: “{{pathd1}}”

patterns: “kafka_*ccs.jar”

use_regex: true

register: vername

  • name : get the complete path

set_fact:

files_found_path: “{{ vername.files.path[1] }}”

files_found_path: “{{ vername.files | map(attribute=‘path’) | list }}”

  • name: debug grepout

debug:

msg: “this is the version {{ vername.files.path }}”

var: files_found_path

tags: vern

- name: get version

set_fact:

versionn: “{{ files_found_path.split(‘-’)[1] }}”

- name: debug grepout

debug:

msg: “this is the version {{ vername.files.path }}”

var: versionn

tags: vern

4

Assuming you are getting ‘files_found_path’ as

TASK [debug] ****************************************************************************
ok: [localhost] => {
“msg”: [
“/opt/confluent-7.2.5/kafka/share/java/kafka/kafka_2.13-7.2.5-ccs.jar”
]
}

I would write my playbook as -

  • debug:
    msg: “{{ files_found_path[0] | regex_replace(kafka_ver_regex, ‘\2’) }}”
    vars:
    kafka_ver_regex: (.kafka_\d.\d\d-)(\d.\d.\d)(-..jar)

That would give me -

TASK [debug] ****************************************************************************
ok: [localhost] => {
“msg”: “7.2.5”
}

5

Thanks (faar aabhari) Abhijeet for your reply. but also i have to store that in variable so i ll have to use set_fact ryt or is there any simpler way in one line because if you see my code i had to use so many set_fact to 1st get one value then again store in another value.

One more point i have to highlight is that when i get output of my shell command jar path is not always the 1st item in list it sometimes 2nd so this is bit annoying .

for example,

  • name: take command grep output

shell:

cmd: “ps -ef | grep -i kafka | grep -i server.prop | awk -F’-cp’ ‘{print $2}’ | awk -F’:’ ‘{print $1}’”

register: grepout

tags: verchk1

here output is not always constant

sometimes

“grepout.stdout”: " /usr/bin/…/share/java/kafka/*\n’ ‘{print $2}’ | awk -F’"

and sometimes

“grepout.stdout”: ‘{print $2}’ | awk -F’" \n’ " /usr/bin/…/share/java/kafka/*"

This creates problems

6

No problem,

You can use set_fact instead of debug task.

When you get a list of elements from the shell command, you can use loop or with_items to iterate over them and set the fact.